UG103.

2: Zigbee Fundamentals

This document describes the key features and characteristics of

a Zigbee solution. It also includes a section on Zigbee 3.0. KEY POINTS

Silicon Labs’ Fundamentals series covers topics that project managers, application de- • Introduces Zigbee and key features—
signers, and developers should understand before beginning to work on an embedded mesh networking, network node types,
networking solution using Silicon Labs chips, networking stacks such as EmberZNet routing concepts, Zigbee stack, Zigbee
Cluster Library, and Zigbee compliance
PRO or Silicon Labs Bluetooth®, and associated development tools. The documents
can be used as a starting place for anyone needing an introduction to developing wire- • Includes a section on Zigbee 3.0.
less networking applications, or who is new to the Silicon Labs development environ-
ment.

silabs.com | Building a more connected world. Rev. 1.4

 UG103.2: Zigbee Fundamentals
Introduction

1. Introduction

Zigbee refers both to:

• An open standard for reliable, cost-effective, low power, wireless device-to-device communication.
• An alliance of over 400 companies who together are defining and using the standard to communicate in a variety of applications
such as smart energy and commercial building automation.

The Zigbee Alliance is led by a set of promoter member companies that make up the board of directors. Silicon Labs is one of the
companies on the board of directors. Zigbee also has a large number of participant and adopter members that work within the technical
and marketing workgroups to create and maintain the standards. The Alliance activities are accomplished through workgroups dedica-
ted to specific areas of the technology. These include a network group, a security group, an application profile group, and several oth-
ers. All Zigbee documents are available on the Zigbee website www.zigbee.org. Zigbee Alliance membership is required to access cur-
rent specification documents. To use the Zigbee technology within a product, companies are required to become members of the Alli-
ance.

Silicon Labs is active in a variety of areas within the Alliance and with helping customers adopt Zigbee technology through its Ember®
Zigbee solutions. The Silicon Labs (originally Ember) Zigbee platform has been a Golden Unit for testing and certification for all new
revisions of the standard to date. Ember Corporation, which was acquired by Silicon Labs in July 2012, was a founding member of the
Zigbee Alliance.

The Zigbee Alliance specifies three major items of interest.

1. Zigbee networking – The basic levels of network interaction, including acceptable RF behavior, methods of creating and joining the
network, discovering routes, and using routes to transmit traffic over the network. The Zigbee Alliance offers other networking specifica-
tions such GreenPower, but the focus of this document is Zigbee PRO. See UG103.15: Silicon Labs Green Power Fundamentals for
more information about the Silicon Labs implementation of Green Power.

2. Zigbee application layer – Describes a set of messages and network settings for a Zigbee application. All devices adhering to these
settings may interoperate. Before Zigbee 3.0, individual application profiles, either public or private, had their own certification program.
With Zigbee 3.0, one common application layer certification program applies to every Zigbee product. The following figure displays
some of the original separate Zigbee application profiles.

Figure 1.1. Zigbee Application Profiles

3. Certification – Zigbee applications are certified by the Alliance to indicate to end users that the device(s) are compliant. Applications
using private profiles receive certification that they are using a properly operating Zigbee stack and will not adversely affect other Zig-
bee networks during operation. For more information, see section 7. Zigbee Compliance.

Because Zigbee is committed to open and interoperable devices, standards have been either adopted where they existed, or devel-
oped, from the physical layer through the application layer, as shown in the following figure. At the physical and MAC (medium access
control) layer, Zigbee adopted the IEEE 802.15.4 standard. The networking, security and application layers have all been developed by
the Zigbee Alliance. An ecosystem of supporting systems such as gateways and commissioning tools has also been developed to sim-
plify the development and deployment of Zigbee networks. Extensions and additions to the standards continue to be developed, and
Silicon Labs is committed to supporting these as they are available.

silabs.com | Building a more connected world. Rev. 1.4 | 2

 UG103.2: Zigbee Fundamentals
Introduction

Figure 1.2. Zigbee Architecture

The following three sections describe the general characteristics of a Zigbee network, discuss the use of IEEE 802.15.4 standard, and
summarize the hardware and software elements of a Zigbee network.

1.1 General Characteristics

Zigbee is intended as a cost-effective and low power solution. It is targeted to several markets including home automation, building
automation, sensor networks, smart energy, and personal health care monitoring. The general characteristics for a Zigbee network are
as follows:
• Low power – Devices can typically operate for several years on AA type batteries using suitable duty cycles. With extremely careful
design and special battery technologies, some Zigbee devices such as gas meters can achieve 20 years of battery life.
• Low data rate – The 2.4 GHz band supports a radio data rate of 250 kbps. Actual sustainable traffic through the network is lower
than this theoretical radio capacity. As such, Zigbee is better used for sampling and monitoring applications or basic control applica-
tions. See AN1138: Zigbee Mesh Network Performance and AN1142: Mesh Network Performance Comparison
• Small and large networks – Zigbee networks vary from several devices to thousands of devices communicating seamlessly. The
networking layer is designed with several different data transfer mechanisms (types of routing) to optimize the network operation
based on the expected use.
• Range – Typical devices provide sufficient range to cover a normal home. Readily available designs with power amplifiers extend
the range substantially. A distributed spread spectrum is used at the physical layer to be more immune to interference.
• Simple network installation, start up and operation – The Zigbee standard supports several network topologies. The simple protocols
for forming and joining networks allow systems to self-configure and fix routing problems as they occur.

