This manual describes the operation of the XBee® /XBee-PRO® SMT ZB RF module, which consists of ZigBee
firmware loaded onto XBee® S2C and S2C PRO hardware. The XBee® /XBee-PRO® SMT ZB RF Modules are
designed to operate within the ZigBee protocol and support the unique needs of low-cost, low-power wireless
sensor networks. The modules require minimal power and provide reliable delivery of data between remote
devices. The modules operate within the ISM 2.4 GHz frequency band.
Digi International Inc.
11001 Bren Road East
Minnetonka, MN 55343 877 912-3444 or 952 912-3444
No part of the contents of this manual may be transmitted or reproduced in any form or by any means without the
written permission of Digi International, Inc.
ZigBee® is a registered trademark of the ZigBee Alliance.
XBee® and XBee-PRO® are registered trademarks of Digi International, Inc.
XBee® and XBee-PRO® S2C SMT ZB embedded RF modules provide wireless connectivity to end-point
devices in ZigBee mesh networks. Utilizing the ZigBee PRO Feature Set, these modules are interoperable
with other ZigBee devices, including devices from other vendors. With XBee, users can have their ZigBee
network up-and-running in a matter of minutes without configuration or additional development. The
programmable XBee® and XBee-PRO® S2C SMT ZB modules incorporate a Freescale SO8 microprocessor
for customization and application development.
XBee® and XBee-PRO® ZB Modules are compatible with other devices that use XBee® “ZB" technology.
These include ConnectPort X gateways, XBee® and XBee-PRO® Adapters, XBee Wall Routers, XBee
Sensors, and other products when designated with the "ZB" product name. Devices that do not have the
"ZB" product name, including Digi's line of DigiMesh and 802.15.4 XBee products, are not compatible with
XBee® and XBee-PRO® ZB Modules.
Network interoperability with ZigBee devices from other vendors requires that the ZigBee Feature Set or
ZigBee PRO Feature Set be deployed on all devices. Contact Digi Support for details.
Dimensions 0.866 X 1.3" (2.199 X 3.302 cm)
Operating Temperature -40 to 85° C (Industrial)
Antenna Options RF Pad, PCB Antenna, or U.FL Connector
RF Specifications
Specification
Performance
Frequency ISM 2.4-2.5GHz
Number of Channels 16 Direct Sequence Channels 16
Channels 11 to 26 11 to 26
Adjustable Power Yes
Interface immunity
Indoor/Urban Range 200 ft 300 Ft
The XBee® /XBee-PRO® SMT ZB RF modules have 16 GPIO (General Purpose Input Output) ports available.
Those available will depend on the module configuration as some GPIO pads are consumed by serial
communication, etc.
See GPIO section for more information on configuring and using GPIO ports
Electrical Specification for GPIO pads
Specification XBee (S2C) XBee-PRO (S2C)
Performance Module Pin Number
Voltage Supply 2.1 to 3.6V
Low Schmitt switching
threshold
threshold
Input current for logic 0
Input current for logic 1 0.5uA
Input pull-up resistor value 29kΩ
Industry Canada (IC) IC: 1846A-XBS2C IC: 1846A-XBPS2C
Europe (DC) ETSI
Australia C-Tick C-Tick
Japan Pending
FCC Approval (USA) Refer to Chapter 12 FCC Requirements. Systems that contain XBee®/ XBee-PRO® ZB RF Modules inherit Digi
Certifications.
Hardware Specifications for Programmable Variant
The following specifications need to be added to the current measurement of the previous table if the module
has the programmable secondary processor. For example, if the secondary processor is running and constantly
collecting DIO samples at a rate while having the RF portion of the XBEE sleeping the new current will be I total
= Ir2 + Is. Where Ir2 is the runtime current of the secondary processor and is the sleep current of the RF
portion of the module of the XBEE-PRO (S2B) listed in the table below.
Specifications of the programmable secondary processor
NOTE: Some lines may not go to the external XBEE pads in the programmable secondary processor version
Design Notes
The XBee modules do not specifically require any external circuitry or specific connections for proper
operation. However, there are some general design guidelines that are recommended for help in
troubleshooting and building a robust design.
Power Supply
Poor power supply can lead to poor radio performance especially if the supply voltage is not kept within
tolerance or is excessively noisy. To help reduce noise a 1uF and 8.2pF capacitor are recommended to be
placed as near to pin1 on the PCB as possible. If using a switching regulator for your power supply, switching
frequencies above 500 kHz are preferred. Power supply ripple should be limited to a maximum 250mV peak to
peak.
Note – For designs using the programmable modules an additional 10uF decoupling cap is recommended near
pin 1 of the module. The nearest proximity to pin 1 of the 3 caps should be in the following order: 8.2pf, 1uF
followed by 10uF.
Recommended Pin Connections
The only required pin connections are VCC, GND, DOUT and DIN. To support serial firmware updates, VCC,
GND, DOUT, DIN, RTS, and DTR should be connected.
All unused pins should be left disconnected. All inputs on the radio can be pulled high with 30k internal pull-up
resistors using the PR software command. No specific treatment is needed for unused outputs.
For applications that need to ensure the lowest sleep current, inputs should never be left floating. Use internal
or external pull-up or pull-down resistors, or set the unused I/O lines to outputs.
Other pins may be connected to external circuitry for convenience of operation including the Associate LED
pad (pad 28) and the Commissioning pad (pad 33). The Associate LED pin will flash differently depending on
the state of the module to the network, and a pushbutton attached to pad 33 can enable various join
functions without having to send UART commands. Please see the commissioning pushbutton and associate
LED section in chapter 7 for more details. The source and sink capabilities are limited to 4mA for pad numbers
3, 4, 5, 10, 12, 14, 15, 16, 17, 25, 26, 28, 29, 30, and 32, and 8mA for pad numbers 7, 8, 24, 31, and 33 on the
module.
The VREF pin (pad 27) is not used on this module. For compatibility with other XBee modules, we recommend
connecting this pin to a voltage reference if analog sampling is desired. Otherwise, connect to GND.
Board Layout
XBee modules do not have any specific sensitivity to nearby processors, crystals or other PCB components.
Other than mechanical considerations, no special PCB placement is required for integrating XBee radios except
for those with integral antennas. In general, Power and GND traces should be thicker than signal traces and be
able to comfortably support the maximum currents.
The radios are also designed to be self sufficient and work with the integrated and external antennas without
the need for additional ground planes on the host PCB. However, considerations should be taken on the
choice of antenna and antenna location. Metal objects that are near an antenna cause reflections and may
reduce the ability for an antenna to efficiently radiate. Using an integral antenna in an enclosed metal box will
greatly reduce the range of a radio. For this type of application an external antenna would be a better choice.
External antennas should be positioned away from metal objects as much as possible. Metal objects next to
the antenna or between transmitting and receiving antennas can often block or reduce the transmission
distance. Some objects that are often overlooked are metal poles, metal studs or beams in structures,
concrete (it is usually reinforced with metal rods), metal enclosures, vehicles, elevators, ventilation ducts,
refrigerators and microwave ovens.
Antennas should reside above or away from any metal objects like batteries, tall electrolytic capacitors or
metal enclosures. Antenna elements radiate perpendicular to the direction they point. Thus a vertical
antenna emits across the horizon.
PCB Antennas should not have any ground planes or metal objects above or below the module at the antenna
location. For best results the module should be in a plastic enclosure, instead of metal one. It should be placed
at the edge of the PCB to which it is mounted. The ground, power and signal planes should be vacant
immediately below the antenna section (See drawing for recommended keep out area).
The S2C modules with the programmable option have a secondary processor with 32k of flash and 2k of RAM.
This allows module integrators to put custom code on the XBEE module to fit their own unique needs. The
DIN, DOUT, RTS, CTS, and RESET lines are intercepted by the secondary processor to allow it to be in control of
the data transmitted and received. All other lines are in parallel and can be controlled by either the EM357 or
the MC9SO8QE micro (see Block Diagram for details). The EM357 by default has control of certain lines. These
lines can be released by the EM357 by sending the proper command(s) to disable the desired DIO line(s) (see
XBEE Command Reference Tables).
In order for the secondary processor to sample with ADCs, the XBEE pin 27 (VREF) needs to be connected to a
reference voltage.
Digi provides a bootloader that can take care of programming the processor over the air or through the serial
interface. This means that over the air updates can be supported through an XMODEM protocol. The
processor can also be programmed and debugged through a one wire interface BKGD (Pin 9).
The XBee Programmable module is equipped with a Freescale MC9S08QExx application processor. This
application processor comes with a supplied bootloader. The following section describes how to interface the
customer's application code running on this processor to the XBee Programmable module's supplied
bootloader. This section discusses how to initiate firmware updates using the supplied bootloader for wired
and over-the-air updates.
Bootloader Software Specifics
Memory Layout
Figure 1 shows the memory map for the MC9S08QE32 application processor.
The supplied bootloader occupies the bottom pages of the flash from 0xF200 to 0xFFFF. Application code
cannot write to this space.
The application code can exist in Flash from address 0x8400 to 0xF1BC. 1k of Flash from 0x8000 to 0x83FF is
reserved for Non Volatile Application Data that will not be erased by the bootloader during a flash update.
A portion of RAM is accessible by both the application and the bootloader. Specifically, there is a shared data
region used by both the application and the bootloader that is located at RAM address 0x200 to 0x215.
Application code should not write anything to AppResetCause or BLResetCause unless informing the
bootloader of the impending reset reason.
Upon reset of any kind, the execution control begins with the bootloader.
If the reset cause is Power-On reset (POR), Pin reset (PIN), or Low Voltage Detect(LVD) reset the bootloader
will not jump to the application code if the override bits are set to RTS(D7)=1, DTR(D5)=0, and DIN(B0)=0.
Otherwise, the bootloader writes the reset cause "NOTHING" to the shared data region, and jumps to the
Application.
Reset causes are defined in the file common. h in an enumeration with the following definitions:
// 0x0000 to 0x00FF are considered valid for APP use.
APP_CAUSE_USE255 = 0x00FF,
APP_CAUSE_FIRMWARE_UPDATE = 0x5981,
APP_CAUSE_BYPASS_MODE = 0x4682,
APP_CAUSE_BOOTLOADER_MENU = 0x6A18,
} APP_RESET_CAUSES;
Otherwise, if the reset cause is a "watchdog" or other reset, the bootloader checks the shared memory region for
the APP_RESET_CAUSE. If the reset cause is:
1. "APP_CAUSE_NOTHING" or 0x0000 to 0x00FF, the bootloader increments the BL_RESET_CAUSES,
verifies that it is still less than BL_CAUSE_BAD_APP, and jumps back to the application. If the
Application does not clear the BL_RESET_CAUSE, it can prevent an infinite loop of running a bad
application that continues to perform illegal instructions or watchdog resets.
2. "APP_CAUSE_FIRMWARE_UPDATE", the bootloader has been instructed to update the application
"over-the-air" from a specific 64 bit address. In this case, the bootloader will attempt to initiate an
Xmodem transfer from the 64 bit address located in Shared RAM.
3. "APP_CAUSE_BYPASS_MODE", the bootloader executes bypass mode. This mode passes the local
UART data directly to the EM250 allowing for direct communication with the EM250. The only way
to exit bypass mode is to reset or power cycle the module. If none of the above is true, the
bootloader will enter "Command mode". In this mode, users can initiate firmware downloads both
wired and over-the-air, check application/bootloader version strings, and enter Bypass mode.
If none of the above is true, the bootloader will enter "Command mode". In this mode, users can initiate firmware
downloads both wired and over-the-air, check application/bootloader version strings, and enter Bypass mode.
Application Version String
Figure 1 shows an "Application version string pointer" area in application flash which holds the pointer to where
the application version string resides. The application's linker command file ultimately determines where this
string is placed in application flash.
It is preferable that the application version string be located at address 0x8400 for MC9S08QE32 parts. The
application string can be any characters terminated by the NULL character (0x00). There is not a strict limit on the
number of characters in the string, but for practical purposes should be kept under 100 bytes including the
terminating NULL character. During an update the bootloader erases the entire application from 0x8400 on. The
last page has the vector table specifically the redirected reset vector. The version string pointer and reset vector
are used to determine if the application is valid.
Application Interrupt Vector Table and Linker Command File
Since the bootloader flash region is read-only, the interrupt vector table is redirected to the region 0xF1C0 to
0xF1FD so that application developers can use hardware interrupts. Note that in order for Application interrupts to
function properly, the Application's linker command file (*.prm extension) must be modified appropriately to allow
the linker to place the developers code in the correct place in memory. For example, the developer desires to use
the serial communications port SCI1 receive interrupt. The developer would add the following line to the
Codewarrior linker command file for the project…
VECTOR ADDRESS 0x0000F1E0 vSci1Rx
This will inform the linker that the interrupt function "vSci1Rx()" should be placed at address 0x0000F1E0. Next,
the developer should add a file to their project "vector_table.c" that creates an array of function pointers to the
ISR routines used by the application…Eg.
extern void _Startup(void);/* _Startup located in Start08.c */
The interrupt routines themselves can be defined in separate files. The "vDummyIsr" function is used in
conjunction with "iWritetoSci1" for debugging purposes.
Bootloader Menu Commands
The bootloader accepts commands from both the local UART and OTA. All OTA commands sent must be Unicast
with only 1 byte in the payload for each command. A response will be returned to the sender. All Broadcast and
multiple byte OTA packets are dropped to help prevent general OTA traffic from being interpreted as a command
to the bootloader while in the menu.
Bypass Mode – “B”
The bootloader provides a "bypass" mode of operation that essentially connects the SCI1 serial communications
peripheral of the Freescale MCU to the EM357's serial Uart channel. This allows direct communication to the
EM357 radio for the purpose of firmware and radio configuration changes. Once in bypass mode, the XCTU utility
can change modem configuration and/or update EM357 firmware. Bypass mode automatically handles any baud
rate up to 115.2kbps. Note that this command is unavailable when module is accessed remotely.
Update Firmware – “F”
The "F" command initiates a firmware download for both wired and over-the-air configurations. Depending on the
source of the command (received via local UART/SPI, or Over the Air), the download will proceed via wired or overthe-air respectively.
Adjust Timeout for Update Firmware – “T”
The "T" command changes the timeout before sending a NAK by Base-Time*2^(T). The Base-Time for the local
UART is different than the Base-Time for Over the Air. During a firmware update, the bootloader will automatically
increase the Timeout if repeat packets are received or multiple NAKs for the same packet without success occur.
Application Version String – “A”
The "A" command provides the version of the currently loaded application. If no application is present, "Unkown"
will be returned.
The "V" command provides the version of the currently loaded bootloader.
The version will return a string in the format BLFFF-HHH-XYZ_DDD where FFF represents the Flash size in kilo bytes,
HHH is the hardware, XYZ is the version, and DDD is the preferred XMODEM packet size for updates. Double the
preferred packet size is also possible, but not guaranteed. For example "BL032-2B0-023_064" will take 64 byte CRC
XMODEM payloads and may take 128 byte CRC XMODEM payloads also. In this case, both 64 and 128 payloads are
handled, but the 64 byte payload is preferred for better Over the Air reliability.
Firmware Updates
Wired Updates
A user can update their application using the bootloader in a wired configuration with the following steps…
a. Plug XBee programmable module into a suitable serial port on a PC.
b. Open a hyperterminal (or similar dumb terminal application) session with 9600 baud, no parity, and 8
data bits with one stop bit.
c. Hit Enter to display the bootloader menu.
d. Hit the "F" key to initiate a wired firmware update.
e. A series of "C" characters Will be displayed within the hyperterminal window. At this point, select the
"transfer->send file" menu item. Select the desired flat binary output file. (The file should start at 0x8400
not 0x0000).
f. Select "Xmodem" as the protocol.
g. Click "Send" on the "Send File" dialog. The file will be downloaded to the XBee Programmable module.
Upon a successful update, the bootloader will jump to the newly loaded application.
Over-The-Air Updates
A user can update their application using the bootloader in an "over-the-air" configuration with the following
steps…(This procedure assumes that the bootloader is running and not the application. The EM357 baud rate must
be set to 9600 baud. The bootloader only operates at 9600 baud. The application must be programmed with some
way to support returning to the bootloader in order to support Over the Air (OTA) updates without local
intervention.)
a. The XBee module sending the file OTA (Host module) should be set up with a series 2 XBee module with
transparent mode firmware.
b. The XBee Programmable module receiving the update (remote module) is configured with API
firmware.
c. Open a hyperterminal session to the host module with 9600 baud, no parity, no hardwareflow control,
8 data bits and 1 stop bit.
d. Enter 3 pluses "+++" to place the EM357 in command mode.
e. Set the Host Module destination address to the target module’s 64 bit address that the host module
will update (ATDH aabbccdd, ATDL eeffgghh, ATCN, where aabbccddeeffgghh is the hexa-decimal 64 bit
address of the target module).
f. Hit Enter and the bootloader command menu will be displayed from the remote module. (Note that the
option "B" doesn't exist for OTA)
g. Hit the "F" key to cause the remote module to request the new firmware file over-the-air.
h. The host module will begin receiving "C" characters indicating that the remote module is requesting an
Xmodem CRC transfer. Using XCTU or another terminal program, Select "XMODEM" file transfer. Select
the Binary file to upload/transfer. Click Send to start the transfer. At the con-clusion of a successful
transfer, the bootloader will jump to the newly loaded application.
Output File Configuration
BKGD Programming
P&E Micro provides a background debug tool that allows flashing applications on the MC9S08QE parts through
their background debug mode port. By default, the Codewarrior tool produces an "ABS" output file for use in
programming parts through the background debug interface. The programmable XBee from the factory has the
BKGD debugging capability disabled. In order to debug, a bootloader with the debug interface enabled needs to be
loaded on the secondary processor or a stand-alone app needs to be loaded.
Bootloader Updates
The supplied bootloader requires files in a "flat binary" format which differs from the default ABS file produced.
The Codewarrior tool also produces a S19 output file. In order to successfully flash new applications, the S19 file
must be converted into the flat binary format. Utilities are available on the web that will convert S19 output to
"BIN" outputs. Often times, the "BIN" file conversion will pad the addresses from 0x0000 to the code space with
the same number. (Often 0x00 or 0xFF) These extra bytes before the APP code starts will need to be deleted from
the bin file before the file can be transferred to the bootloader.
The XBee RF Modules interface to a host device through a logic-level asynchronous serial port, or a Serial
Peripheral Interface (SPI) port. Through its serial ports, the module can communicate with any logic and
voltage compatible UART or SPI; or through a level translator to any serial device (for example: through a RS232 or USB interface board).
Two wire serial interface (TWI) is also available, but not supported by Digi. For information on the TWI see the
EM357 specification.
UART Communications
UART Data Flow
Devices that have a UART interface can connect directly to the pins of the RF module as shown in the figure
below.
UART Serial Data
Data enters the module UART through the DIN (pin 4) as an asynchronous serial signal. The signal should idle
high when no data is being transmitted. Each data byte consists of a start bit (low), 8 data bits (least significant
bit first) and a stop bit (high). The following figure illustrates the serial bit pattern of data passing through the
module.
Serial communications depend on the two UARTs (the microcontroller's and the RF module's) to be configured
with compatible settings (baud rate, parity, start bits, stop bits, data bits).
The UART baud rate, parity, and stop bits settings on the XBee module can be configured with the BD, NB, and
SB commands respectively. See the command table in chapter 10 for details.
SPI Communications
The XBee modules support SPI communications in the slave mode. Slave mode receives the clock signal and data
from the master and returns data to the master. The SPI port uses the following signals on the XBee:
• SPI_MOSI (Master Out, Slave In) – inputs serial data from the master
• SPI_MISO (Master In, Slave Out) – outputs serial data to the master
• SPI_SCLK (Serial Clock) – clocks data transfers on MOSI and MISO
• SPI_nSSEL (Slave Select) – enables serial communication with the slave
In this mode the following apply:
• Data/Clock rates up to 5MBPS are possible.
• Data is MSB first
• Frame Format mode 0 is used (see below)
Frame Format for SPI communications
SPI mode is chip to chip communication. Digi does not supply SPI communication option of Device Development
Evaluation Boards.
SPI Operation
When the slave select (SPI_nSSEL) signal is asserted by the master, SPI transmit data is driven to the output pin
SPI_MISO, and SPI data is received from the input pin SPI_MOSI. The SPI_nSSEL pin has to be asserted to enable
the transmit serializer to drive data to the output signal SPI_MISO. A falling edge on SPI_nSSEL resets the SPI slave
shift registers.
If the input buffer is empty, the SPI serializer transmits a busy token (0xFF).
The SPI slave controller must guarantee that there is time to move new transmit data from the transmit buffer into
the hardware serializer. To provide sufficient time, the SPI slave controller inserts a byte of padding at the start of
every new string of transmit data. Whenever the transmit buffer is empty and data is placed into the transmit
buffer the SPI hardware inserts a byte of padding onto the front of the transmission as if this byte was placed there
by software.
The XBee modules maintain small buffers to collect received serial and RF data, which is illustrated in the
figure below. The serial receive buffer collects incoming serial characters and holds them until they can be
processed. The serial transmit buffer collects data that is received via the RF link that will be transmitted out
the UART or SPI port.
Serial Receive Buffer
When serial data enters the RF module through the DIN Pin (pin 4), the data is stored in the serial receive buffer
until it can be processed. Under certain conditions, the module may not be able to process data in the serial
receive buffer immediately. If large amounts of serial data are sent to the module, CTS flow control may be
required to avoid overflowing the serial receive buffer.
Cases in which the serial receive buffer may become full and possibly overflow:
1. If the module is receiving a continuous stream of RF data, the data in the serial receive buffer
will not be transmitted until the module is no longer receiving RF data.
2. If the module is transmitting an RF data packet, the module may need to discover the
destination address or establish a route to the destination. After transmitting the data, the
module may need to retransmit the data if an acknowledgment is not received, or if the
transmission is a broad-cast. These issues could delay the processing of data in the serial receive
buffer.
Serial Transmit Buffer
When RF data is received, the data is moved into the serial transmit buffer and sent out the UART or SPI port. If the
serial transmit buffer becomes full enough such that all data in a received RF packet won’t fit in the serial transmit
buffer, the entire RF data packet is dropped.
Cases in which the serial transmit buffer may become full resulting in dropped RF packets
1. If the RF data rate is set higher than the interface data rate of the module, the module could
receive data faster than it can send the data to the host.
2. If the host does not allow the module to transmit data out from the serial transmit buffer
because of being held off by hardware flow control.
Serial Flow Control
The nRTS and nCTS module pins can be used to provide RTS and/or CTS flow control. CTS flow control provides an
indication to the host to stop sending serial data to the module. RTS flow control allows the host to signal the
module to not send data in the serial transmit buffer out the UART or SPI port. RTS and CTS flow control are
enabled using the D6 and D7 commands.
nCTS Flow Control
If CTS flow control is enabled (D7 command), when the serial receive buffer is 17 bytes away from being full, the
module de-asserts nCTS (sets it high) to signal to the host device to stop sending serial data. nCTS is re-asserted
after the serial receive buffer has 34 bytes of space.
nRTS Flow Control
If RTS flow control is enabled (D6 command), data in the serial transmit buffer will not be sent out the DOUT pin as
long as nRTS is de-asserted (set high). The host device should not de-assert nRTS for long periods of time to avoid
filling the serial transmit buffer. If an RF data packet is received, and the serial transmit buffer does not have
enough space for all of the data bytes, the entire RF data packet will be discarded.
Note: If the XBee is sending data out the UART or SPI port when nRTS is de-asserted (set high), the XBee could
send up to 5 characters out the UART or SPI port after RTS is de-asserted.
Serial Interface Protocols
The XBee modules support both transparent and API (Application Programming Interface) serial interfaces.
Transparent Operation
When operating in transparent mode, the modules act as a serial line replacement. All UART or SPI data received
through the respective ports is queued up for RF transmission. When RF data is received, the data is sent out
through the UART or SPI port. The module configuration parameters are configured using the AT command mode
interface.
Data is buffered in the serial receive buffer until one of the following causes the data to be packetized and
transmitted:
•No serial characters are received for the amount of time determined by the RO (Packetization Time-out)
parameter. If RO = 0, packetization begins when a character is received.
•The Command Mode Sequence (GT + CC + GT) is received. Any character buffered in the serial receive
buffer before the sequence is transmitted.
•The maximum number of characters that will fit in an RF packet is received.
API operation is an alternative to transparent operation. The frame-based API extends the level to which a host
application can interact with the networking capabilities of the module. When in API mode, all data entering and
leaving the module is contained in frames that define operations or events within the module.
Transmit Data Frames (received through the DIN pin (pin 3)) include:
•RF Transmit Data Frame
•Command Frame (equivalent to AT commands) Receive Data Frames (sent out the DOUT pin (pin 2))
include:
•RF-received data frame •Command response
•Event notifications such as reset, associate, disassociate, etc.
The API provides alternative means of configuring modules and routing data at the host application layer. A host
application can send data frames to the module that contain address and payload information instead of using
command mode to modify addresses. The module will send data frames to the application containing status
packets; as well as source, and payload information from received data packets. The API operation option
facilitates many operations such as the examples cited below:
-> Transmitting data to multiple destinations without entering Command Mode
-> Receive success/failure status of each transmitted RF packet
-> Identify the source address of each received packet
A Comparison of Transparent and API Operation
The following table compares the advantages of transparent and API modes of operation:
Transparent Operation Features
Simple Interface All received serial data is transmitted unless the module is in command mode.
Easy to support It is easier for an application to support transparent operation and command mode.
API Operation Features
Transmitting RF data to multiple remotes only requires changing the address in the API frame.
Easy to manage data
transmissions to
multiple destinations
Received data frames
indicate the sender's
address
This Process is much faster than transparent operation where the application must enter AT
command mode, change the address, exit command mode, and then transmit data.
Each API transmission can return a transmit status frame indicating the success or reason for
failure
All received RF data API frames indicate the source address.
API transmit and receive frames can expose ZigBee addressing fields including source and
destination endpoints, cluster ID and profile ID. This makes it easy to support ZDO commands
and public profile traffic.
Page 38
XBee®/XBee‐PRO® SMT ZB RF Modules
Advanced Networking
diagnostics
Remote Configuration
API frames can provide indication of IO samples from remote devices, and node identification
messages.
Set/read configuration commands can be sent to remote devices to configure them as needed
using the API.
As a general rule of thumb, API firmware is recommended when a device:
•sends RF data to multiple destinations
•sends remote configuration commands to manage devices in the network
•receives IO samples from remote devices
•receives RF data packets from multiple devices, and the application needs to know which device sent
which packet
•must support multiple ZigBee endpoints, cluster IDs, and/or profile IDs
•uses the ZigBee Device Profile services.
If the above conditions do not apply (e.g. a sensor node, router, or a simple application), then AT firmware might
be suitable. It is acceptable to use a mixture of devices running API and AT firmware in a network.
When not receiving or transmitting data, the RF module is in Idle Mode. The module shifts into the other modes of
operation under the following conditions:
•Transmit Mode (Serial data in the serial receive buffer is ready to be packetized)
•Receive Mode (Valid RF data is received through the antenna)
•Sleep Mode (End Devices only)
•Command Mode (Command Mode Sequence is issued)
Transmit Mode
When serial data is received and is ready for packetization, the RF module will exit Idle Mode and attempt to
transmit the data. The destination address determines which node(s) will receive the data.
Prior to transmitting the data, the module ensures that a 16-bit network address and route to the destination node
have been established.
If the destination 16-bit network address is not known, network address discovery will take place. If a route is not
known, route discovery will take place for the purpose of establishing a route to the destination node. If a module
with a matching network address is not discovered, the packet is discarded. The data will be transmitted once a
route is established. If route discovery fails to establish a route, the packet will be discarded.
When data is transmitted from one node to another, a network-level acknowledgement is transmitted back across
the established route to the source node. This acknowledgement packet indicates to the source node that the data
packet was received by the destination node. If a network acknowledgement is not received, the source node will
re-transmit the data.
It is possible in rare circumstances for the destination to receive a data packet, but for the source to not receive
the network acknowledgment. In this case, the source will retransmit the data, which could cause the destination
to receive the same data packet multiple times. The XBee modules do not filter out duplicate packets. The
application should include provisions to address this potential issue
See Data Transmission and Routing in chapter 4 for more information.
Receive Mode
If a valid RF packet is received, the data is transferred to the serial transmit buffer.
Command Mode
To modify or read RF Module parameters, the module must first enter into Command Mode - a state in which
incoming serial characters are interpreted as commands. Refer to the API Mode section in chapter 9 for an
alternate means of configuring modules.
AT Command Mode
To Enter AT Command Mode:
Send the 3-character command sequence “+++” and observe guard times before and
after the command characters. [Refer to the “Default AT Command Mode Sequence”
below.]
Default AT Command Mode Sequence (for transition to Command Mode):
•No characters sent for one second [GT (Guard Times) parameter = 0x3E8]
•Input three plus characters (“+++”) within one second [CC (Command Sequence
Character) parameter = 0x2B.]
•No characters sent for one second [GT (Guard Times) parameter = 0x3E8]
Once the AT command mode sequence has been issued, the module sends an "OK\r" out the
UART or SPI pad. The "OK\r" characters can be delayed if the module has not finished
transmitting received serial data.
When command mode has been entered, the command mode timer is started (CT command),
and the module is able to receive AT commands on the UART or SPI port.
All of the parameter values in the sequence can be modified to reflect user preferences.
NOTE: Failure to enter AT Command Mode is most commonly due to baud rate
mismatch. By default, the BD (Baud Rate) parameter = 3 (9600 bps).
To Send AT Commands:
Send AT commands and parameters using the syntax shown below.
To read a parameter value stored in the RF module’s register, omit the parameter field.
The preceding example would change the RF module Destination Address (Low) to “0x1F”. To
store the new value to non-volatile (long term) memory, subsequently send the WR (Write)
command.
For modified parameter values to persist in the module’s registry after a reset, changes must be
saved to non-volatile memory using the WR (Write) Command. Otherwise, parameters are
restored to previously saved values after the module is reset.
Command Response
When a command is sent to the module, the module will parse and execute the command. Upon
successful execution of a command, the module returns an “OK” message. If execution of a
command results in an error, the module returns an “ERROR” message.
Applying Command Changes
Any changes made to the configuration command registers through AT commands will not take
effect until the changes are applied. For example, sending the BD command to change the baud
rate will not change the actual baud rate until changes are applied. Changes can be applied in
one of the following ways:
•The AC (Apply Changes) command is issued.
•AT command mode is exited.
To Exit AT Command Mode:
1. Send the ATCN (Exit Command Mode) command (followed by a carriage return).
[OR]
2. If no valid AT Commands are received within the time specified by CT (Command Mode Timeout)
Command, the RF module automatically returns to Idle Mode.
For an example of programming the RF module using AT Commands and descriptions of each
configurable parameter, please see the Command Reference Table chapter.
Sleep modes allow the RF module to enter states of low power consumption when not in use. The XBee RF
modules support both pin sleep (sleep mode entered on pin transition) and cyclic sleep (module sleeps for a fixed
time). XBee sleep modes are discussed in detail in chapter 6.
ZigBee is an open global standard built on the IEEE 802.15.4 MAC/PHY. ZigBee defines a network layer above the
802.15.4 layers to support advanced mesh routing capabilities. The ZigBee specification is developed by a growing
consortium of companies that make up the ZigBee Alliance. The Alliance is made up of over 300 members,
including semiconductor, module, stack, and software developers.
ZigBee Stack Layers
The ZigBee stack consists of several layers including the PHY, MAC, Network, Application Support Sublayer (APS),
and ZigBee Device Objects (ZDO) layers. Technically, an Application Framework (AF) layer also exists, but will be
grouped with the APS layer in remaining discussions. The ZigBee layers are shown in the figure below.
A description of each layer appears in the following table:
ZigBee Layer Description
Defines the physical operation of the ZigBee device including
receive sensitivity, channel rejection, output power, number
PHY
of channels, chip modulation, and transmission rate
specifications. Most ZigBee applications operate on the 2.4
GHz ISM band at a 250kbps data rate. See the IEEE 802.15.4
specification for details.
Manages RF data transactions between neighboring devices
MAC
Network
APS (AF)
ZDO
(point to point). The MAC includes services such as
transmission retry and acknowledgement management, and
collision avoidance techniques (CSMA-CA)
Adds routing capabilities that allow RF data packets to
traverse multiple devices (multiple "hops") to route data from
source to destination (peer to peer).
Application layer that defines various addressing objects
including profiles, clusters, and endpoints.
Application layer that provides device and service discovery
features and advanced network management capabilities.
Networking concepts
Device Types
ZigBee defines three different device types: coordinator, router, and end device.