1.2 IEEE 802.15.4

Zigbee networks are based on the IEEE 802.15.4 MAC and physical layer. The 802.15.4 standard operates at 250 kbps in the 2.4 GHz
band and 40 kbps/20 kbps in the 900/868 MHz bands. A number of chip companies provide solutions in the 2.4 GHz band with a small-
er number supporting the 900/868 MHz band. Zigbee PRO uses 802.15.4-2003.

The 802.15.4 MAC layer is used for basic message handling and congestion control. This MAC layer includes mechanisms for forming
and joining a network, a CSMA mechanism for devices to listen for a clear channel, as well as a link layer to handle retries and ac-
knowledgment of messages for reliable communications between adjacent devices. The Zigbee network layer builds on these underly-
ing mechanisms to provide reliable end-to-end communications in the network. The 802.15.4 standard is available from www.ieee.org.

The 802.15.4 standard provides some options within the MAC layer that are not used by Zigbee in any current stack profiles. An exam-
ple of such an unused feature is guaranteed time slots (GTS), which would be employed by a network to synchronize radio activity
across devices. As such, these items are not normally included in the Zigbee software stack to save code space. Zigbee also has made
specific changes to the 802.15.4 MAC that are documented in Annex D of the Zigbee specification.

silabs.com | Building a more connected world. Rev. 1.4 | 3

 UG103.2: Zigbee Fundamentals
Introduction

1.3 Hardware and Software Elements

A Zigbee solution requires implementation of a Zigbee radio and associated microprocessor (together in a single chip or separately),
and implementation of an application on top of a Zigbee stack. EmberZNet PRO is the Silicon Labs implementation of the Zigbee PRO
stack.

Typically a developer can purchase a Zigbee radio and software as a bundled package, although some third party software stacks have
been developed. In general, the hardware and software provider includes reference designs for the hardware and sample applications
for the software. Based on these, hardware developers can customize the hardware to their specific needs. Alternatively, a number of
module providers can deliver compact and low cost custom modules.

Because of the embedded nature of typical Zigbee applications, software application development is typically interrelated with the hard-
ware design to provide an optimal solution. Silicon Labs provides both a standard networking API based on the Zigbee specification
and an application framework, which gives customers a way to rapidly develop their applications based on Zigbee application profiles
and the Zigbee cluster library (ZCL), as described below. Alternatively, a number of third party software development firms specialize in
developing Zigbee applications and can assist in new product development.

1.3.1 About Application Frameworks

Any application may be built from scratch, but this is a slow and sometimes tedious process. The alternative is to take working code
and modify it to meet the requirements of your application. Adapting a working design is an easier and more efficient approach to build-
ing an application, especially your first application using a new technology. To assist with this, Silicon Labs supplies not only example
application code but also several application frameworks as part of their development platforms. Application frameworks contain a body
of Silicon Labs-supplied code stored in libraries and plugins. Silicon Labs also provides Simplicity Studio, which includes an Integrated
Development Environment (IDE) based on Application Builder (AppBuilder). AppBuilder is an interactive GUI tool that allows you to
configure the libraries and plugins contained in the application framework to implement devices, such as a Zigbee Smart Energy meter,
defined by the Zigbee application layer,such as Zigbee 3.0.

The main Zigbee Application Framework is based around the ZCL (Zigbee Cluster Library) for use when developing Zigbee PRO appli-
cations. It provides a set of embedded code to implement ZCL-based cluster handling, required network tasks, and typical Zigbee appli-
cation-state machines and includes support for the Zigbee 3.0 application layer, as well as a number of legacy Zigbee profiles (such as
Home Automation (HA), Smart Energy (SE), and Zigbee Light Link (ZLL)). This code, developed according to best practices recommen-
ded by Silicon Labs and by the specifications for different Zigbee application profiles, acts as a sort of “software reference design” that
can be run on either a System-On-Chip (such as the EFR32) or on a host processor connected to a Network Control Processor based
on the EZSP (EmberZNet Serial Protocol). See UG391:Zigbee Application Framework Developer’s Guide for details. Other Zigbee ap-
plication frameworks provided in the EmberZNet SDK include Green Power and Customizable Network Coprocessor (NCP).

Application frameworks provide common basic services and, when combined with device-specific code, provide you with most of the
over-the-air behavior for your device. Your task then becomes developing hardware-specific functionality and interfacing it to the Silicon
Labs software. Some common usage case examples based on the relevant application framework are provided with each Silicon Labs
stack release.

Using an application framework has the following benefits:

• Helps maintain a common approach, and adherence to the protocol standard and Silicon Labs best practices.
• Helps ensure protocol compliance, since Silicon Labs runs the application framework through pre-compliance testing.
• Allows easy merge of updates, because Silicon Labs maintains the core components.
• Offers the ability to include complex tasks in your application, without the need to manage those tasks at the code level.
• Provides configurability while maintaining protocol compliance and the ability to easily upgrade to future stack releases.
• Provides a callback/command interface and set of configuration points that are farther removed from the stack than usual, and there-
fore are more in tune with application-layer behaviors than stack-level behaviors.
• Provides a set of plugins to illustrate “best practice” implementations of higher-level application logic, such as debug functionality
and CLI functionality, ZCL cluster implementations, application state machines, or complex functionality.

All of the above help to achieve faster time to market and more robust solutions.