Node Types / Sample of a Basic ZigBee Network Topology
A coordinator has the following characteristics: it
•Selects a channel and PAN ID (both 64-bit and 16-bit) to start the network
•Can allow routers and end devices to join the network
•Can assist in routing data
•Cannot sleep--should be mains powered
•Can buffer RF data packets for sleeping end device children.
A router has the following characteristics: it
•Must join a ZigBee PAN before it can transmit, receive, or route data
•After joining, can allow routers and end devices to join the network
•After joining, can assist in routing data
•Cannot sleep--should be mains powered.
•Can buffer RF data packets for sleeping end device children.
An end device has the following characteristics: it
•Must join a ZigBee PAN before it can transmit or receive data
•Cannot allow devices to join the network
•Must always transmit and receive RF data through its parent. Cannot route data.
•Can enter low power modes to conserve power and can be battery-powered.
An example of such a network is shown below:
In ZigBee networks, the coordinator must select a PAN ID (64-bit and 16-bit) and channel to start a network. After
that, it behaves essentially like a router. The coordinator and routers can allow other devices to join the network
and can route data.
After an end device joins a router or coordinator, it must be able to transmit or receive RF data through that router
or coordinator. The router or coordinator that allowed an end device to join becomes the "parent" of the end
device. Since the end device can sleep, the parent must be able to buffer or retain incoming data packets destined
for the end device until the end device is able to wake and receive the data.
PAN ID
ZigBee networks are called personal area networks or PANs. Each network is defined with a unique PAN identifier
(PAN ID). This identifier is common among all devices of the same network. ZigBee devices are either
preconfigured with a PAN ID to join, or they can discovery nearby networks and select a PAN ID to join.
ZigBee supports both a 64-bit and a 16-bit PAN ID. Both PAN IDs are used to uniquely identify a network. Devices
on the same ZigBee network must share the same 64-bit and 16-bit PAN IDs. If multiple ZigBee networks are
operating within range of each other, each should have unique PAN IDs.
The 16-bit PAN ID is used as a MAC layer addressing field in all RF data transmissions between devices in a
network. However, due to the limited addressing space of the 16-bit PAN ID (65,535 possibilities); there is a
possibility that multiple ZigBee networks (within range of each other) could use the same 16-bit PAN ID. To resolve
potential 16-bit PAN ID conflicts, the ZigBee Alliance created a 64-bit PAN ID.
The 64-bit PAN ID (also called the extended PAN ID), is intended to be a unique, non-duplicated value. When a
coordinator starts a network, it can either start a network on a preconfigured 64-bit PAN ID, or it can select a
random 64-bit PAN ID. The 64-bit PAN ID is used during joining; if a device has a preconfigured 64-bit PAN ID, it will
only join a network with the same 64-bit PAN ID. Otherwise, a device could join any detected PAN and inherit the
PAN ID from the network when it joins. The 64-bit PAN ID is included in all ZigBee beacons and is used in 16-bit
PAN ID conflict resolution.
Routers and end devices are typically configured to join a network with any 16-bit PAN ID as long as the 64-bit PAN
ID is valid. Coordinators typically select a random 16-bit PAN ID for their network.
Since the 16-bit PAN ID only allows up to 65,535 unique values, and since the 16-bit PAN ID is randomly selected,
provisions exist in ZigBee to detect if two networks (with different 64-bit PAN IDs) are operating on the same 16bit PAN ID. If such a conflict is detected, the ZigBee stack can perform PAN ID conflict resolution to change the 16bit PAN ID of the network in order to resolve the conflict. See the ZigBee specification for details.
To summarize, ZigBee routers and end devices should be configured with the 64-bit PAN ID of the network they
want to join. They typically acquire the 16-bit PAN ID when they join a network.
Operating Channel
ZigBee utilizes direct-sequence spread spectrum modulation and operates on a fixed channel. The 802.15.4 PHY
defines 16 operating channels (channels 11 to 26) in the 2.4 GHz frequency band. XBee modules support all 16
channels.
ZigBee Application Layers: In Depth
This section provides a more in-depth look at the ZigBee application stack layers (APS, ZDO) including a discussion
on ZigBee endpoints, clusters, and profiles. Much of the material in this section can introduce unnecessary details
of the ZigBee stack that are not required in many cases.
Skip this section if
•The XBee does not need to interoperate or talk to non-Digi ZigBee devices
•The XBee simply needs to send data between devices.
Read this section if
•The XBee may talk to non-Digi ZigBee devices
•The XBee requires network management and discovery capabilities of the ZDO layer
•The XBee needs to operate in a public application profile (smart energy, home automation, etc.)
Application Support Sublayer (APS)
The APS layer in ZigBee adds support for application profiles, cluster IDs, and endpoints.
Application Profiles
Application profiles specify various device descriptions including required functionality for various devices. The
collection of device descriptions forms an application profile. Application profiles can be defined as "Public" or
"Private" profiles. Private profiles are defined by a manufacturer whereas public profiles are defined, developed,
and maintained by the ZigBee Alliance. Each application profile has a unique profile identifier assigned by the
ZigBee Alliance.
Examples of public profiles include:
•Home Automation
•Smart Energy
•Commercial Building Automation
The Smart Energy profile, for example, defines various device types including an energy service portal, load
controller, thermostat, in-home display, etc. The Smart Energy profile defines required functionality for each
device type. For example, a load controller must respond to a defined command to turn a load on or off. By
defining standard communication protocols and device functionality, public profiles allow interoperable ZigBee
solutions to be developed by independent manufacturers.
Digi XBee ZB firmware operates on a private profile called the Digi Drop-In Networking profile. However, the API
firmware in the module can be used in many cases to talk to devices in public profiles or non-Digi private profiles.
See the API Operations chapter for details.
Clusters
A cluster is an application message type defined within a profile. Clusters are used to specify a unique function,
service, or action. For example, the following are some clusters defined in the home automation profile:
•On/Off - Used to switch devices on or off (lights, thermostats, etc.)
•Level Control - Used to control devices that can be set to a level between on and off
•Color Control - Controls the color of color capable devices.
Each cluster has an associated 2-byte cluster identifier (cluster ID). The cluster ID is included in all application
transmissions. Clusters often have associated request and response messages. For example, a smart energy
gateway (service portal) might send a load control event to a load controller in order to schedule turning on or off
an appliance. Upon executing the event, the load controller would send a load control report message back to the
gateway.
Devices that operate in an application profile (private or public) must respond correctly to all required clusters. For
example, a light switch that will operate in the home automation public profile must correctly implement the
On/Off and other required clusters in order to interoperate with other home automation devices. The ZigBee
Alliance has defined a ZigBee Cluster Library (ZCL) that contains definitions or various general use clusters that
could be implemented in any profile. XBee modules implement various clusters in the Digi private profile. In
addition, the API can be used to send or receive messages on any cluster ID (and profile ID or endpoint). See the
Explicit Addressing ZigBee Command API frame in chapter 3 for details.
Endpoints
The APS layer includes supports for endpoints. An endpoint can be thought of as a running application, similar to a
TCP/IP port. A single device can support one or more endpoints. Each application endpoint is identified by a 1-byte
value, ranging from 1 to 240. Each defined endpoint on a device is tied to an application profile. A device could, for
example, implement one endpoint that supports a Smart Energy load controller, and another endpoint that
supports other functionality on a private profile.
Profile ID 0x0000 is reserved for the ZigBee Device Profile. This profile is implemented on all ZigBee devices. Device
Profile defines many device and service discovery features and network management capabilities. Endpoint 0 is a
reserved endpoint that supports the ZigBee Device Profile. This endpoint is called the ZigBee Device Objects (ZDO)
endpoint.
ZigBee Device Objects (ZDO)
The ZDO (endpoint 0) supports the discovery and management capabilities of the ZigBee Device Profile. A
complete listing of all ZDP services is included in the ZigBee specification. Each service has an associated cluster ID.
The XBee ZB firmware allows applications to easily send ZDO messages to devices in the network using the API. See
the ZDO Transmissions section in chapter 4 for details.
Coordinator Operation
Forming a Network
The coordinator is responsible for selecting the channel, PAN ID (16-bit and 64-bit), security policy, and stack
profile for a network. Since a coordinator is the only device type that can start a network, each ZigBee network
must have one coordinator. After the coordinator has started a network, it can allow new devices to join the
network. It can also route data packets and communicate with other devices on the network.
To ensure the coordinator starts on a good channel and unused PAN ID, the coordinator performs a series of scans
to discover any RF activity on different channels (energy scan) and to discover any nearby operating PANs (PAN
scan). The process for selecting the channel and PAN ID are described in the following sections.
Channel Selection
When starting a network, the coordinator must select a "good" channel for the network to operate on. To do this,
it performs an energy scan on multiple channels (frequencies) to detect energy levels on each channel. Channels
with excessive energy levels are removed from its list of potential channels to start on.
PAN ID Selection
After completing the energy scan, the coordinator scans its list of potential channels (remaining channels after the
energy scan) to obtain a list of neighboring PANs. To do this, the coordinator sends a beacon request (broadcast)
transmission on each potential channel. All nearby coordinators and routers (that have already joined a ZigBee
network) will respond to the beacon request by sending a beacon back to the coordinator. The beacon contains
information about the PAN the device is on, including the PAN identifiers (16-bit and 64-bit). This scan (collecting
beacons on the potential channels) is typically called an active scan or PAN scan.
After the coordinator completes the channel and PAN scan, it selects a random channel and unused 16-bit PAN ID
to start on.
Security Policy
The security policy determines which devices are allowed to join the network, and which device(s) can
authenticate joining devices. See chapter 5 for a detailed discussion of various security policies.
Persistent Data
Once a coordinator has started a network, it retains the following information through power cycle or reset events:
•Child table (end device children that are joined to the coordinator).
The coordinator will retain this information indefinitely until it leaves the network. When the coordinator leaves a
network and starts a new network, the previous PAN ID, operating channel, and child table data are lost.
XBee ZB Coordinator Startup
The following commands control the coordinator network formation process.
Network formation commands used by the coordinator to form a network.
Command Description
ID
SC
Used to determine the 64-bit PAN ID. If set to 0 (default), a random 64-bit PAN ID
will be selected.
Determines the scan channels bitmask (up to 16 channels) used by the coordinator
when forming a network. The coordinator will perform an energy scan on all
enabled SC channels. It will then perform a PAN ID scan and then form the
network on one of the SC channels
SD
Set the scan duration period. This value determined how long the coordinator
performs an energy scan or PAN ID scan on a given channel.
ZS Set the ZigBee Stack profile for the network.
EE Enable or disable security in the network.
NK
Set the network security key for the network. If set to 0 (default), a random
network security key will be used.
KY
Set the trust center link key for the network. If set to 0 (default), a random link key
will be used.
EO
Once the coordinator starts a network, the network configuration settings and child table data persist through
power cycles as mentioned in the "Persistent Data" section.
When the coordinator has successfully started a network, it
•Allows other devices to join the network for a time (see NJ command)
•Sets AI=0 •Starts blinking the Associate LED
•Sends an API modem status frame ("coordinator started") out the UART (API firmware only).
These behaviors are configurable using the following commands:
Set the security policy for the network.
Command Description
NJ Sets the permit-join time on the coordinator, measured in seconds.
D5 Enables the Associate LED functionality
LT Sets the Associate LED blink time when joined. Default is 1 blink per second.
If any of the command values in the network formation commands table changes, the coordinator will leave its
current network and start a new network, possibly on a different channel. Note that command changes must be
applied (AC or CN command) before taking effect.
Permit Joining
The permit joining attribute on the coordinator is configurable with the NJ command. NJ can be configured to
always allow joining, or to allow joining for a short time.
Joining Always Enabled
If NJ=0xFF (default), joining is permanently enabled. This mode should be used carefully. Once a network has been
deployed, the application should strongly consider disabling joining to prevent unwanted joins from occurring.
Joining Temporarily Enabled
If NJ < 0xFF, joining will be enabled only for a number of seconds, based on the NJ parameter. The timer is started
once the XBee joins a network. Joining will not be re-enabled if the module is power cycled or reset. The following
mechanisms can restart the permit-joining timer:
•Changing NJ to a different value (and applying changes with the AC or CN commands)
•Pressing the commissioning button twice (enables joining for 1 minute)
•Issuing the CB command with a parameter of 2 (software emulation of a 2 button press - enables joining
for 1 minute).
Resetting the Coordinator
When the coordinator is reset or power cycled, it checks its PAN ID, operating channel and stack profile against the
network configuration settings (ID, CH, and ZS). It also verifies the saved security policy against the security
configuration settings (EE, NK, KY). If the coordinator's PAN ID, operating channel, stack profile, or security policy is
not valid based on its network and security configuration settings, then the coordinator will leave the network and
attempt to form a new network based on its network formation command values.
To prevent the coordinator from leaving an existing network, the WR command should be issued after all network
formation commands have been configured in order to retain these settings through power cycle or reset events.
Leaving a Network
There are a couple of mechanisms that will cause the coordinator to leave its current PAN and start a new network
based on its network formation parameter values. These include the following:
•Change the ID command such that the current 64-bit PAN ID is invalid.
•Change the SC command such that the current channel (CH) is not included in the channel mask.
•Change the ZS or any of the security command values (excluding NK).
•Issue the NR0 command to cause the coordinator to leave.
•Issue the NR1 command to send a broadcast transmission, causing all devices in the network to leave
and migrate to a different channel.
•Press the commissioning button 4 times or issue the CB command with a parameter of 4.
Note that changes to ID, SC, ZS, and security command values only take effect when changes are applied (AC or CN
commands).
In rare occasions, it may become necessary to replace an existing coordinator in a network with a new physical
device. If security is not enabled in the network, a replacement XBee coordinator can be configured with the PAN
ID (16-bit and 64-bit), channel, and stack profile settings of a running network in order to replace an existing
coordinator.
NOTE: Having two coordinators on the same channel, stack profile, and PAN ID (16-bit and 64-bit) can cause
problems in the network and should be avoided. When replacing a coordinator, the old coordinator should be
turned off before starting the new coordinator.
To replace a coordinator, the following commands should be read from a device on the network:
AT Command Description
OP Read the operating 64-bit PAD ID.
OI Read the operating 16-bit PAN ID.
CH Read the operating channel.
ZS Read the stack profile.
Each of the commands listed above can be read from any device on the network. (These parameters will be the
same on all devices in the network.) After reading these commands from a device on the network, these
parameter values should be programmed into the new coordinator using the following commands.
AT Command Description
ID
II
Set the 64-bit PAN ID to match
the read OP value.
Set the initial 16-bit PAN ID to
match the read OI value.
Set the scan channels bitmask to
enable the read operating
channel (CH command). For
SC
example, if the operating channel
is 0x0B, set SC to 0x0001. If the
operating channel is 0x17, set SC
to 0x1000.
ZS
Set the stack profile to match the
read ZS value.
Note: II is the initial 16-bit PAN ID. Under certain conditions, the ZigBee stack can change the 16-bit PAN ID of the
network. For this reason, the II command cannot be saved using the WR command. Once II is set, the coordinator
leaves the network and starts on the 16-bit PAN ID specified by II.
Example: Starting a Coordinator
1. Set SC and ID to the desired scan channels and PAN ID values. (The defaults should suffice.)
2. If SC or ID is changed from the default, issue the WR command to save the changes.
3. If SC or ID is changed from the default, apply changes (make SC and ID changes take effect) either by
sending the AC command or by exiting AT command mode.
4. The Associate LED will start blinking once the coordinator has selected a channel and PAN ID.
5. The API Modem Status frame ("Coordinator Started") is sent out the UART (API firmware only).
6. Reading the AI command (association status) will return a value of 0, indicating a successful startup.
7. Reading the MY command (16-bit address) will return a value of 0, the ZigBee-defined 16-bit address of
the coordinator.
After startup, the coordinator will allow joining based on its NJ value.
Example: Replacing a Coordinator (security disabled)
1. Read the OP, OI, CH, and ZS commands on the running coordinator.
2. Set the ID, SC, and ZS parameters on the new coordinator, followed by WR command to save these parameter
values.
3. Turn off the running coordinator.
4. Set the II parameter on the new coordinator to match the read OI value on the old coordinator.
5. Wait for the new coordinator to start (AI=0).
Router Operation
Routers must discover and join a valid ZigBee network before they can participate in a ZigBee network. After a
router has joined a network, it can allow new devices to join the network. It can also route data packets and
communicate with other devices on the network.