1.3.2 Other Tools and Utilities

As well as Simplicity Studio and AppBuilder, Silicon Labs also offers a number of utilities and tools that are available to developers
during the application process including:
• Over-the-air (OTA) bootloaders to allow upgrades to system software after deployment
• Gateway interfaces to interface the Zigbee network to other systems
• Programming tools for the microprocessor

silabs.com | Building a more connected world. Rev. 1.4 | 4

 UG103.2: Zigbee Fundamentals
Zigbee Mesh Networking

2. Zigbee Mesh Networking

Embedded mesh networks make radio systems more reliable by allowing radios to relay messages for other radios. For example, if a
node cannot send a message directly to another node, the embedded mesh network relays the message through one or more interme-
diary nodes.

EmberZNet PRO supports three types of mesh network topologies, as shown in the following figure:
• Star Network
• Full Mesh Network
• Hybrid Mesh Network

Figure 2.1. Star, Full Mesh, and Hybrid Mesh Networks

Note: Blue devices in the above figure are end devices and can be sleepy or mobile. See section 3. Network Node Types, for more
information about end devices.

2.1 Star Network

In a star network, one hub is the central point of all communications. The hub can become bottlenecked with network/processing band-
width. This topology is not very mesh-like, and transmission is limited by the hub’s communication radius. Outlying nodes can be bat-
tery powered. In the EmberZNet PRO stack, this topology is formed by a group of end devices with a coordinator node as their parent.
The coordinator node serves as network hub.

2.2 Full Mesh Network

In a full mesh network, all nodes are router nodes, including the coordinator after it forms the network. Because all nodes can relay
information for all other nodes, this topology is least vulnerable to link failure; it is highly unlikely that one device might act as a single
point of failure for the entire network.

2.3 Hybrid Mesh Network

A hybrid mesh network topology combines star and mesh strategies. Several star networks exist, but their hubs can communicate as a
mesh network. A hybrid network allows for longer distance communication than a star topology and more capability for hierarchical de-
sign than a mesh topology. This topology is formed by the EmberZNet PRO stack, using router devices as hubs for the star subnet-
works, where each hub can have end devices attached to it.

The choice of topology must take into consideration which nodes are line-powered or battery-powered, expected battery lifetime,
amount of network traffic required, latency requirements, the cost of the solution, as well as other factors. See UG103.3: Software De-
sign Fundamentals, for more information.

silabs.com | Building a more connected world. Rev. 1.4 | 5

 UG103.2: Zigbee Fundamentals
Network Node Types

3. Network Node Types

The Zigbee specification supports at most one coordinator, multiple routers, and multiple end devices within a single network. These
node types are described in the following sections.

3.1 Coordinator

A Zigbee coordinator (ZC) is responsible for forming a centralized network. A coordinator is a router with some additional functionality
and uses a hardcoded network address of 0x0000. Zigbee coordinator functions include selecting an appropriate channel after scan-
ning available channels, and selecting an extended PAN ID (see section 5.4 Extended PAN IDs for more information about extended
PAN IDs). After forming the network, the coordinator acts as a router.

The coordinator also has added responsibilities, such as being the trust center or network manager. The trust center manages the se-
curity settings and authorizations for the network. The network manager monitors and corrects network issues such as PAN ID conflicts
or channel changing due to interference. These choices are up to the application developer, and in some cases are made by an appli-
cation layer, such as Zigbee 3.0, being used.
Note: Only a network coordinator can be designated as a centralized trust center for security purposes when starting a network.

3.2 Routers

Zigbee router devices provide routing services to network devices. Routers can also serve as end devices. Unlike end devices, routers
are not designed to sleep and should generally remain on as long as a network is established.

3.3 End Devices

Zigbee end devices (ZED) are leaf nodes. They communicate only through their parent nodes and, unlike router devices, cannot relay
messages intended for other nodes.

Depending on the network stack, end devices can be of several types:

• Sleepy end devices (EmberNodeType EMBER_SLEEPY_END_DEVICE) power down their radio when idle, and thus conserve re-
sources. However, they must poll their parent node to receive incoming messages and acknowledgments; no data is sent to the
sleepy end device until the end device requests it. Sleepy end devices are also sometimes known as rx-off-when-idle devices. This
is a standard Zigbee device type.
• Non-sleepy end devices (EmberNodeType EMBER_END_DEVICE) do not route messages for other devices but they remain pow-
ered during operation. These devices are known as Rx-on-when-idle devices. This is a standard Zigbee device type.

The EmberZNet PRO stack supports sleepy and non-sleepy end devices. The choice of node type must also be considered carefully.
For example, in a very dense network, it is not always advantageous for all line powered nodes to be routers due to possible interfer-
ence issues which may occur when a child node tries to find a parent node to communicate with. It is important to try to create a bal-
anced network where all nodes have redundant paths, but without too many routers in close proximity to create interference.

silabs.com | Building a more connected world. Rev. 1.4 | 6

 UG103.2: Zigbee Fundamentals
Zigbee Routing Concepts

4. Zigbee Routing Concepts

4.1 Overview

Zigbee has several routing mechanisms that can be used based on the network and expected traffic patterns. In Zigbee Specification,
document 05-3474, section 3.4 describes the frame format for command frames, and section 3.6.3 describes routing behavior.

The application designer should choose which mechanism to use as part of the system architecture and design. In actual practice one
application may use several of these routing mechanisms, because some devices may be performing one-to-one communications while
other devices may be communicating to a central monitoring device. The types of routing discussed below are:
• Table Routing
• Broadcast Routing
• Multicast Routing
• Many-to-One/Source Routing

4.1.1 Table Routing

Routes are formed when one node sends a route request to discover the path to another node. After a route is discovered between the
two nodes, the source node sends its message to the first node in the route, as specified in the source node's routing table. Each inter-
mediate node uses its own routing table to forward the message to the next node (that is "hop") along the route until the message
reaches its destination. If a route fails, a route error is sent back to the originator of the message who can then rediscover the route.