Discovering ZigBee Networks
To discover nearby ZigBee networks, the router performs a PAN (or active) scan, just like the coordinator does
when it starts a network. During the PAN scan, the router sends a beacon request (broadcast) transmission on the
first channel in its scan channels list. All nearby coordinators and routers operating on that channel (that are
already part of a ZigBee network) respond to the beacon request by sending a beacon back to the router. The
beacon contains information about the PAN the nearby device is on, including the PAN identifier (PAN ID), and
whether or not joining is allowed. The router evaluates each beacon received on the channel to determine if a
valid PAN is found. A router considers a PAN to be valid if the PAN:
•Has a valid 64-bit PAN ID (PAN ID matches ID if ID > 0)
•Has the correct stack profile (ZS command)
•Is allowing joining.
If a valid PAN is not found, the router performs the PAN scan on the next channel in its scan channels list and
continues scanning until a valid network is found, or until all channels have been scanned. If all channels have been
scanned and a valid PAN was not discovered, all channels will be scanned again.
The ZigBee Alliance requires that certified solutions not send beacon request messages too frequently. To meet
certification requirements, the XBee firmware attempts 9 scans per minute for the first 5 minutes, and 3 scans per
minute thereafter. If a valid PAN is within range of a joining router, it should typically be discovered within a few
seconds.
Joining a Network
Once the router discovers a valid network, it sends an association request to the device that sent a valid beacon
requesting a join on the ZigBee network. The device allowing the join then sends an association response frame
that either allows or denies the join.
When a router joins a network, it receives a 16-bit address from the device that allowed the join. The 16-bit
address is randomly selected by the device that allowed the join.
Authentication
In a network where security is enabled, the router must then go through an authentication process. See the
Security chapter for a discussion on security and authentication.
After the router is joined (and authenticated, in a secure network), it can allow new devices to join the network.
Persistent Data
Once a router has joined a network, it retains the following information through power cycle or reset events:
•PAN ID
•Operating channel
•Security policy and frame counter values
•Child table (end device children that are joined to the coordinator).
The router will retain this information indefinitely until it leaves the network. When the router leaves a network,
the previous PAN ID, operating channel, and child table data are lost.
XBee ZB Router Joining
When the router is powered on, if it is not already joined to a valid ZigBee network, it immediately attempts to find
and join a valid ZigBee network.
Note: The DJ command can be set to 1 to disable joining. The DJ parameter cannot be written with WR, so a power
cycle always clears the DJ setting.
The following commands control the router joining process.
Command Description
ID
Sets the 64-bit PAN ID to join. Setting ID=0 allows the router to join any 64-bit PAN
ID.
Set the scan channels bitmask that determines which channels a router will scan to
SC
find a valid network. SC on the router should be set to match SC on the
coordinator. For example, setting SC to 0x281 enables scanning on channels 0x0b,
0x12, and 0x14, in that order.
SD
Set the scan duration, or time that the router will listen for beacons on each
channel.
ZS Sets the stack profile on the device.
EE
KY
Enable or disable security in the network. This must be set to match the EE value
(security policy) of the coordinator.
Set the trust center link key. If set to 0 (default), the link key is expected to be
obtained (unencrypted) during joining.
Once the router joins a network, the network configuration settings and child table data persist through power
cycles as mentioned in the "Persistent Data" section previously. If joining fails, the status of the last join attempt
can be read in the AI command register.
If any of the above command values change, when command register changes are applied (AC or CN commands),
the router will leave its current network and attempt to discover and join a new valid network. When a ZB router
has successfully joined a network, it:
•Allows other devices to join the network for a time
•Sets AI=0
•Starts blinking the Associate LED
•Sends an API modem status frame ("associated") out the UART (API firmware only).
These behaviors are configurable using the following commands:
Command Description
Sets the permit-join time on the router, or the time that it will allow new devices to
NJ
join the network, measured in seconds. If NJ=0xFF, permit joining will always be
enabled.
D5 Enables the Associate LED functionality.
LT
Sets the Associate LED blink time when joined. Default is 2 blinks per second
(router).
Permit Joining
The permit joining attribute on the router is configurable with the NJ command. NJ can be configured to always
allow joining, or to allow joining for a short time.
Joining Always Enabled
If NJ=0xFF (default), joining is permanently enabled. This mode should be used carefully. Once a network has been
deployed, the application should strongly consider disabling joining to prevent unwanted joins from occurring.
Joining Temporarily Enabled
If NJ < 0xFF, joining will be enabled only for a number of seconds, based on the NJ parameter. The timer is started
once the XBee joins a network. Joining will not be re-enabled if the module is power cycled or reset. The following
mechanisms can restart the permit-joining timer:
•Changing NJ to a different value (and applying changes with the AC or CN commands)
•Pressing the commissioning button twice (enables joining for 1 minute)
•Issuing the CB command with a parameter of 2 (software emulation of a 2 button press - enables joining
for 1 minute)
•Causing the router to leave and rejoin the network.
Router Network Connectivity
Once a router joins a ZigBee network, it remains connected to the network on the same channel and PAN ID as
long as it is not forced to leave. (See Leaving a Network section for details.) If the scan channels (SC), PAN ID (ID)
and security settings (EE, KY) do not change after a power cycle; the router will remain connected to the network
after a power cycle.
If a router may physically move out of range of the network it initially joined, the application should include
provisions to detect if the router can still communicate with the original network. If communication with the
original network is lost, the application may choose to force the router to leave the network (see Leaving a
Network section for details). The XBee firmware includes two provisions to automatically detect the presence of a
network, and leave if the check fails.
Power-On Join Verification
The JV command (join verification) enables the power-on join verification check. If enabled, the XBee will attempt
to discover the 64-bit address of the coordinator when it first joins a network. Once it has joined, it will also
attempt to discover the 64-bit address of the coordinator after a power cycle event. If 3 discovery attempts fail,
the router will leave the network and try to join a new network. Power-on join verification is disabled by default
(JV defaults to 0).
Network Watchdog
The NW command (network watchdog timeout) can be used for a powered router to periodically check for the
presence of a coordinator to verify network connectivity. The NW command specifies a timeout in minutes where
the router must receive communication from the coordinator or data collector. The following events restart the
network watchdog timer:
•RF data received from the coordinator
•RF data sent to the coordinator and an acknowledgment was received
•Many-to-one route request was received (from any device)
•Changing the value of NW.
If the watchdog timer expires (no valid data received for NW time), the router will attempt to discover the 64-bit
address of the coordinator. If the address cannot be discovered, the router records one watchdog timeout. Once
three consecutive network watchdog timeouts have expired (3 * NW) and the coordinator has not responded to
the address discovery attempts, the router will leave the network and attempt to join a new network. Anytime a
router receives valid data from the coordinator or data collector, it will clear the watchdog timeouts counter and
restart the watchdog timer. The watchdog timer (NW command) is settable to several days. The network watchdog
feature is disabled by default (NW defaults to 0).
There are a couple of mechanisms that will cause the router to leave its current PAN and attempt to discover and
join a new network based on its network joining parameter values.
These include the following:
•Change the ID command such that the current 64-bit PAN ID is invalid.
•Change the SC command such that the current channel (CH) is not included in the channel mask.
•Change the ZS or any of the security command values.
•Issue the NR0 command to cause the router to leave.
•Issue the NR1 command to send a broadcast transmission, causing all devices in the network to leave
and migrate to a different channel.
•Press the commissioning button 4 times or issue the CB command with a parameter of 4.
Note that changes to ID, SC, ZS, and security command values only take effect when changes are applied (AC or CN
commands).
When the router is reset or power cycled, it checks its PAN ID, operating channel and stack profile against the
network configuration settings (ID, SC, and ZS). It also verifies the saved security policy is valid based on the
security configuration commands (EE, KY). If the router's PAN ID, operating channel, stack profile, or security policy
is invalid, the router will leave the network and attempt to join a new network based on its network joining
command values.
To prevent the router from leaving an existing network, the WR command should be issued after all network
joining commands have been configured in order to retain these settings through power cycle or reset events.
Example: Joining a Network
After starting a coordinator (that is allowing joins), the following steps will cause a router to join the network:
1. Set ID to the desired 64-bit PAN ID, or to 0 to join any PAN.
2. Set SC to the list of channels to scan to find a valid network.
3. If SC or ID is changed from the default, apply changes (make SC and ID changes take effect) by issuing
the AC or CN command.
4. The Associate LED will start blinking once the router has joined a PAN.
5. If the Associate LED is not blinking, the AI command can be read to determine the cause of join failure.
6. Once the router has joined, the OP and CH commands will indicate the operating 64-bit PAN ID and
channel the router joined.
7. The MY command will reflect the 16-bit address the router received when it joined.
8. The API Modem Status frame ("Associated") is sent out the UART (API firmware only).
9. The joined router will allow other devices to join for a time based on its NJ setting.
End Device Operation
Similar to routers, end devices must also discover and join a valid ZigBee network before they can participate in a
network. After an end device has joined a network, it can communicate with other devices on the network. Since
end devices are intended to be battery powered and therefore support low power (sleep) modes, end devices
cannot allow other devices to join, nor can they route data packets.
Discovering ZigBee Networks
End devices go through the same process as routers to discover networks by issuing a PAN scan. After sending the
broadcast beacon request transmission, the end device listens for a short time in order to receive beacons sent by
nearby routers and coordinators on the same channel. The end device evaluates each beacon received on the
channel to determine if a valid PAN is found. An end device considers a PAN to be valid if the PAN:
•Has a valid 64-bit PAN ID (PAN ID matches ID if ID > 0)
•Has the correct stack profile (ZS command)
•Is allowing joining
•Has capacity for additional end devices (see End Device Capacity section below).
If a valid PAN is not found, the end device performs the PAN scan on the next channel in its scan channels list and
continues this process until a valid network is found, or until all channels have been scanned. If all channels have
been scanned and a valid PAN was not discovered, the end device may enter a low power sleep state and scan
again later.
If scanning all SC channels fails to discover a valid PAN, XBee ZB modules will attempt to enter a low power state
and will retry scanning all SC channels after the module wakes from sleeping. If the module cannot enter a low
power state, it will retry scanning all channels, similar to the router. To meet ZigBee Alliance requirements, the end
device will attempt up to 9 scans per minute for the first 5 minutes, and 3 scans per minute thereafter.
Note: The XBee ZB end device will not enter sleep until it has completed scanning all SC channels for a valid
network.
Once the end device discovers a valid network, it joins the network, similar to a router, by sending an association
request (to the device that sent a valid beacon) to request a join on the ZigBee network. The device allowing the
join then sends an association response frame that either allows or denies the join. When an end device joins a
network, it receives a 16-bit address from the device that allowed the join. The 16- bit address is randomly
selected by the device that allowed the join.
Parent Child Relationship
Since an end device may enter low power sleep modes and not be immediately responsive, the end device relies
on the device that allowed the join to receive and buffer incoming messages in its behalf until it is able to wake
and receive those messages. The device that allowed an end device to join becomes the parent of the end device,
and the end device becomes a child of the device that allowed the join.
End Device Capacity
Routers and coordinators maintain a table of all child devices that have joined called the child table. This table is a
finite size and determines how many end devices can join. If a router or coordinator has at least one unused entry
in its child table, the device is said to have end device capacity. In other words, it can allow one or more additional
end devices to join. ZigBee networks should have sufficient routers to ensure adequate end device capacity.
In the XBee ZB firmware, a coordinator can support 20 end devices, and a router can support 24 end devices.
In ZB firmware, the NC command (number of remaining end device children) can be used to determine how many
additional end devices can join a router or coordinator. If NC returns 0, then the router or coordinator device has
no more end device capacity. (Its child table is full.)
Also of note, since routers cannot sleep, there is no equivalent need for routers or coordinators to track joined
routers. Therefore, there is no limit to the number of routers that can join a given router or coordinator device.
(There is no "router capacity" metric.)
Authentication
In a network where security is enabled, the end device must then go through an authentication process. See
chapter 5 for a discussion on security and authentication.
Persistent Data
The end device can retain its PAN ID, operating channel, and security policy information through a power cycle.
However, since end devices rely heavily on a parent, the end device does an orphan scan to try and contact its
parent. If the end device does not receive an orphan scan response (called a coordinator realignment command), it
will leave the network and try to discover and join a new network. When the end device leaves a network, the
previous PAN ID and operating channel settings are lost.
Orphan Scans
When an end device comes up from a power cycle, it performs an orphan scan to verify it still has a valid parent.
The orphan scan is sent as a broadcast transmission and contains the 64-bit address of the end device. Nearby
routers and coordinator devices that receive the broadcast check their child tables for an entry that contains the
end device's 64-bit address. If an entry is found with a matching 64-bit address, the device sends a coordinator
realignment command to the end device that includes the end device's 16-bit address, 16-bit PAN ID, operating
channel, and the parent's 64-bit and 16-bit addresses.
If the orphaned end device receives a coordinator realignment command, it is considered joined to the network.
Otherwise, it will attempt to discover and join a valid network.
When an end device is powered on, if it is not joined to a valid ZigBee network, or if the orphan scan fails to find a
parent, it immediately attempts to find and join a valid ZigBee network.
Note: The DJ command can be set to 1 to disable joining. The DJ parameter cannot be written with WR, so a power
cycle always clears the DJ setting.
Similar to a router, the following commands control the end device joining process.
Network joining commands used by an end device to join a network.
Command Description
ID
Sets the 64-bit PAN ID to join. Setting ID=0 allows the router to join any 64-bit PAN
ID.
Set the scan channels bitmask that determines which channels a router will scan to
SC
find a valid network. SC on the end device should be set to match SC on the
coordinator and router. For example, setting SC to 0x281 enables scanning on
channels 0x0b, 0x12, and 0x14, in that order.
SD
Set the scan duration, or time that the end device will listen for beacons on each
channel.
ZS Sets the stack profile on the device.
EE
KY
Enable or disable security in the network. This must be set to match the EE value
(security policy) of the coordinator.
Set the trust center link key. If set to 0 (default), the link key is expected to be
obtained (unencrypted) during joining.
Once the end device joins a network, the network configuration settings can persist through power cycles as
mentioned in the "Persistent Data" section previously. If joining fails, the status of the last join attempt can be read
in the AI command register.
If any of these command values changes, when command register changes are applied, the end device will leave
its current network and attempt to discover and join a new valid network.
When a ZB end device has successfully started a network, it
•Sets AI=0
•Starts blinking the Associate LED
•Sends an API modem status frame ("associated") out the UART (API firmware only)
•Attempts to enter low power modes.
These behaviors are configurable using the following commands:
Command Description
D5 Enables the Associate LED functionality.
LT
SM, SP, ST,
SN, SO
Sets the Associate LED blink time when joined. Default is 2 blinks per second
(router).
Parameters that configure the sleep mode characteristics. (See Managing End
Devices chapter for details.)
The XBee ZB end device sends regular poll transmissions to its parent when it is awake. These poll transmissions
query the parent for any new received data packets. The parent always sends a MAC layer acknowledgment back
to the end device. The acknowledgment indicates whether the parent has data for the end device or not.
If the end device does not receive an acknowledgment for 3 consecutive poll requests, it considers itself
disconnected from its parent and will attempt to discover and join a valid ZigBee network. (See "Managing End
Devices" chapter for details.)
Resetting the End Device
When the end device is reset or power cycled, if the orphan scan successfully locates a parent, the end device then
checks its PAN ID, operating channel and stack profile against the network configuration settings (ID, SC, and ZS). It
also verifies the saved security policy is valid based on the security configuration commands (EE, KY). If the end
device's PAN ID, operating channel, stack profile, or security policy is invalid, the end device will leave the network
and attempt to join a new network based on its network joining command values.
To prevent the end device from leaving an existing network, the WR command should be issued after all network
joining commands have been configured in order to retain these settings through power cycle or reset events.
Leaving a Network
There are a couple of mechanisms that will cause the end device to leave its current PAN and attempt to discover
and join a new network based on its network joining parameter values. These include the following:
•The ID command changes such that the current 64-bit PAN ID is invalid.
•The SC command changes such that the current operating channel (CH) is not included in the channel
mask.
•The ZS or any of the security command values change.
•The NR0 command is issued to cause the end device to leave.
•The NR1 command is issued to send a broadcast transmission, causing all devices in the network to leave
and migrate to a different channel.
•The commissioning button is pressed 4 times or the CB command is issued with a parameter of 4.
•The end device's parent is powered down or the end device is moved out of range of the parent such
that the end device fails to receive poll acknowledgment messages.
Note that changes to command values only take effect when changes are applied (AC or CN commands).
Example: Joining a Network
After starting a coordinator (that is allowing joins), the following steps will cause an XBee end device to join the
network:
1. Set ID to the desired 64-bit PAN ID, or to 0 to join any PAN.
2. Set SC to the list of channels to scan to find a valid network.
3. If SC or ID is changed from the default, apply changes (make SC and ID changes take effect) by issuing
the AC or CN command.
4. The Associate LED will start blinking once the end device has joined a PAN.
5. If the Associate LED is not blinking, the AI command can be read to determine the cause of join failure.
6. Once the end device has joined, the OP and CH commands will indicate the operating 64-bit PAN ID and
channel the end device joined.
7. The MY command will reflect the 16-bit address the router received when it joined.
8. The API Modem Status frame ("Associated") is sent out the UART (API firmware only).
9. The joined end device will attempt to enter low power sleep modes based on its sleep configuration
commands (SM, SP, SN, ST, SO).
As mentioned previously, routers and end devices must scan one or more channels to discover a valid network to
join. When a join attempt begins, the XBee sends a beacon request transmission on the lowest channel specified in
the SC (scan channels) command bitmask. If a valid PAN is found on the channel, the XBee will attempt to join the
PAN on that channel. Otherwise, if a valid PAN is not found on the channel, it will attempt scanning on the next
higher channel in the SC command bitmask.
The XBee will continue to scan each channel (from lowest to highest) in the SC bitmask until a valid PAN is found or
all channels have been scanned. Once all channels have been scanned, the next join attempt will start scanning on
the lowest channel specified in the SC command bitmask.
For example, if the SC command is set to 0x400F, the XBee would start scanning on channel 11 (0x0B) and scan
until a valid beacon is found, or until channels 11, 12, 13, 14, and 25 have been scanned (in that order). Once an
XBee router or end device joins a network on a given channel, if the XBee is told to leave (see "Leaving a Network"
section), it will leave the channel it joined on and continue scanning on the next higher channel in the SC bitmask.
For example, if the SC command is set to 0x400F, and the XBee joins a PAN on channel 12 (0x0C), if the XBee leaves
the channel, it will start scanning on channel 13, followed by channels 14 and 25 if a valid network is not found.
Once all channels have been scanned, the next join attempt will start scanning on the lowest channel specified in
the SC command bitmask.
Managing Multiple ZigBee Networks
In some applications, multiple ZigBee networks may exist in proximity of each other. The application may need
provisions to ensure the XBee joins the desired network. There are a number of features in ZigBee to manage
joining among multiple networks. These include the following:
•PAN ID Filtering
•Preconfigured Security Keys
•Permit Joining
•Application Messaging
Pan ID Filtering
The XBee can be configured with a fixed PAN ID by setting the ID command to a non-zero value. If the PAN ID is set
to a non-zero value, the XBee will only join a network with the same PAN ID.
Preconfigured Security Keys
Similar to PAN ID filtering, this method requires a known security key be installed on a router to ensure it will join a
ZigBee network with the same security key. If the security key (KY command) is set to a non-zero value, and if
security is enabled (EE command), an XBee router or end device will only join a network with the same security
key.
Permit Joining
The Permit Joining parameter can be disabled in a network to prevent unwanted devices from joining. When a new
device must be added to a network, permit-joining can be enabled for a short time on the desired network. In the
XBee firmware, joining is disabled by setting the NJ command to a value less than 0xFF on all routers and
coordinator devices. Joining can be enabled for a short time using the commissioning push-button (see Network
Commissioning chapter for details) or the CB command.
Application Messaging
If the above mechanisms are not feasible, the application could build in a messaging framework between the
coordinator and devices that join its network. For example, the application code in joining devices could send a
transmission to the coordinator after joining a network, and wait to receive a defined reply message. If the
application does not receive the expected response message after joining, the application could force the XBee to
leave and continue scanning (see NR parameter).
All ZigBee devices have two different addresses, a 64-bit and a 16-bit address. The characteristics of each are
described below.
64-bit device Address
The 64-bit address is a unique device address assigned during manufacturing. This address is unique to each
physical device. The 64-bit address includes a 3-byte Organizationally Unique Identifier (OUI) assigned by the IEEE.
The 64-bit address is also called the extended address.
16-bit Device Address
A device receives a 16-bit address when it joins a ZigBee network. For this reason, the 16-bit address is also called
the "network address". The 16-bit address of 0x0000 is reserved for the coordinator. All other devices receive a
randomly generated address from the router or coordinator device that allows the join. The 16-bit address can
change under certain conditions:
•An address conflict is detected where two devices are found to have the same 16-bit address
•A device leaves the network and later joins (it can receive a different address)
All ZigBee transmissions are sent using the source and destination 16-bit addresses. The routing tables on ZigBee
devices also use 16-bit addresses to determine how to route data packets through the network. However, since
the 16-bit address is not static, it is not a reliable way to identify a device.
To solve this problem, the 64-bit destination address is often included in data transmissions to guarantee data is
delivered to the correct destination. The ZigBee stack can discover the 16-bit address, if unknown, before
transmitting data to a remote.
Application Layer Addressing
ZigBee devices can support multiple application profiles, cluster IDs, and endpoints. (See "ZigBee Application
Layers - In Depth" in chapter 3.) Application layer addressing allows data transmissions to be addressed to specific
profile IDs, cluster IDs, and endpoints. Application layer addressing is useful if an application must •Interoperate
with other ZigBee devices outside of the Digi application profile
•Utilize service and network management capabilities of the ZDO
•Operate on a public application profile such as Home Controls or Smart Energy.
The API firmware provides a simple yet powerful interface that can easily send data to any profile ID, endpoint,
and cluster ID combination on any device in a ZigBee network.
Data Transmission
ZigBee data packets can be sent as either unicast or broadcast transmissions. Unicast transmissions route data
from one source device to one destination device, whereas broadcast transmissions are sent to many or all devices
in the network.
Broadcast Transmissions
Broadcast transmissions within the ZigBee protocol are intended to be propagated throughout the entire network
such that all nodes receive the transmission. To accomplish this, the coordinator and all routers that receive a
broadcast transmission will retransmit the packet three times.
Note: when a router or coordinator delivers a broadcast transmission to an end device child, the transmission is
only sent once (immediately after the end device wakes and polls the parent for any new data). See Parent
Operation section in chapter 6 for details.
Each node that transmits the broadcast will also create an entry in a local broadcast transmission table. This entry
is used to keep track of each received broadcast packet to ensure the packets are not endlessly transmitted. Each
entry persists for 8 seconds. The broadcast transmission table holds 8 entries.
For each broadcast transmission, the ZigBee stack must reserve buffer space for a copy of the data packet. This
copy is used to retransmit the packet as needed. Large broadcast packets will require more buffer space. This
information on buffer space is provided for general knowledge; the user does not and cannot change any buffer
spacing. Buffer spacing is handled automatically by the XBee module.
Since broadcast transmissions are retransmitted by each device in the network, broadcast messages should be
used sparingly.
Unicast Transmissions
Unicast transmissions are sent from one source device to another destination device. The destination device could
be an immediate neighbor of the source, or it could be several hops away. Unicast transmissions that are sent
along a multiple hop path require some means of establishing a route to the destination device. See the "RF Packet
Routing" section in chapter 4 for details.
Address Resolution
As mentioned previously, each device in a ZigBee network has both a 16-bit (network) address and a 64-bit
(extended) address. The 64-bit address is unique and assigned to the device during manufacturing, and the 16-bit
address is obtained after joining a network. The 16-bit address can also change under certain conditions.
When sending a unicast transmission, the ZigBee network layer uses the 16-bit address of the destination and each
hop to route the data packet. If the 16-bit address of the destination is not known, the ZigBee stack includes a
discovery provision to automatically discover the destination device's 16-bit address before routing the data.
To discover a 16-bit address of a remote, the device initiating the discovery sends a broadcast address discovery
transmission. The address discovery broadcast includes the 64-bit address of the remote device whose 16-bit
address is being requested. All nodes that receive this transmission check the 64-bit address in the payload and
compare it to their own 64-bit address. If the addresses match, the device sends a response packet back to the
initiator. This response includes the remote's 16-bit address. When the discovery response is received, the initiator
will then transmit the data.
Address Table
Each ZigBee device maintains an address table that maps a 64-bit address to a 16-bit address. When a transmission
is addressed to a 64-bit address, the ZigBee stack searches the address table for an entry with a matching 64-bit
address, in hopes of determining the destination's 16-bit address. If a known 16-bit address is not found, the
ZigBee stack will perform address discovery to discover the device's current 16- bit address.
Sample Address Table
64-bit Address 16-bit Address
0013 A200 4000 0001 0x4414
0013 A200 400A 3568 0x1234
0013 A200 4004 1122 0xC200
0013 A200 4002 1123 0xFFFE (unknown)
The XBee modules can store up to 20 address table entries. For applications where a single device (e.g.
coordinator) may send unicast transmissions to more than 20 devices, the application should implement an
address table to store the 16-bit and 64-bit addresses for each remote device. Any XBee that will send data to
more than 20 remotes should also use API firmware. The application can then send both the 16-bit and 64-bit
addresses to the XBee in the API transmit frames which will significantly reduce the number of 16-bit address
discoveries and greatly improve data throughput.
If an application will support an address table, the size should ideally be larger than the maximum number of
destination addresses the device will communicate with. Each entry in the address table should contain a 64-bit
destination address and its last known 16-bit address.
When sending a transmission to a destination 64-bit address, the application should search the address table for a
matching 64-bit address. If a match is found, the 16-bit address should be populated into the 16-bit address field of
the API frame. If a match is not found, the 16-bit address should be set to 0xFFFE (unknown) in the API transmit
frame.
The API provides indication of a remote device's 16-bit address in the following frames:
•All receive data frames
Rx Data (0x90)
Rx Explicit Data (0x91)
IO Sample Data (0x92)
Node Identification Indicator (0x95)
Route Record Indicator (0xA1)
etc.
•Transmit status frame (0x8B)
The application should always update the 16-bit address in the address table when one of these frames is received
to ensure the table has the most recently known 16-bit address. If a transmission failure occurs, the application
should set the 16-bit address in the table to 0xFFFE (unknown).
Each unicast transmission may support up to 84 bytes of RF payload. (Enabling security or using source routing can
reduce this number. See the NP command for details.) However, the XBee ZB firmware supports a ZigBee feature
called fragmentation that allows a single large data packet to be broken up into multiple RF transmissions and
reassembled by the receiver before sending data out its UART or SPI port. This is shown in the image below.
The API transmit frame can include up to 255 bytes of data, which will be broken up into multiple transmissions
and reassembled on the receiving side. If one or more of the fragmented messages are not received by the
receiving device, the receiver will drop the entire message, and the sender will indicate a transmission failure in
the Tx Status API frame.
Applications that do not wish to use fragmentation should avoid sending more than the maximum number of bytes
in a single RF transmission. See the "Maximum RF Payload Size" section for details.
Devices will not receive or reassemble fragmented RF packets if RTS flow control is enabled (D6 command).
DATA Transmission Examples
AT Firmware
To send a data packet in AT firmware, the DH and DL commands must be set to match the 64-bit address of the
destination device. DH must match the upper 4-bytes, and DL must match the lower 4 bytes. Since the coordinator
always receives a 16-bit address of 0x0000, a 64-bit address of 0x0000000000000000 is defined as the
coordinator's address (in ZB firmware). The default values of DH and DL are 0x00, which sends data to the
coordinator.
Example 1: Send a transmission to the coordinator
(In this example, a '\r' refers to a carriage return character.)
A router or end device can send data in two ways. First, set the destination address (DH and DL commands) to
0x00.
1. Enter command mode ('+++')
2. After receiving an OK\r, issue the following commands: a. ATDH0\r b. ATDL0\r c. ATCN\r
3. Verify that each of the 3 commands returned an OK\r response.
4. After setting these command values, all serial characters will be sent as a unicast transmission to the
coordinator.
Alternatively, if the coordinator's 64-bit address is known, DH and DL can be set to the coordinator's 64-bit
address. Suppose the coordinator's address is 0x0013A200404A2244.
1. Enter command mode ('+++')
2. After receiving an OK\r, issue the following commands: a. ATDH13A200\r b. ATDL404A2244\ c. ATCN\r
3. Verify that each of the 3 commands returned an OK\r response.
4. After setting these command values, all serial characters will be sent as a unicast transmission to the
coordinator.
API Firmware
Use the transmit request, or explicit transmit request frame (0x10 and 0x11 respectively) to send data to the
coordinator. The 64-bit address can either be set to 0x0000000000000000, or to the 64-bit address of the
coordinator. The 16-bit address should be set to 0xFFFE when using the 64-bit address of all 0x00s.
To send an ascii "1" to the coordinator's 0x00 address, the following API frame can be used:
If the explicit transmit frame is used, the cluster ID should be set to 0x0011, the profile ID to 0xC105, and the
source and destination endpoints to 0xE8 (recommended defaults for data transmissions in the Digi profile.) The
same transmission could be sent using the following explicit transmit frame:
Notice the 16-bit address is set to 0xFFFE. This is required when sending to a 64-bit address of 0x00s.
Now suppose the coordinator's 64-bit address is 0x0013A200404A2244. The following transmit request API frame
(0x10) will send an ASCII "1" to the coordinator:
Notice the destination 16-bit address is set to 0xFFFE for broadcast transmissions.
RF Packet Routing
Unicast transmissions may require some type of routing. ZigBee includes several different ways to route data, each
with its own advantages and disadvantages. These are summarized in the table below.
source and destination, possibly
traversing multiple nodes ("hops").
Each device knows who to send data to
eventually reaching the destination.
configured reverse routes on all devices
into the device that sends the broadcast
the packet should traverse to get from
source to destination.
beyond about 40 destination devices.
send data to a single gateway or
collector device.
networks (over 40 remote devices)
Note – End devices do not make use of these routing protocols. Rather, an end device sends a unicast transmission
to its parent and allows the parent to route the data packet in its behalf.
Link Status Transmission
Before discussing the various routing protocols, it is worth understanding the primary mechanism in ZigBee for
establishing reliable bi-directional links. This mechanism is especially useful in networks that may have a mixture of
devices with varying output power and/or receiver sensitivity levels.
Each coordinator or router device periodically sends a link status message. This message is sent as a 1-hop
broadcast transmission, received only by one-hop neighbors. The link status message contains a list of neighboring
devices and incoming and outgoing link qualities for each neighbor. Using these messages, neighboring devices can
determine the quality of a bi-directional link with each neighbor and use that information to select a route that
works well in both directions.
For example, consider a network of two neighboring devices that send periodic link status messages. Suppose that
the output power of device A is +18dBm, and the output power of device B is +3dBm (considerably less than the
output power of device A). The link status messages might indicate the following:
This mechanism enables devices A and B to recognize that the link is not reliable in both directions and select a
different neighbor when establishing routes. (Such links are called asymmetric links, meaning the link quality is not
similar in both directions.)
When a router or coordinator device powers on, it sends link status messages every couple seconds to attempt to
discover link qualities with its neighbors quickly. After being powered on for some time, the link status messages
are sent at a much slower rate (about every 3-4 times per minute).
AODV Mesh Routing
ZigBee employs mesh routing to establish a route between the source device and the destination. Mesh routing
allows data packets to traverse multiple nodes (hops) in a network to route data from a source to a destination.
Routers and coordinators can participate in establishing routes between source and destination devices using a
process called route discovery. The Route discovery process is based on the AODV (Ad-hoc On-demand Distance
Vector routing) protocol.
Routing under the AODV protocol is accomplished using tables in each node that store the next hop (intermediary
node between source and destination nodes) for a destination node. If a next hop is not known, route discovery
must take place in order to find a path. Since only a limited number of routes can be stored on a Router, route
discovery will take place more often on a large network with communication between many different nodes.
When a source node must discover a route to a destination node, it sends a broadcast route request command.
The route request command contains the source network address, the destination network address and a path
cost field (a metric for measuring route quality). As the route request command is propagated through the
network (refer to the Broadcast Transmission), each node that re-broadcasts the message updates the path cost
field and creates a temporary entry in its route discovery table.
Sample Route Request (Broadcast) Transmission Where R3 is Trying to Discover a Route to R6
When the destination node receives a route request, it compares the ‘path cost’ field against previously received
route request commands. If the path cost stored in the route request is better than any previously received, the
destination node will transmit a route reply packet to the node that originated the route request. Intermediate
nodes receive and forward the route reply packet to the source node (the node that originated route request).
Sample Route Reply (Unicast) Where R6 Sends a Route Reply to R3.