4.1.2 Broadcast Routing

Broadcast routing is a mechanism to send a message to all devices in a network. Network-level broadcast options exist to send to rout-
ers only, to all non-sleeping nodes including end devices, or also to send to sleeping end devices. A broadcast message is repeated by
all router-capable devices in the network three times to ensure delivery to all devices. While a broadcast is a reliable means of sending
a message, it should be used sparingly because of the impact on network performance. Repeated broadcasts can limit any other traffic
that may be occurring in the network. Broadcasts are also not a reliable means of delivery to a sleeping device because the parent
device is responsible for buffering the message for the sleeping child but may drop the message before the end device wakes to re-
ceive it.

4.1.3 Multicast Routing

Multicast routing provides a one-to-many routing option. A multicast is used when one device wants to send a message to a group of
devices, such as a light switch sending an on command to a bank of 10 lights. Under this mechanism, all the devices are joined into a
multicast group. Only those devices that are members of the group will receive messages, although other devices will route these multi-
cast messages. A multicast is a filtered limited broadcast. It should be used only as necessary in applications, because over-use of
broadcast mechanisms can degrade network performance. A multicast message is never acknowledged.

4.1.4 Many-to-One/Source Routing

Many-to-one routing is a simple mechanism to allow an entire network to have a path to a central control or monitoring device. Under
normal table routing, the central device and the devices immediately surrounding it would need routing table space to store a next hop
for each device in the network, as well as an entry to the central device itself. Given the memory limited devices often used in Zigbee
networks, these large tables are undesirable.

Under many-to-one routing, the central device, known as a “concentrator,” sends a single route discovery that established a single
route table entry in all routers to provide the next hop to the central device. This yields a result similar to that of table routing, but with a
single many-to-one route request rather than many individual, one-to-one route requests from each router towards the concentrator.

All devices in the network then have a next hop path to the concentrator and only a single table entry is used. However, often the cen-
tral device also needs to send messages back out into the network. This would result in a more significant increase in route table size,
particularly for those nodes closest to the concentrator, since they are relay points in the concentrator’s many outbound routes to the
rest of the network. Instead, incoming messages to the concentrator first use a route record message to store the sequence of hops
used along the route. The concentrator then stores these next hop routes in reverse order as “source routes” in a locally held table
known as a “source route table”. Outgoing messages include this source route in the network header of the message. The message is
then routed using next hops from the network header instead of from the route table. This provides for large scalable networks without
increasing the memory requirements of all devices. It should be noted that the concentrator requires some additional memory if it is
storing these source routes.

For detailed information on message delivery, refer to the Zigbee specification available from www.zigbee.org.

silabs.com | Building a more connected world. Rev. 1.4 | 7

 UG103.2: Zigbee Fundamentals
Zigbee Routing Concepts

4.2 Using Link Quality to Aid in Routing

The information in this section is provided for those wishing to understand the details of the network layer's operation, which can prove
useful during troubleshooting. Otherwise, link status messages are handled automatically by the stack and application writers need not
be concerned with it.

Links in wireless networks often have asymmetrical link quality due to variations in the local noise floor, receiver sensitivity, and transmit
power. The routing layer must use knowledge of the quality of links in both directions in order to establish working routes and to opti-
mize the reliability and efficiency of those routes. It can also use the knowledge to establish reliable two-way routes with a single dis-
covery.

Zigbee routers keep track of inbound link quality in the neighbor table, typically by averaging LQI (Link Quality Indicator) measurements
made by the physical layer. To handle link asymmetry, the Zigbee PRO stack profile specifies that routers obtain and store costs of
outgoing links as measured by their neighbors. This is accomplished by exchanging link status information through periodic one hop
broadcasts, referred to as "link status" messages. The link status algorithm is explained below, as implemented in EmberZNet PRO.

4.2.1 Description of Relevant Neighbor Table Fields

Zigbee routers store information about neighboring Zigbee devices in a neighbor table. For each router neighbor, the entry includes the
following fields:
• average incoming LQI
• outgoing cost
• age

The incoming LQI field is an exponentially weighted moving average of the LQI for all incoming packets from the neighbor. The incom-
ing cost for the neighbor is computed from this value using a lookup table.

The outgoing cost is the incoming cost reported by the neighbor in its neighbor exchange messages. An outgoing cost of 0 means the
cost is unknown. An entry is called "two-way" if it has a nonzero outgoing cost, and "one-way" otherwise.

The age field measures the amount of time since the last neighbor exchange message was received. A new entry starts at age 0. The
age is incremented every EM_NEIGHBOR_AGING_PERIOD, currently 16 seconds. Receiving a neighbor exchange packet resets the age to
EM_MIN_NEIGHBOR_AGE, as long as the age is already at least EM_MIN_NEIGHBOR_AGE (currently defined to be 3). This makes it possible
to recognize nodes that have been recently added to the table and avoid evicting them, which reduces thrashing in a dense network. If
the age is greater than EM_STALE_NEIGHBOR (currently 6), the entry is considered stale and the outgoing cost is reset to 0.

4.2.2 Link Status Messages

Routers send link status messages every 16 seconds plus or minus 2 seconds of jitter. If the router has no two-way links it sends them
eight times faster. The packet is sent as a one-hop broadcast with no retries. In the EmberZNet PRO stack, they are sent as Zigbee
network command frames.