Note: R6 could send multiple replies if it identifies a better route.
Retries and Acknowledgments
ZigBee includes acknowledgment packets at both the Mac and Application Support (APS) layers. When data is
transmitted to a remote device, it may traverse multiple hops to reach the destination. As data is transmitted from
one node to its neighbor, an acknowledgment packet (Ack) is transmitted in the opposite direction to indicate that
the transmission was successfully received. If the Ack is not received, the transmitting device will retransmit the
data, up to 4 times. This Ack is called the Mac layer acknowledgment.
In addition, the device that originated the transmission expects to receive an acknowledgment packet (Ack) from
the destination device. This Ack will traverse the same path that the data traversed, but in the opposite direction.
If the originator fails to receive this Ack, it will retransmit the data, up to 2 times until an Ack is received. This Ack is
called the ZigBee APS layer acknowledgment.
Refer to the ZigBee specification for more details.
Many-to-One Routing
In networks where many devices must send data to a central collector or gateway device, AODV mesh routing
requires significant overhead. If every device in the network had to discover a route before it could send data to
the data collector, the network could easily become inundated with broadcast route discovery messages.
Many-to-one routing is an optimization for these kinds of networks. Rather than require each device to do its own
route discovery, a single many-to-one broadcast transmission is sent from the data collector to establish reverse
routes on all devices. This is shown in the figure below. The left side shows the many broadcasts the devices can
send when they create their own routes and the route replies generated by the data collector. The right side shows
the benefits of many-to-one routing where a single broadcast creates reverse routes to the data collector on all
routers.
The many-to-one broadcast is a route request message with the target discovery address set to the address of the
data collector. Devices that receive this route request create a reverse many-to-one routing table entry to create a
path back to the data collector. The ZigBee stack on a device uses historical link quality information about each
neighbor to select a reliable neighbor for the reverse route.
When a device sends data to a data collector, and it finds a many-to-one route in its routing table, it will transmit
the data without performing a route discovery. The many-to-one route request should be sent periodically to
update and refresh the reverse routes in the network.
Applications that require multiple data collectors can also use many-to-one routing. If more than one data
collector device sends a many-to-one broadcast, devices will create one reverse routing table entry for each
collector.
In ZB firmware, the AR command is used to enable many-to-one broadcasting on a device. The AR command sets a
time interval (measured in 10 second units) for sending the many to one broadcast transmission. (See the
command table for details.)
Source Routing
In applications where a device must transmit data to many remotes, AODV routing would require performing one
route discovery for each destination device to establish a route. If there are more destination devices than there
are routing table entries, established AODV routes would be overwritten with new routes, causing route
discoveries to occur more regularly. This could result in larger packet delays and poor network performance.
ZigBee source routing helps solve these problems. In contrast to many-to-one routing that establishes routing
paths from many devices to one data collector, source routing allows the collector to store and specify routes for
many remotes.
To use source routing, a device must use the API firmware, and it must send periodic many-to-one route request
broadcasts (AR command) to create a many-to-one route to it on all devices. When remote devices send RF data
using a many-to-one route, they first send a route record transmission. The route record transmission is unicast
along the many-to-one route until it reaches the data collector. As the route record traverses the many-to-one
route, it appends the 16-bit address of each device in the route into the RF payload. When the route record
reaches the data collector, it contains the address of the sender, and the 16-bit address of each hop in the route.
The data collector can store the routing information and retrieve it later to send a source routed packet to the
remote. This is shown in the images below.
Acquiring source routes requires the remote devices to send a unicast to a data collector (device that sends many-to-one route
request broadcasts). There are several ways to force remotes to send route record transmissions.
1. If the application on remote devices periodically sends data to the data collector, each transmission will force a route
record to occur.
2. The data collector can issue a network discovery command (ND command) to force all XBee devices to send a network
discovery response. Each network discovery response will be prefaced by a route record.
3. Periodic IO sampling can be enabled on remotes to force them to send data at a regular rate. Each IO sample would be
prefaced by a route record. (See chapter 8 for details.)
4. If the NI string of the remote device is known, the DN command can be issued with the NI string of the remote in the
payload. The remote device with a matching NI string would send a route record and a DN response.
Storing Source Routes
When a data collector receives a route record, it sends it out the UART as a Route Record Indicator API frame
(0xA1). To use source routing, the application should receive these frames and store the source route information.
Sending a Source Routed Transmission
To send a source routed transmission, the application should send a Create Source Route API frame (0x21) to the
XBee to create a source route in its internal source route table. After sending the Create Source Route API frame,
the application can send data transmission or remote command request frames as needed to the same
destination, or any destination in the source route. Once data must be sent to a new destination (a destination not
included in the last source route), the application should first send a new Create Source Route API frame. The XBee
can buffer one source route that includes up to 10 hops (excluding source and destination).
For example, suppose a network exists with a coordinator and 5 routers (R1, R2, R3, R4, and R5) with known
source routes as shown below.
To send a source-routed packet to R3, the application must send a Create Source Route API frame (0x21) to the
XBee, with a destination of R3, and 2 hops (R1 and R2). If the 64- bit address of R3 is 0x0013A200 404a1234 and
the 16-bit addresses of R1, R2, and R3 are:
0x0012 - length
0x21 - API ID (create source route)
0x00 - frame ID (set to 0 always)
0x0013A200 404A1234 - 64-bit address of R3 (destination)
0xEEFF - 16-bit address of R3 (destination)
0x00 - Route options (set to 0)
0x02 - Number of intermediate devices in the source route
0xCCDD - Address of furthest device (1-hop from target)
0xAABB - Address of next-closer device
0x5C - Checksum (0xFF - SUM (all bytes after length))
Repairing Source Routes
It is possible in a network to have an existing source route fail (i.e. a device in the route moves or goes down, etc.).
If a device goes down in a source routed network, all routes that used the device will be broken.
As mentioned previously, source routing must be used with many-to-one routing. (A device that uses source
routing must also send a periodic many-to-one broadcast in order to keep routes fresh). If a source route is broken,
remote devices must send in new route record transmissions to the data collector to provide it with a new source
route. This requires that remote devices periodically send data transmissions into the data collector. See the
earlier "Acquiring Source Routes" section for details.
ZigBee includes acknowledgment packets at both the Mac and Application Support (APS) layers. When data is
transmitted to a remote device, it may traverse multiple hops to reach the destination. As data is transmitted from
one node to its neighbor, an acknowledgment packet (Ack) is transmitted in the opposite direction to indicate that
the transmission was successfully received. If the Ack is not received, the transmitting device will retransmit the
data, up to 4 times. This Ack is called the Mac layer acknowledgment.
In addition, the device that originated the transmission expects to receive an acknowledgment packet (Ack) from
the destination device. This Ack will traverse the same path that the data traversed, but in the opposite direction.
If the originator fails to receive this Ack, it will retransmit the data, up to 2 times until an Ack is received. This Ack is
called the ZigBee APS layer acknowledgment.
Refer to the ZigBee specification for more details.
Encrypted Transmissions
Encrypted transmissions are routed similar to non-encrypted transmissions with one exception. As an encrypted
packet propagates from one device to another, each device decrypts the packet using the network key, and
authenticates the packet by verifying packet integrity. It then re-encrypts the packet with its own source address
and frame counter values, and sends the message to the next hop. This process adds some overhead latency to
unicast transmissions, but it helps prevent replay attacks. See chapter 5 for details.
Maximum RF Payload Size
XBee ZB firmware includes a command (ATNP) that returns the maximum number of RF payload bytes that can be
sent in a unicast transmission. Querying the NP command, like most other commands, returns a HEXADECIMAL
value. This number will change based on whether security is enabled or not. If security is enabled (EE command),
the maximum number of RF payload bytes decreases since security requires additional overhead.
After reading the NP value, the following conditions can affect the maximum number of data bytes in a single RF
transmission:
•Broadcast transmissions can support 8 bytes more than unicast transmissions.
•If source routing is used, the 16-bit addresses in the source route are inserted into the RF payload space.
For example, if NP returns 84 bytes, and a source route must traverse 3 intermediate hops (3 16-bit
addresses); the total number of bytes that can be sent in one RF packet is 78.
•Enabling APS encryption (API tx option bit set) will reduce the number of payload bytes by 4.
Throughput
Throughput in a ZigBee network can vary by a number of variables, including: number of hops, encryption
enabled/ disabled, sleeping end devices, failures/route discoveries. Our empirical testing showed the following
throughput performance in a robust operating environment (low interference).
* Data throughput measurements were made setting the serial interface rate to 115200 bps, and measuring the
time to send 100,000 bytes from source to destination. During the test, no route discoveries or failures occurred.
ZDO Transmissions
ZigBee defines a ZigBee Device Objects layer (ZDO) that can provide device and service discovery and network
management capabilities. This layer is described below.
ZigBee Device Objects (ZDO)
The ZigBee Device Objects (ZDO) is supported to some extent on all ZigBee devices. The ZDO is an endpoint that
implements services described in the ZigBee Device Profile in the ZigBee specification. Each service has an assigned
cluster ID, and most service requests have an associated response. The following table describes some common
ZDO services.
Cluster Name Cluster ID Description
Request a 16-bit address of the
Request
radio with a matching 64-bit
address (required parameter).
Request a list of endpoints from a
Request
remote device.
Request data from a neighbor table
of a remote device.
Request to retrieve routing table
Request
entries from a remote device.
Response that includes the 16-bit
Response
address of a device.
Cluster Name Cluster ID Description
Response that includes a neighbor
0x8031
table data from a remote device.
Response that includes routing
Routing Table
Response 0x8032
Refer to the ZigBee specification for a detailed description of all ZigBee Device Profile services.
Sending a ZDO Command
To send a ZDO command, an explicit transmit API frame must be used and formatted correctly. The source and
destination endpoints must be set to 0, and the profile ID must be set to 0. The cluster ID must be set to match the
cluster ID of the appropriate service. For example, to send an active endpoints request, the cluster ID must be set
to 0x0005.
The first byte of payload in the API frame is an application sequence number (transaction sequence number) that
can be set to any single byte value. This same value will be used in the first byte of the ZDO response. All remaining
payload bytes must be set as required by the ZDO. All multi-byte values must be sent in little endian byte order.
Receiving ZDO Commands and Responses
In XBee ZB firmware, ZDO commands can easily be sent using the API. In order to receive incoming ZDO
commands, receiver application addressing must be enabled with the AO command. (See examples later in this
section.) Not all incoming ZDO commands are passed up to the application.
When a ZDO message is received on endpoint 0 and profile ID 0, the cluster ID indicates the type of ZDO message
that was received. The first byte of payload is generally a sequence number that corresponds to a sequence
number of a request. The remaining bytes are set as defined by the ZDO. Similar to a ZDO request, all multi-byte
values in the response are in little endian byte order.
Example 1: Send a ZDO LQI Request to read the neighbor table contents of a remote.
Looking at the ZigBee specification, the cluster ID for an LQI Request is 0x0031, and the payload only
requires a single byte (start index). This example will send an LQI request to a remote device with a 64-bit
address of 0x0013A200 40401234. The start index will be set to 0, and the transaction sequence number
will be set to 0x76
API Frame:
7E 0016 11 01 0013A200 40401234 FFFE 00 00 0031 0000 00 00 76 00 CE
0x0016 - length
0x11 - Explicit transmit request
0x01 - frame ID (set to a non-zero value to enable the transmit status message, or set to 0 to disable)
0x0013A200 40401234 - 64-bit address of the remote
0xFFFE - 16-bit address of the remote (0xFFFE = unknown). Optionally, set to the 16-bit address of the
destination if known.
0x00 - Source endpoint
0x00 - Destination endpoint
0x0031 - Cluster ID (LQI Request, or Neighbor table request)
0xCE - Checksum (0xFF - SUM (all bytes after length))
Description:
This API frame sends a ZDO LQI request (neighbor table request) to a remote device to obtain data from
its neighbor table. Recall that the AO command must be set correctly on an API device to enable the
explicit API receive frames in order to receive the ZDO response.
Example 2: Send a ZDO Network Address Request to discover the 16-bit address of a remote.
Looking at the ZigBee specification, the cluster ID for a network Address Request is 0x0000, and the
payload only requires the following:
[64-bit address] + [Request Type] + [Start Index]
This example will send a Network Address Request as a broadcast transmission to discover the 16-bit
address of the device with a 64-bit address of 0x0013A200 40401234. The request type and start index
will be set to 0, and the transaction sequence number will be set to 0x44
0x33 - Checksum (0xFF - SUM (all bytes after length))
Description:
This API frame sends a broadcast ZDO Network Address Request to obtain the 16-bit address of a device
with a 64-bit address of 0x0013A200 40401234. Note the bytes for the 64-bit address were inserted in
little endian byte order. All multi-byte fields in the API payload of a ZDO command must have their data
inserted in little endian byte order. Also recall that the AO command must be set correctly on an API
device to enable the explicit API receive frames in order to receive the ZDO response.
Transmission Timeouts
The ZigBee stack includes two kinds of transmission timeouts, depending on the nature of the destination device.
For destination devices such as routers whose receiver is always on, a unicast timeout is used. The unicast timeout
estimates a timeout based on the number of unicast hops the packet should traverse to get data to the destination
device. For transmissions destined for end devices, the ZigBee stack uses an extended timeout that includes the
unicast timeout (to route data to the end device's parent), and it includes a timeout for the end device to finish
sleeping, wake, and poll the parent for data.
The ZigBee stack includes some provisions for a device to detect if the destination is an end device or not. The
ZigBee stack uses the unicast timeout unless it knows the destination is an end device.
The XBee API includes a transmit options bit that can be set to specify if the extended timeout should be used for a
given transmission. If this bit is set, the extended timeout will be used when sending RF data to the specified
destination. To improve routing reliability, applications should set the extended timeout bit when sending data to
end devices if:
•The application sends data to 10 or more remote devices, some of which are end devices, AND
•The end devices may sleep longer than the unicast timeout
Equations for these timeouts are computed in the following sections.
Note: The timeouts in this section are worst-case timeouts and should be padded by a few hundred milliseconds.
These worst-case timeouts apply when an existing route breaks down (e.g. intermediate hop or destination device
moved).
Unicast Timeout
The unicast timeout is settable with the NH command. The actual unicast timeout is computed as ((50 * NH) +
100). The default NH value is 30 which equates to a 1.6 second timeout.
The unicast timeout includes 3 transmission attempts (1 attempt and 2 retries). The maximum total timeout is
about:
3 * ((50 * NH) + 100).
For example, if NH=30 (0x1E), the unicast timeout is about
3 * ((50 * 30) + 100), or
3 * (1500 + 100), or
3 * (1600), or
4800 ms, or
4.8 seconds.
Extended Timeout
The worst-case transmission timeout when sending data to an end device is somewhat larger than when
transmitting to a router or coordinator. As described later in chapter 6, RF data packets are actually sent to the
parent of the end device, who buffers the packet until the end device wakes to receive it. The parent will buffer an
RF data packet for up to (1.2 * SP) time.
To ensure the end device has adequate time to wake and receive the data, the extended transmission timeout to
an end device is:
(50 * NH) + (1.2 * SP)
This timeout includes the packet buffering timeout (1.2 * SP) and time to account for routing through the mesh
network (50 * NH).
If an acknowledgment is not received within this time, the sender will resend the transmission up to two more
times. With retries included, the longest transmission timeout when sending data to an end device is:
The SP value in both equations must be entered in millisecond units. (The SP command setting uses 10ms units and
must be converted to milliseconds to be used in this equation.)
For example, suppose a router is configured with NH=30 (0x1E) and SP=0x3E8 (10,000 ms), and that it is either
trying to send data to one of its end device children, or to a remote end device. The total extended timeout to the
end device is about:
3 * ((50 * NH) + (1.2 * SP)), or
3 * (1500 + 12000), or
3 * (13500), or
40500 ms, or
40.5 seconds.
Transmission Examples
Example 1: Send a unicast API data transmission to the coordinator using 64-bit address 0, with payload
"TxData".
API Frame:
7E 0014 10 01 00000000 00000000 FFFE 00 00 54 78 44 61 74 61 AB
Field Composition:
0x0014 - length
0x10 - API ID (tx data)
0x01 - frame ID (set greater than 0 to enable the tx-status response)
0x00000000 00000000 - 64-bit address of coordinator (ZB definition)
0xFFFE - Required 16-bit address if sending data to 64-bit address of 0.