The payload contains a list of short IDs of all non-stale neighbors, along with their incoming and outgoing costs. The incoming cost is
always a value between 1 and 7. The outgoing cost is a value between 0 and 7, with the value 0 indicating an unknown outgoing cost.
For frame format details, refer to the Zigbee specification. Link status messages are also automatically decoded by the Simplicity Studio
Network Analyzer for easy reference.

Upon receipt of a link status message, either a neighbor entry already exists for that neighbor, or one is added if there is space or if the
neighbor selection policy decides to replace an old entry with it. If the entry does not get into the table, the packet is simply dropped. If it
does get in, then the outgoing cost field is updated with the incoming cost to the receiving node as listed in the sender's neighbor ex-
change message. If the receiver is not listed in the message, the outgoing cost field is set to 0. The age field is set to
EM_MIN_NEIGHBOR_AGE.

silabs.com | Building a more connected world. Rev. 1.4 | 8

 UG103.2: Zigbee Fundamentals
Zigbee Routing Concepts

4.2.3 How Two-way Costs are Used by the Network Layer

As mentioned above, the routing algorithm makes use of the bidirectional cost information to avoid creating broken routes, and to opti-
mize the efficiency and robustness of established routes. For the reader familiar with the Zigbee route discovery process, this subsec-
tion gives details of how the outgoing cost is used. The mechanism is surprisingly simple, but provides all the benefits mentioned
above.

Upon receipt of a route request command frame, the neighbor table is searched for an entry corresponding to the transmitting device. If
no such entry is found, or if the outgoing cost field of the entry has a value of 0, the frame is discarded and route request processing is
terminated.

If an entry is found with non-zero outgoing cost, the maximum of the incoming and outgoing costs is used for the purposes of the path
cost calculation, instead of only the incoming cost. This value is also used to increment the path cost field of the route request frame
prior to retransmission.

4.2.4 Key Concept: Rapid Response

Rapid response allows a node that has been powered on or reset to rapidly acquire two-way links with its neighbors, minimizing the
amount of time the application must wait for the stack to be ready to participate in routing. This feature is 100% Zigbee-compatible.

If a link status message is received that contains no two-way links, and the receiver has added the sender to its neighbor table, then the
receiver sends its own link status message immediately in order to get the sender started quickly. The message is still jittered by 2
seconds to avoid collisions with other rapid responders. To avoid a chain reaction, rapid responders must themselves have at least one
two-way link.

4.2.5 Key Concept: Connectivity Management

By nature Zigbee devices are RAM-constrained, but often Zigbee networks are dense. This means that each router is within radio range
of a large number of other routers. In such cases, the number of neighbors can exceed the maximum number of entries in a device's
neighbor table. In such cases, the wrong choice of which neighbors to keep can lead to routing inefficiencies or worse — a disconnec-
ted network. The EmberZNet PRO stack employs 100% Zigbee-compatible algorithms to manage the selection of neighbors in dense
networks to optimize network connectivity.

4.3 Route Discovery & Repair

Routing in Zigbee is automatically handled by the networking layer, and the application developer usually does not need to be con-
cerned with its behavior. However, it is useful to have a feel for how the network behaves when a route needs to be discovered or
repaired.

silabs.com | Building a more connected world. Rev. 1.4 | 9

 UG103.2: Zigbee Fundamentals
Zigbee Routing Concepts

4.3.1 Route Discovery

Route discovery is initiated when a unicast message is sent from one device to another and there is no pre-existing route.

We assume that there is no existing route so the networking software will begin the process of route discovery. For simplicity, assume
that the routing tables of all devices are blank.

For example, assume that device A needs to send a message to device C, as shown in the following figure. Device A will broadcast a
message to the entire network asking the device C to reply. This broadcast message also serves to establish a temporary route back to
A, as each intermediate device records the device from which it received the message. Routes are updated on intermediate nodes —
note that these are temporary entries that have a shorter lifetime than regular entries and are not intended for re-use. Because A is a
one-hop neighbor, B and D do not need to store routing information about it.

Figure 4.1. Example Network

C could use either B or D as its next hop back to A. Zigbee leaves this choice to the implementation; Silicon Labs uses a weighting
algorithm to choose the most apparently reliable next hop.

When the message reaches device C, C sends a special unicast message (called a Route Response message) back to A using the
temporary route constructed in step 1, as shown in the following figure. This message is used by intermediate devices to establish a
(permanent) route back to C.

Figure 4.2. Unicast Message (Green) with Acknowledgement (Blue)

Because C is a one-hop neighbor, B does not need to store routing information about it. D is not involved in this part of the discovery
process because it was not selected by A in the above step. When the message reaches device A, the route discovery is complete and
the new route can be used to send data messages from A to C.

Zigbee PRO networks will detect asymmetric RF links and avoid them during route discovery. This improves the reliability of the discov-
ery process and the resulting routes.

Routes that have not been used within a certain timeout period (1 minute in EmberZNet 3.0 and later) are marked for re-use and new
routes may then overwrite that memory location. In some cases a new route may be needed and one or more intermediate devices will
not have an available routing table entry; in this case the message will be reported as undeliverable to the sending node.

silabs.com | Building a more connected world. Rev. 1.4 | 10

 UG103.2: Zigbee Fundamentals
Zigbee Routing Concepts

The application specifies if an end-to-end acknowledgment should be sent by the receiver (this is called an APS acknowledgment). If
yes, the sender will be notified upon successful delivery of in the case of a timeout waiting for acknowledgment. In the case of a time-
out, the route may need to be repaired.