0xAB - Checksum (0xFF - SUM (all bytes after length))
Description:
This transmission sends the string "TxData" to the coordinator, without knowing the coordinator device's
64-bit address. A 64-bit address of 0 is defined as the coordinator in ZB firmware. If the coordinator's 64bit address was known, the 64-bit address of 0 could be replaced with the coordinator's 64-bit address,
and the 16-bit address could be set to 0.
Example 2 - Send a broadcast API data transmission that all devices can receive (including sleeping end
devices), with payload "TxData".
API Frame:
7E 0014 10 01 00000000 0000FFFF FFFE 00 00 54 78 44 61 74 61 AD
Field Composition:
0x0014 - length
0x10 - API ID (tx data)
0x01 - frame ID (set to a non-zero value to enable the tx-status response)
0x00000000 0000FFFF - Broadcast definition (including sleeping end devices
0xAD - Checksum (0xFF - SUM (all bytes after length))
Description:
This transmission sends the string "TxData" as a broadcast transmission. Since the destination address is
set to 0xFFFF, all devices, including sleeping end devices can receive this broadcast.
If receiver application addressing is enabled, the XBee will report all received data frames in the explicit
format (0x91) to indicate the source and destination endpoints, cluster ID, and profile ID that each packet
was received on. (Status messages like modem status and route record indicators are not affected.)
To enable receiver application addressing, set the AO command to 1 using the AT command frame (0x08).
Here's how to do this:
API Frame:
7E 0005 08 01 414F 01 65
Field Composition:
0x0005 - length
0x08 - API ID (at command)
0x01 - frame ID (set to a non-zero value to enable AT command response frames)
0x414F - ASCII representation of 'A','O' (the command being issued) 0x01 - Parameter value
0x65 - Checksum (0xFF - SUM (all bytes after length))
Description:
Setting AO=1 is required for the XBee to use the explicit receive API frame (0x91) when RF data packets
are received. This is required if the application needs indication of source or destination endpoint, cluster
ID, and/ or profile ID values used in received ZigBee data packets. ZDO messages can only be received if
AO=1.
ZigBee supports various levels of security that can be configured depending on the needs of the application.
Security provisions include:
•128-bit AES encryption
•Two security keys that can be preconfigured or obtained during joining
•Support for a trust center
•Provisions to ensure message integrity, confidentiality, and authentication.
The first half of this chapter describes various security features defined in the ZigBee-PRO specification, while the
last half illustrates how the XBee and XBee-PRO modules can be configured to support these features
Security Modes
The ZigBee standard supports three security modes – residential, standard, and high security. Residential security
was first supported in the ZigBee 2006 standard. This level of security requires a network key be shared among
devices. Standard security adds a number of optional security enhancements over residential security, including an
APS layer link key. High security adds entity authentication, and a number of other features not widely supported.
XBee ZB modules primarily support standard security, although end devices that support residential security can
join and interoperate with standard security devices. The remainder of this chapter focuses on material that is
relevant to standard security.
ZigBee Security Model
ZigBee security is applied to the Network and APS layers. Packets are encrypted with 128-bit AES encryption. A
network key and optional link key can be used to encrypt data. Only devices with the same keys are able to
communicate together in a network. Routers and end devices that will communicate on a secure network must
obtain the correct security keys.
Network Layer Security
The network key is used to encrypt the APS layer and application data. In addition to encrypting application
messages, network security is also applied to route request and reply messages, APS commands, and ZDO
commands. Network encryption is not applied to MAC layer transmissions such as beacon transmissions, etc. If
security is enabled in a network, all data packets will be encrypted with the network key.
Packets are encrypted and authenticated using 128-bit AES. This is shown in the figure below.
The network header of encrypted packets includes a 32-bit frame counter. Each device in the network maintains a
32-bit frame counter that is incremented for every transmission. In addition, devices track the last known 32- bit
frame counter for each of its neighbors. If a device receives a packet from a neighbor with a smaller frame counter
than it has previously seen, the packet is discarded. The frame counter is used to protect against replay attacks.
If the frame counter reaches a maximum value of 0xFFFFFFFF, it does not wrap to 0 and no more transmissions can
be sent. Due to the size of the frame counters, reaching the maximum value is a very unlikely event for most
applications. The following table shows the required time under different conditions, for the frame counter to
reach its maximum value.
Average Transmission Rate Time until 32-bit frame counter expires
To clear the frame counters without compromising security, the network key can be changed in the network.
When the network key is updated, the frame counters on all devices reset to 0. (See the Network Key Updates
section for details.)
Message Integrity code
The network header, APS header, and application data are all authenticated with 128-bit AES. A hash is performed
on these fields and is appended as a 4-byte message integrity code (MIC) to the end of the packet. The MIC allows
receiving devices to ensure the message has not been changed. The MIC provides message integrity in the ZigBee
security model. If a device receives a packet and the MIC does not match the device’s own hash of the data, the
packet is dropped.
Network Layer Encryption and Decryption
Packets with network layer encryption are encrypted and decrypted by each hop in a route. When a device
receives a packet with network encryption, it decrypts the packet and authenticates the packet. If the device is not
the destination, it then encrypts and authenticates the packet, using its own frame counter and source address in
the network header section.
Since network encryption is performed at each hop, packet latency is slightly longer in an encrypted network than
in a non-encrypted network. Also, security requires 18 bytes of overhead to include a 32-bit frame counter, an 8byte source address, 4-byte MIC, and 2 other bytes. This reduces the number of payload bytes that can be sent in a
data packet.
Network Key Updates
ZigBee supports a mechanism for changing the network key in a network. When the network key is changed, the
frame counters in all devices reset to 0.
APS Layer Security
APS layer security can be used to encrypt application data using a key that is shared between source and
destination devices. Where network layer security is applied to all data transmissions and is decrypted and reencrypted on a hop-by-hop basis, APS security is optional and provides end-to-end security using an APS link key
that only the source and destination device know. APS security can be applied on a packet-by-packet basis. APS
security cannot be applied to broadcast transmissions.
If APS security is enabled, packets are encrypted and authenticated using 128-bit AES. This is shown in the figure
below:
If APS security is enabled, the APS header and data payload are authenticated with 128-bit AES. A hash is
performed on these fields and appended as a 4-byte message integrity code (MIC) to the end of the packet. This
MIC is different than the MIC appended by the network layer. The MIC allows the destination device to ensure the
message has not been changed. If the destination device receives a packet and the MIC does not match the
destination device’s own hash of the data, the packet is dropped.
APS Link Keys
There are two kinds of APS link keys – trust center link keys and application link keys. A trust center link key is
established between a device and the trust center, where an application link key is established between a device
and another device in the network where neither device is the trust center.
APS Layer Encryption and Decryption
Packets with APS layer encryption are encrypted at the source and only decrypted by the destination. Since APS
encryption requires a 5-byte header and a 4-byte MIC, the maximum data payload is reduced by 9 bytes when APS
encryption is used.
Network and APS Layer Encryption
Network and APS layer encryption can both be applied to data. The following figure demonstrates the
authentication and encryption performed on the final ZigBee packet when both are applied.
ZigBee defines a trust center device that is responsible for authenticating devices that join the network.
The trust center also manages link key distribution in the network.
Forming and Joining a secure Network
The coordinator is responsible for selecting a network encryption key. This key can either be preconfigured or
randomly selected. In addition, the coordinator generally operates as a trust center and must therefore select the
trust center link key. The trust center link key can also be preconfigured or randomly selected.
Devices that join the network must obtain the network key when they join. When a device joins a secure network,
the network and link keys can be sent to the joining device. If the joining device has a pre-configured trust center
link key, the network key will be sent to the joining device encrypted by the link key. Otherwise, if the joining
device is not pre-configured with the link key, the device could only join the network if the network key is sent
unencrypted (“in the clear”). The trust center must decide whether or not to send the network key unencrypted to
joining devices that are not pre-configured with the link key. Sending the network key unencrypted is not
recommended as it can open a security hole in the network. To maximize security, devices should be preconfigured with the correct link key.
Implementing Security on the XBee
If security is enabled in the XBee ZB firmware, devices acquire the network key when they join a network. Data
transmissions are always encrypted with the network key, and can optionally be end-to-end encrypted with the
APS link key. The following sections discuss the security settings and options in the XBee ZB firmware.
Enabling Security
To enable security on a device, the EE command must be set to 1. If the EE command value is changed and changes
are applied (e.g. AC command), the XBee module will leave the network (PAN ID and channel) it was operating on,
and attempt to form or join a new network.
If EE is set to 1, all data transmissions will be encrypted with the network key. When security is enabled, the
maximum number of bytes in a single RF transmission will be reduced. See the NP command for details.
Note: The EE command must be set the same on all devices in a network. Changes to the EE command should be
written to non-volatile memory (to be preserved through power cycle or reset events) using the WR command.
Setting the Network Security Key
The coordinator must select the network security key for the network. The NK command (write-only) is used to set
the network key. If NK=0 (default), a random network key will be selected. (This should suffice for most
applications.) Otherwise, if NK is set to a non-zero value, the network security key will use the value specified by
NK. NK is only supported on the coordinator.
Routers and end devices with security enabled (ATEE=1) acquire the network key when they join a network. They
will receive the network key encrypted with the link key if they share a pre-configured link key with the
coordinator. See the following section for details.
Setting the APS Trust Center Link Key
The coordinator must also select the trust center link key, using the KY command. If KY=0 (default), the coordinator
will select a random trust center link key (not recommended). Otherwise, if KY is set greater than 0, this value will
be used as the pre-configured trust center link key. KY is write-only and cannot be read.
Note: Application link keys (sent between two devices where neither device is the coordinator) are not supported
in ZB firmware at this time.
Random Trust Center Link Keys
If the coordinator selects a random trust center link key (KY=0, default), then it will allow devices to join the
network without having a pre-configured link key. However, this will cause the network key to be sent
unencrypted over-the-air to joining devices and is not recommended.
Pre-configured Trust Center Link Keys
If the coordinator uses a pre-configured link key (KY > 0), then the coordinator will not send the network key
unencrypted to joining devices. Only devices with the correct pre-configured link key will be able to join and
communicate on the network.
Enabling APS Encryption
APS encryption is an optional layer of security that uses the link key to encrypt the data payload. Unlike network
encryption that is decrypted and encrypted on a hop-by-hop basis, APS encryption is only decrypted by the
destination device. The XBee must be configured with security enabled (EE set to 1) to use APS encryption.
APS encryption can be enabled in API firmware on a per-packet basis. To enable APS encryption for a given
transmission, the "enable APS encryption" transmit options bit should be set in the API transmit frame. Enabling
APS encryption decreases the maximum payload size by 9 bytes.
Using a Trust Center
The EO command can be used to define the coordinator as a trust center. If the coordinator is a trust center, it will
be alerted to all new join attempts in the network. The trust center also has the ability to update or change the
network key on the network.
In ZB firmware, a secure network can be established with or without a trust center. Network and APS layer
encryption are supported if a trust center is used or not.
Updating the Network Key with a Trust Center
If the trust center has started a network and the NK value is changed, the coordinator will update the network key
on all devices in the network. (Changes to NK will not force the device to leave the network.) The network will
continue to operate on the same channel and PAN ID, but the devices in the network will update their network
key, increment their network key sequence number, and restore their frame counters to 0.
Updating the Network Key without a Trust Center
If the coordinator is not running as a trust center, the network reset command (NR1) can be used to force all
devices in the network to leave the current network and rejoin the network on another channel. When devices
leave and reform then network, the frame counters are reset to 0. This approach will cause the coordinator to
form a new network that the remaining devices should join. Resetting the network in this manner will bring the
coordinator and routers in the network down for about 10 seconds, and will likely cause the 16-bit PAN ID and 16bit addresses of the devices to change.
XBee Security Examples
This section covers some sample XBee configurations to support different security modes. Several AT commands
are listed with suggested parameter values. The notation in this section includes an '=' sign to indicate what each
command register should be set to - for example, EE=1. This is not the correct notation for setting command values
in the XBee. In AT command mode, each command is issued with a leading 'AT' and no '=' sign - for example ATEE1.
In the API, the two byte command is used in the command field, and parameters are populated as binary values in
the parameter field.
Example 1: Forming a network with security (pre-configured link keys)
1. Start a coordinator with the following settings:
a. ID=2234 (arbitrarily selected)
b. EE=1
c. NK=0
d. KY=4455
e. WR (save networking parameters to preserve them through power cycle)
2. Configure one or more routers or end devices with the following settings:
a. ID=2234
b. EE=1
c. KY=4455
d. WR (save networking parameters to preserve them through power cycle)
3. Read the AI setting on the coordinator and joining devices until they return 0 (formed or joined a
network).
In this example, EE, ID, and KY are set the same on all devices. After successfully joining the secure network, all
application data transmissions will be encrypted by the network key. Since NK was set to 0 on the coordinator, a
random network key was selected. And since the link key (KY) was configured the same on all devices, to a nonzero value, the network key was sent encrypted by the pre-configured link key (KY) when the devices joined.
Example 1: Forming a network with security (obtaining keys during joining)
1. Start a coordinator with the following settings:
e. WR (save networking parameters to preserve them through power cycle)
2. Configure one or more routers or end devices with the following settings:
a. ID=2235
b. EE=1
c. KY=0
d. WR (save networking parameters to preserve them through power cycle)
3. Read the AI setting on the coordinator and joining devices until they return 0 (formed or joined a
network).
In this example, EE, ID, and KY are set the same on all devices. Since NK was set to 0 on the coordinator, a random
network key was selected. And since KY was set to 0 on all devices, the network key was sent unencrypted ("in the
clear") when the devices joined. This approach introduces security vulnerability into the network and is not
Network commissioning is the process whereby devices in a mesh network are discovered and configured for
operation. The XBee modules include several features to support device discovery and configuration. In addition to
configuring devices, a strategy must be developed to place devices to ensure reliable routes.
To accommodate these requirements, the XBee modules include various features to aid in device placement,
configuration, and network diagnostics.
Device Configuration
XBee/XBee-PRO ZB modules can be configured locally through serial commands (AT or API), or remotely through
remote API commands. API devices can send configuration commands to set or read the configuration settings of
any device in the network.
Device Placement
For a mesh network installation to be successful, the installer must be able to determine where to place individual
XBee devices to establish reliable links throughout the mesh network.
Link Testing
A good way to measure the performance of a mesh network is to send unicast data through the network from one
device to another to determine the success rate of many transmissions. To simplify link testing, the modules
support a loopback cluster ID (0x12) on the data endpoint (0xE8). Any data sent to this cluster ID on the data
endpoint will be transmitted back to the sender. This is shown in the figure below:
The configuration steps to send data to the loopback cluster ID depend on the firmware type.
AT Firmware
To send data to the loopback cluster ID on the data endpoint of a remote device, set the CI command value to
0x12. The SE and DE commands should be set to 0xE8 (default value). The DH and DL commands should be set to
the address of the remote (0 for the coordinator, or the 64-bit address of the remote). After exiting command
mode, any received serial characters will be transmitted to the remote device, and returned to the sender.
Send an Explicit Addressing ZigBee Command API frame (0x11) using 0x12 as the cluster ID and 0xE8 as the source
and destination endpoint. Data packets received by the remote will be echoed back to the sender.
RSSI Indicators
It is possible to measure the received signal strength on a device using the DB command. DB returns the RSSI value
(measured in –dBm) of the last received packet. However, this number can be misleading. The DB value only
indicates the received signal strength of the last hop. If a transmission spans multiple hops, the DB value provides
no indication of the overall transmission path, or the quality of the worst link – it only indicates the quality of the
last link and should be used sparingly.
The DB value can be determined in hardware using the RSSI/PWM module pin (pin 6). If the RSSI PWM
functionality is enabled (P0 command), when the module receives data, the RSSI PWM is set to a value based on
the RSSI of the received packet. (Again, this value only indicates the quality of the last hop.) This pin could
potentially be connected to an LED to indicate if the link is stable or not.
Device Discovery
Network Discovery
The network discovery command can be used to discover all Digi modules that have joined a network. Issuing the
ND command sends a broadcast node discovery command throughout the network. All devices that receive the
command will send a response that includes the device’s addressing information, node identifier string (see NI
command), and other relevant information. This command is useful for generating a list of all module addresses in
a network.
When a device receives the node discovery command, it waits a random time before sending its own response.
The maximum time delay is set on the ND sender with the NT command. The ND originator includes its NT setting
in the transmission to provide a delay window for all devices in the network. Large networks may need to increase
NT to improve network discovery reliability. The default NT value is 0x3C (6 seconds).