4.3.2 Route Repair

When a unicast message is sent with an acknowledgment requested, the sending device will be informed when the message is suc-
cessfully delivered. If does not receive this acknowledgment, it can then take steps to repair the route. Route repair follows exactly the
same steps as route discovery, above, but the damaged node (B, in the following figure) does not participate, resulting in a different
route choice.

Figure 4.3. Network with a Damaged Node

The routing table for A is updated to reflect that the next hop is D and the message is successfully delivered along the new path, as
shown in the following figure.

Figure 4.4. Alternative Route

If no alternative path is available, the sender is informed that the message could not be delivered. In EmberZNet PRO this is denoted
by a response with EmberStatus of EMBER_DELIVERY_FAILED (0x66).

EmberZNet PRO will attempt to deliver a message again before performing the route repair. Route repair is performed automatically
when EMBER_APS_OPTION_RETRY and EMBER_APS_OPTION_ENABLE_ROUTE_DISCOVERY are both set in the message op-
tions.

silabs.com | Building a more connected world. Rev. 1.4 | 11

 UG103.2: Zigbee Fundamentals
Zigbee Routing Concepts

4.4 Retries and Acknowledgements

Zigbee and its underlying network layers provide a system of retries and acknowledgments that are designed to efficiently manage the
uncertainty of RF communication. It is not necessary to understand this concept in order to start using Zigbee but it may be of interest
to some application developers in specific situation.

This section discusses retries and acknowledgments layer-by-layer:

• MAC retries and ACKs (802.15.4)
• NWK retries (Zigbee NWK layer)
• APS retries and ACKs (Zigbee APS layer)

4.4.1 MAC Retries and ACKs (802.15.4)

The following figure illustrates the MAC layer transmission retry process.

Figure 4.5. MAC Retry and ACK Process

The MAC layer attempts transmission five times.

• Unicast retries will occur if the channel was not clear (CCA fail) or if the MAC level ACK was not received from the next hop destina-
tion.
• Broadcast retries will occur in the case of CCA failure but broadcast does not use the MAC ACK capability.

These retries occur very quickly — maximum retry time for complete failure is approximately 37 ms. Note that the MAC ACK is sent
back immediately from the sender without additional CCA — see 802.15.4 documentation for more information.

silabs.com | Building a more connected world. Rev. 1.4 | 12

 UG103.2: Zigbee Fundamentals
Zigbee Routing Concepts

4.4.2 NWK Retries

NWK retries in Zigbee are vendor specific. The following figure illustrates the EmberZNet PRO stack’s NWK layer transmission retry
process.

Figure 4.6. NWK Retry Process

The NWK retries occur only if the MAC layer indicates a transmission failure. They operate on a longer time scale than MAC retries and
therefore give the network additional robustness in the presence of medium term (1-500 ms) interference.

Silicon Labs’ interference research shows that the NWK layer retries are important for overcoming temporary interference from WiFi in
certain situations.
• Unicast: unicast behavior is as described in the flow chart, with retries for up to 500 ms.
• Broadcast: broadcast messages are re-sent every 500 ms up to a total of 3 times (including the initial broadcast) or until all neigh-
bors are heard to rebroadcast the message themselves (thereby ensuring complete delivery).

silabs.com | Building a more connected world. Rev. 1.4 | 13

 UG103.2: Zigbee Fundamentals
Zigbee Routing Concepts

4.4.3 APS Retries and ACKs

The following figure illustrates the APS layer retry process.

Figure 4.7. APS Layer Retry Process

The APS layer has an ACK flag that controls whether it uses the additional logic to wait for an acknowledgment and retry if the acknowl-
edgment is not heard. This represents a full end-to-end acknowledgment from the recipient device.

The APS layer can be further optionally configured to repair the route to the destination in the case that the APS sending fails.

There is no equivalent of the APS end-to-end acknowledgment for broadcast messages.

4.4.4 Conclusions

If it is possible to send a message to the destination, the automatic Zigbee behavior provides “maximum effort”, by attempting retries
over several different time scales, from 1 ms to several seconds with optional end-to-end ACK and route repair.

If the delivery fails, the application is recommended to wait a more significant amount of time before retrying — this gives the interfer-
ence or failure time to clear up. In cases of extreme bandwidth congestion the application retries may actually contribute to the problem.

silabs.com | Building a more connected world. Rev. 1.4 | 14

 UG103.2: Zigbee Fundamentals
The Zigbee Stack

5. The Zigbee Stack

Zigbee provides several separate stacks. This has been a point of confusion so the summary is as follows:
• Zigbee 2004 – was released in 2004 and supported a home control lighting profile. This stack was never extensively deployed with
customers and is no longer supported.
• Zigbee 2006 – was released in 2006 and supports a single stack known as the Zigbee stack (explained below)
• Zigbee 2007 – was released in Q4 2007. It has two feature sets, Zigbee and Zigbee PRO.

The Zigbee and Zigbee PRO stacks are complete implementations of the MAC, networking layer, security services and the application
framework. Devices implementing Zigbee and Zigbee PRO can interoperate by acting as end devices in the other type of network. For
example, if a network is formed around a Zigbee PRO coordinator it can have Zigbee PRO routers only, but both Zigbee and Zigbee
PRO end devices.
Note: The Zigbee 2004 stack is not interoperable with Zigbee 2007.

The Zigbee stack is formed around a central coordinator and uses tree addressing to establish the network.

The Zigbee PRO stack is formed around a central coordinator but uses a stochastic addressing scheme to avoid limitations with the
tree. Table routing is always used and additional features are available such as:
• Network level multicasts
• Many-to-one and source routing
• Frequency agility
• Asymmetric link handling
• PAN ID Conflict
• Fragmentation
• Standard or high security.

Silicon Labs does not recommend deploying systems on the Zigbee stack because the Zigbee PRO stack has a number of features
that are necessary for long term network stability and reliability. The Zigbee stack is typically used for small static networks.

Note: The Zigbee 2007 specification includes updates to the Zigbee stack to allow the use of frequency agility and fragmentation on
this stack. This stack remains interoperable with the Zigbee 2006 stack and therefore use of these features must be limited to those
networks which can handle the complexity of some devices having these capabilities and some devices not.

silabs.com | Building a more connected world. Rev. 1.4 | 15

 UG103.2: Zigbee Fundamentals
The Zigbee Stack

5.1 Zigbee Device Object (ZDO)

The ZDO (occupying Endpoint 0 of each node) is a stack-level entity defined by the Zigbee Networking Specification for use in network
management and information gathering. This section explains how to make use of ZDO functionality in the EmberZNet PRO stack
and/or EZSP. An example is described as well.

The ZDO, an entity in the stack, provides network management capabilities that nodes can use to learn about each other, about the
network in general, or for managing certain stack-level functionality. Many features of the ZDO must be implemented as part of a Zig-
bee Certified Platform (ZCP) and are therefore available on all Zigbee/Zigbee PRO devices, making them useful as an interoperable
way to gather and manage system properties on a Zigbee network. The ZDO is implemented through an OTA messaging format known
as the Zigbee Device Profile.

The following basic guidelines apply to using the ZDO for network management:
• Broadcasts requests must go to broadcast mask 0xFFFD.
• Requests should always be sent to destination endpoint 0, with application profile ID 0x0000, as this indicates that the ZDO (intrinsic
in the stack) should handle the message. The source endpoint generally corresponds to the portion of your application that is inter-
ested in the requested information.
• Responses to ZDO requests come from source endpoint 0, to the destination endpoint corresponding to the source used for the
Request.
• If a particular ZDO request handler is not implemented in the stack, a status of EMBER_ZDP_NOT_SUPPORTED (0x84) will be
returned in the response.
• Cluster ID is used to indicate the Request or Response type being sent.
• Most requests can be sent as unicast, broadcast or multicast, just like any other APS frame (with optional APS acknowledgment),
and responses arrive in the IncomingMessageHandler like any other incoming APS message. However, certain kinds of requests, if
sent as broadcasts, will not generate a response (in the interest of network bandwidth conservation). There are also certain requests
that cannot be sent broadcast. For more information on these specific commands, please refer to the Zigbee core specification
(document 14-0392).
• stack/include/zigbee-device-stack.h details the frame formats for commands and responses, including which request types
are supported by the EmberZNet PRO stack. Further details about the frame formats are described in the Zigbee Device Profile
(ZDP) section of the Zigbee Networking Specification (Zigbee document 14-0392).
• Utility functions (provided as source code) for generating ZDO requests and parsing certain kinds of responses can be found in app/
util/zigbee-framework/zigbee-device-library.h. (There is also an alternative version of this file for EZSP host applications.)
• All ZDO requests and responses begin with the transaction sequence number as the first byte of the payload. This allows a request
to be matched up with a particular response.

As an example, take the case of a Permit Joining ZDO transaction, which facilitates remote control of the PermitAssociation flag for one
or more devices in the network.

For this particular transaction, zigbee-device-stack.h explains the following:

1. // Request: [duration:1] [permit:1]

2. // Response: [status:1]
3. #define PERMIT_JOINING_REQUEST 0x0036
4. #define PERMIT_JOINING_RESPONSE 0x8036

So a Permit Joining request frame would use cluster ID 0x0036, and the response would arrive with cluster ID 0x8036. For the request
parameters, the “duration” byte corresponds to the argument you would ordinarily provide to the PermitJoining call in EmberZNet PRO/
EZSP (a number of seconds to allow joining, or 0x00 for always-off joining, or 0xFF for always-on joining). The “permit authentication”
byte used to only be applicable when the recipient of the request was the Trust Center (TC) and controlled whether the TC should
perform authentication on "certain devices" during future association requests.

However, the current Zigbee Networking Specification mandates that the permit authentication byte be 0x01, so this argument in the
Permit Joining Request is no longer significant.

Permit Join Requests are typically sent as a broadcast to all routers, allowing devices to join at any router in the network after this
broadcast request has been processed.

silabs.com | Building a more connected world. Rev. 1.4 | 16

 UG103.2: Zigbee Fundamentals
The Zigbee Stack

5.1.1 Advanced ZDO Usage

The EmberZNet PRO stack supports a feature where the application (host MCU) can be informed about ZDO requests that the stack is
handling, and/or informed about those that the stack does not handle (so that it can handle them itself), and also to give the host a
chance to handle endpoint requests (instead of handling these automatically) if desired. Please see comments in the hal/ember-
configuration.c file (in SoC platforms) or app/util/ezsp/ezsp-enum.h (for NCP platforms) for descriptions of these available op-
tions, which are described as EzspConfigId values 0x26-0x28 (in EZSP) or the following configuration defines (in SoC API):

EMBER_APPLICATION_RECEIVES_SUPPORTED_ZDO_REQUESTS
EMBER_APPLICATION_HANDLES_UNSUPPORTED_ZDO_REQUESTS
EMBER_APPLICATION_HANDLES_ENDPOINT_ZDO_REQUESTS
EMBER_APPLICATION_HANDLES_BINDING_ZDO_REQUESTS

5.2 Zigbee Profiles

Before Zigbee 3.0, application profiles, or simply profiles, sat on top of the basic Zigbee stack. These were developed to specify the
OTA messages required for device interoperability. A given application profile could be certified on either the Zigbee or Zigbee PRO
stack. Now, Zigbee 3.0 has introduced an all-encompassing application layer specification for defining OTA behavior for all Zigbee ap-
plications.