ZDO Discovery
The ZigBee Device Profile includes provisions to discover devices in a network that are supported on all ZigBee
devices (including non-Digi products). These include the LQI Request (cluster ID 0x0031) and the Network Update
Request (cluster ID 0x0038). The LQI Request can be used to read the devices in the neighbor table of a remote
device, and the Network Update Request can be used to have a remote device do an active scan to discover all
nearby ZigBee devices. Both of these ZDO commands can be sent using the XBee Explicit API transmit frame (0x11).
See the API chapter for details. Refer to the ZigBee specification for formatting details of these two ZDO frames.
Joining Announce
All ZigBee devices send a ZDO Device Announce broadcast transmission when they join a ZigBee network (ZDO
cluster ID 0x0013). These frames will be sent out the XBee's UART as an Explicit Rx Indicator API frame (0x91) if AO
is set to 1. The device announce payload includes the following information:
The 16-bit and 64-bit addresses are received in little-endian byte order (LSB first). See the ZigBee specification for
details.
Commissioning Pushbutton and Associate LED
The XBee modules support a set of commissioning and LED behaviors to aid in device deployment and
commissioning. These include the commissioning pushbutton definitions and associate LED behaviors. These
features can be supported in hardware as shown below.
Commissioning Pushbutton and Associate LED Functionalities
The commissioning pushbutton definitions provide a variety of simple functions to aid in deploying devices in a
network. The commissioning button functionality on pin 20 is enabled by setting the D0 command to 1 (enabled by
default).
Button
Presses
1
2
4
If modules is joined to a network
• Sends a node identification broadcast
transmission
• Sends a broadcast transmission to
enable joining on the coordinator and all
devices in the network for 1 minute. (If
joining is permanently enabled on a
device (NJ = 0xFF), this action has no
effect on that device.)
• Issues ATRE to restore module
parameters to default values, including
ID and SC.
• The device attempts to join a network
based on its ID and SC settings.
If module is not joined to a
network
• Wakes an end device for 30
seconds
• Blinks a numeric error code on
the Associate pin indicating the
cause of join failure (see section
6.4.2).
• N/A
• Issues ATRE to restore module
parameters to default values,
including ID and SC.
• The device attempts to join a
network based on its ID and SC
settings.
Button presses may be simulated in software using the ATCB command. ATCB should be issued with a parameter
set to the number of button presses to execute. (e.g. sending ATCB1 will execute the action(s) associated with a
single button press.)
The node identification frame is similar to the node discovery response frame – it contains the device’s address,
node identifier string (NI command), and other relevant data. All API devices that receive the node identification
frame send it out their UART as an API Node Identification Indicator frame (0x95).
The Associate pin (pin 15) can provide indication of the device’s network status and diagnostics information. To
take advantage of these indications, an LED can be connected to the Associate pin as shown in the figure above.
The Associate LED functionality is enabled by setting the D5 command to 1 (enabled by default). If enabled, the
Associate pin is configured as an output and will behave as described in the following sections.
Joined Indication
The Associate pin indicates the network status of a device. If the module is not joined to a network, the Associate
pin is set high. Once the module successfully joins a network, the Associate pin blinks at a regular time interval.
This is shown in the following figure.
The LT command defines the blink time of the Associate pin. If set to 0, the device uses the default blink time
(500ms for coordinator, 250ms for routers and end devices).
Diagnostics Support
The Associate pin works with the commissioning pushbutton to provide additional diagnostics behaviors to aid in
deploying and testing a network. If the commissioning push button is pressed once, and the device has not joined a
network, the Associate pin blinks a numeric error code to indicate the cause of join failure. The number of blinks is
equal to (AI value – 0x20). For example, if AI=0x22, 2 blinks occur.
If the commissioning push button is pressed once, and the device has joined a network, the device transmits a
broadcast node identification packet. If the Associate LED functionality is enabled (D5 command), a device that
receives this transmission will blink its Associate pin rapidly for 1 second.
The following figures demonstrate these behaviors.
ZigBee end devices are intended to be battery-powered devices capable of sleeping for extended periods of time.
Since end devices may not be awake to receive RF data at a given time, routers and coordinators are equipped
with additional capabilities (including packet buffering and extended transmission timeouts) to ensure reliable
data delivery to end devices.
End Device Operation
When an end device joins a ZigBee network, it must find a router or coordinator device that is allowing end devices
to join. Once the end device joins a network, a parent-child relationship is formed between the end device and the
router or coordinator that allowed it to join. See chapter 3 for details.
When the end device is awake, it sends poll request messages to its parent. When the parent receives a poll
request, it checks a packet queue to see if it has any buffered messages for the end device. It then sends a MAC
layer acknowledgment back to the end device that indicates if it has data to send to the end device or not.
If the end device receives the acknowledgment and finds that the parent has no data for it, the end device can
return to idle mode or sleep. Otherwise, it will remain awake to receive the data. This polling mechanism allows
the end device to enter idle mode and turn its receiver off when RF data is not expected in order to reduce current
consumption and conserve battery life.
The end device can only send data directly to its parent. If an end device must send a broadcast or a unicast
transmission to other devices in the network, it sends the message directly to its parent and the parent performs
any necessary route or address discoveries to route the packet to the final destination.
Parent Operation
Each router or coordinator maintains a child table that contains the addresses of its end device children. A router
or coordinator that has unused entries in its child table is said to have end device capacity, or the ability to allow
new end devices to join. If the child table is completely filled (such that the number of its end device children
matches the number of child table entries), the device cannot allow any more end devices to join to it.
Since the end device children are not guaranteed to be awake at a given time, the parent is responsible for
managing incoming data packets in behalf of its end device children. If a parent receives an RF data transmission
destined for one of its end device children, and if the parent has enough unused buffer space, it will buffer the
packet. The data packet will remain buffered until a timeout expires, or until the end device sends a poll request to
retrieve the data.
The parent can buffer one broadcast transmission for all of its end device children. When a broadcast transmission
is received and buffered, the parent sets a flag in its child table when each child polls and retrieves the packet.
Once all children have received the broadcast packet, the buffered broadcast packet is discarded. If all children
have not received a buffered broadcast packet and a new broadcast is received, the old broadcast packet is
discarded, the child table flags are cleared, and the new broadcast packet is buffered for the end device children.
This is demonstrated in the figure below.
When an end device sends data to its parent that is destined for a remote device in the network, the parent
buffers the data packet until it can establish a route to the destination. The parent may perform a route or 16-bit
address discovery in behalf of its end device children. Once a route is established, the parent sends the data
transmission to the remote device.
End Device Poll Timeouts
To better support mobile end devices (end devices that can move around in a network); parent router and
coordinator devices have a poll timeout for each end device child. If an end device does not send a poll request to
its parent within the poll timeout, the parent will remove the end device from its child table. This allows the child
table on a router or coordinator to better accommodate mobile end devices in the network.
Packet Buffer Usage
Packet buffer usage on a router or coordinator varies depending on the application. The following activities can
require use of packet buffers for up to several seconds:
•Route and address discoveries
•Application broadcast transmissions
•Stack broadcasts (e.g. ZDO "Device Announce" messages when devices join a network)
•Unicast transmissions (buffered until acknowledgment is received from destination or retries exhausted)
•Unicast messages waiting for end device to wake.
Applications that use regular broadcasting or that require regular address or route discoveries will use up a
significant number of buffers, reducing the buffer availability for managing packets for end device children.
Applications should reduce the number of required application broadcasts, and consider implementing an external
address table or many-to-one and source routing if necessary to improve routing efficiency.
Non-Parent Device Operation
Devices in the ZigBee network treat data transmissions to end devices differently than transmissions to other
routers and coordinators. Recall that when a unicast transmission is sent, if a network acknowledgment is not
received within a timeout, the device resends the transmission. When transmitting data to remote coordinator or
router devices, the transmission timeout is relatively short since these devices are powered and responsive.
However, since end devices may sleep for some time, unicast transmissions to end devices use an extended
timeout mechanism in order to allow enough time for the end device to wake and receive the data transmission
from its parent.
If a non-parent device does not know the destination is an end device, it will use the standard unicast timeout for
the transmission. However, provisions exist in the Ember ZigBee stack for the parent to inform the message sender
that the destination is an end device. Once the sender discovers the destination device is an end device, future
transmissions will use the extended timeout. See the XBee Router / Coordinator Configuration section in this
chapter for details.
XBee End Device Configuration
XBee end devices support two different sleep modes:
•Pin Sleep
•Cyclic Sleep.
Pin sleep allows an external microcontroller to determine when the XBee should sleep and when it should wake by
controlling the Sleep_RQ pin. In contrast, cyclic sleep allows the sleep period and wake times to be configured
through the use of AT commands. The sleep mode is configurable with the SM command.
In both pin and cyclic sleep modes, XBee end devices poll their parent every 100ms while they are awake to
retrieve buffered data. When a poll request has been sent, the end device enables the receiver until an
acknowledgment is received from the parent. (It generally takes less than 10ms from the time the poll request is
sent until the acknowledgment is received.) The acknowledgment indicates if the parent has buffered data for the
end device child or not. If the acknowledgment indicates the parent has pending data, the end device will leave the
receiver on to receive the data. Otherwise, the end device will turn off the receiver and enter idle mode (until the
next poll request is sent) to reduce current consumption (and improve battery life).
Once the module enters sleep mode, the On/Sleep pin (pin 26) is de-asserted (low) to indicate the module is
entering sleep mode. If CTS hardware flow control is enabled (D7 command), the CTS pin (pin 25) is de-asserted
(high) when entering sleep to indicate that serial data should not be sent to the module. The module will not
respond to serial or RF data when it is sleeping. Applications that must communicate serially to sleeping end
devices are encouraged to observe CTS flow control.
When the XBee wakes from sleep, the On/Sleep pin is asserted (high), and if flow control is enabled, the CTS pin is
also asserted (low). If the module has not joined a network, it will scan all SC channels after waking to try and find
a valid network to join.
Pin Sleep
Pin sleep allows the module to sleep and wake according to the state of the Sleep_RQ pin (pin 9). Pin sleep mode is
enabled by setting the SM command to 1.
When Sleep_RQ is asserted (high), the module will finish any transmit or receive operations and enter a low power
state. For example, if the module has not joined a network and Sleep_RQ is asserted (high), the module will sleep
once the current join attempt completes (i.e. when scanning for a valid network completes). The module will wake
from pin sleep when the Sleep_RQ pin is de-asserted (low).
In the figure above, t1, t2, and t3 represent the following events:
•T1 - Time when Sleep_RQ is asserted (high)
•T2 - Time when the XBee enters sleep (CTS state change only if hardware flow control is enabled)
•T3 - Time when Sleep_RQ is de-asserted (low) and the module wakes.
The time between T1 and T2 varies depending on the state of the module. In the worst case scenario, if the end
device is trying to join a network, or if it is waiting for an acknowledgment from a data transmission, the delay
could be up to a few seconds.
When the XBee is awake and is joined to a network, it sends a poll request to its parent to see if the parent has any
buffered data for it. The end device will continue to send poll requests every 100ms while it is awake.
Parent and remote devices must be configured to buffer data correctly and to utilize adequate transmission
timeouts. See the XBee Router / Coordinator Configuration section in this chapter for details.
Cyclic Sleep
Cyclic sleep allows the module to sleep for a specified time and wake for a short time to poll its parent for any
buffered data messages before returning to sleep again. Cyclic sleep mode is enabled by setting the SM command
to 4 or 5. SM5 is a slight variation of SM4 that allows the module to be woken prematurely by asserting the
Sleep_RQ pin (pin 10). In SM5, the XBee can wake after the sleep period expires, or if a high-to-low transition
occurs on the Sleep_RQ pin. Setting SM to 4 disables the pin wake option.
In cyclic sleep, the module sleeps for a specified time, and then wakes and sends a poll request to its parent to
discover if the parent has any pending data for the end device. If the parent has buffered data for the end device,
or if serial data is received, the XBee will remain awake for a time. Otherwise, it will enter sleep mode
immediately.
The On/Sleep line is asserted (high) when the module wakes, and is de-asserted (low) when the module sleeps. If
hardware flow control is enabled (D7 command), the CTS pin will assert (low) when the module wakes and can
receive serial data, and de-assert (high) when the module sleeps.
In the figure above, t1, t2, and t3 represent the following events:
•T1 - Time when the module wakes from cyclic sleep
•T2 - Time when the module returns to sleep
•T3 - Later time when the module wakes from cyclic sleep.
The wake time and sleep time are configurable with software commands as described in the sections below.
Wake Time (Until Sleep)
In cyclic sleep mode (SM=4 or 5), if serial or RF data is received, the module will start a sleep timer (time until
sleep). Any data received serially or over the RF link will restart the timer. The sleep timer value is settable with the
ST command. While the module is awake, it will send poll request transmissions every 100ms to check its parent
for buffered data messages. The module returns to sleep when the sleep timer expires, or if the SI command is
sent to it. The following image shows this behavior.
The sleep period is configured based on the SP, SN, and SO commands. The following table lists the behavior of
these commands.
Command Range
SP 0x20-oxAF0 (x10ms)
SN 1-0xFFFF
SD 0-0xFF
Description
module.
Configures the number of sleep periods
multiplier.
Defines options for sleep mode behavior.
0x02 - Always wake
0X04-Enable extended sleep (sleep for full
SP*SN) time)
The XBee module supports both a short cyclic sleep and an extended cyclic sleep that make use of these
commands. These two modes allow the sleep period to be configured according to the application requirements.
Short Cyclic Sleep
In short cyclic sleep mode, the sleep behavior of the module is defined by the SP and SN commands, and the SO
command must be set to 0x00 (default) or 0x02. In short cyclic sleep mode, the SP command defines the sleep
period and is settable up to 28 seconds. When the XBee enters short cyclic sleep, it remains in a low power state
until the SP time has expired.
After the sleep period expires, the XBee sends a poll request transmission to its parent to determine if its parent
has any buffered data waiting for the end device. Since router and coordinator devices can buffer data for end
device children up to 30 seconds, the SP range (up to 28 seconds) allows the end device to poll regularly enough to
receive buffered data. If the parent has data for the end device, the end device will start its sleep timer (ST) and
continue polling every 100ms to receive data. If the end device wakes and finds that its parent has no data for it,
the end device can return to sleep immediately.
The SN command can be used to control when the On/Sleep line is asserted (high). If SN is set to 1 (default), the
On/Sleep line will be set high each time the XBee wakes from sleep. Otherwise, if SN is greater than 1, the
On/Sleep line will only be set high if RF data is received, or after SN wake cycles occur. This allows an external
device to remain powered off until RF data is received, or until a number of sleep periods have expired (SN sleep
periods). This mechanism allows the XBee to wake at regular intervals to poll its parent for data without waking an
external device for an extended time (SP * SN time). This is shown in the figure below.
Note: SP controls the packet buffer time on routers and coordinators. SP should be set on all router and
coordinator devices to match the longest end device SP time. See the XBee Router / Coordinator Configuration
section for details.
Extended Cyclic Sleep
In extended cyclic sleep operation, an end device can sleep for a multiple of SP time which can extend the sleep
time up to several days. The sleep period is configured using the SP and SN commands. The total sleep period is
equal to (SP * SN) where SP is measured in 10ms units. The SO command must be set correctly to enable extended
sleep.
Since routers and coordinators can only buffer incoming RF data for their end device children for up to 30 seconds,
if an end device sleeps longer than 30 seconds, devices in the network need some indication when an end device is
awake before they can send data to it. End devices that use extended cyclic sleep should send a transmission (such
as an IO sample) when they wake to inform other devices that they are awake and can receive data. It is
recommended that extended sleep end devices set SO to wake for the full ST time in order to provide other
devices with enough time to send messages to the end device.
Similar to short cyclic sleep, end devices running in this mode will return to sleep when the sleep timer expires, or
when the SI command is received.
Transmitting RF Data
An end device may transmit data when it wakes from sleep and has joined a network. End devices transmit directly
to their parent and then wait for an acknowledgment to be received. The parent will perform any required address
and route discoveries to help ensure the packet reaches the intended destination before reporting the
transmission status to the end device.
Receiving RF Data
After waking from sleep, an end device sends a poll request to its parent to determine if the parent has any
buffered data for it. In pin sleep mode, the end device polls every 100ms while the Sleep_RQ pin is de-asserted
(low). In cyclic sleep mode, the end device will only poll once before returning to sleep unless the sleep timer (ST)
is started (serial or RF data is received). If the sleep timer is started, the end device will continue to poll every
100ms until the sleep timer expires.