The following are the application profile groups that existed before Zigbee 3.0:
• Home Automation (HA) – Devices for typical residential and small commercial installations.
• Smart Energy (SE) – For utility meter reading and interaction with household devices.
• Commercial Building Automation (CBA) – Devices for large commercial buildings and networks.
• Telecom Application (TA) – Wireless applications within the telecom area.
• Health Care (HC) – Monitoring of personal health in the home or hospital environment.
• Retail – Monitoring and information delivery in a retail environment.
• Zigbee Light Link – Wireless control of LED lighting systems.

The Zigbee Cluster Library (ZCL) forms a generic basis for the Zigbee common application layer. This library defines common elements
that are shared such as data types and allows reuse of simple devices such as on/off switch protocols between different profiles.

Application profiles defined the roles and functions of devices in a Zigbee network. Two types of application profiles were administered
by the Alliance:
• Public Application Profiles, developed by members of the Zigbee Alliance so that devices from different manufacturers can interoper-
ate.
• Manufacturer-Specific Application Profiles, developed by product developers creating private networks for their own applications
where interoperability is not required.

The Zigbee Alliance now recognizes one application layer, described by the Zigbee Base Device Behavior specification, which defines
roles and functions of a device in any Zigbee network. These application behaviors interoperate with any other application.

If you develop a product based upon your own private application layer specification, the Zigbee Alliance requires you to make use of a
unique private profile identifier to ensure the product can successfully co-exist with other products. The Zigbee Alliance issues these
unique private profile IDs to member companies upon request and at no charge.

An application can also be developed using the common Zigbee 3.0 application layer with private extensions for specific features from a
manufacturer.

5.3 Zigbee Addressing Scheme

The Zigbee PRO stack uses a stochastic address assignment mechanism. Under this mechanism the coordinator is still used to start
the network. Each device (routers and end devices) that joins the network is given a randomly assigned address from the device it is
joining. The device conducts conflict detection on this address using network level messages to ensure the address is unique. This
address is then retained by a device, even if the parent address changes. A best practice is to let the stack assign the addresses.

Under Zigbee PRO, provisions are intended for a commissioning tool for setup and configuration of networks. This commissioning tool
can be used to assign addresses manually. Silicon Labs does not provide a commissioning tool. If you are interested in such a tool,
refer to the ZCL Commissioning cluster for information.

silabs.com | Building a more connected world. Rev. 1.4 | 17

 UG103.2: Zigbee Fundamentals
The Zigbee Stack

5.4 Extended PAN IDs

Developers who have used stack software earlier than Zigbee 2007 should note a new and important change to PAN IDs. Zigbee has
added an 8 byte extended PAN ID (EPID or XPID) to facilitate provisioning and PAN ID conflict detection. The extended PAN ID is now
included in the beacon payload, following the existing 3 bytes.

The Extended PAN ID was a network parameter in Zigbee 2007 (EmberZNet 3.x and later) was used in the Zigbee and Zigbee PRO
feature sets. This extended PAN ID [EPID] can be thought of as an extension of the basic, 16-bit PAN ID [PID]. The EPID is a 64-bit
value set for the entire network by the Zigbee Coordinator [ZC] at the time the personal area network [PAN] is formed and must not
change while the PAN is operating (unlike the PID). Like the PID, all nodes within the same PAN share an EPID.

Other than the scanning and joining processes, the EPID rarely appears in transmitted Zigbee packets due to its large overhead (8
bytes) in the header. However, the EPID is used during a Network Update transaction, where a Network Manager node informs the
other devices in the PAN about a PID conflict or channel change, so that nodes moving to a new channel and/or PID can rely on the
consistency of the EPID as a criteria for finding the moved network.

Note that the Extended PAN ID (even if it’s zero) is saved to the non-volatile memory (as part of the tokens) once the node associates
into the network, so it won't change over the lifetime of the device until the device leaves the network.

5.4.1 Use in Scanning / Forming / Joining Process

The EPID will appear (along with other information such as the PID) in the beacon results seen by an Active Scan, and when a join is
initiated, if the EPID is non-zero in the join request, the EPID will take precedence as the primary criteria for matching up the joiner to
the joinee (16-bit PAN ID is effectively ignored). However, it’s possible that some networks (particularly older ones) may not want (or be
able) to deal with EPIDs, so the EPID would be zero in the network parameters from the beacon. If the requested EPID is all zeroes
during the JoinNetwork operation, only the 16-bit PAN ID will be used to match to a network for the purposes of association.

While Silicon Labs provides some sample utilities for scanning networks and parsing the results to make a determination about which
one to join, this process will generally vary depending on the application design and the level of strictness desired in selecting a net-
work. (For example, the joining device may simply want any available Zigbee PRO network; it may seek a closed network designed
specifically for the application in use but may not care which closed network it chooses; or it may seek one specific closed network with
specific properties, in the way that I want my laptop at home to join not just any home WiFi network, but my WiFi network in my home.)

The following guidelines apply to EmberZNet in determining the expected behavior of a Form/Join action given a particular EPID:
• If an all-zero value is specified for extended PAN ID during FormNetwork, the stack will generate a random 64-bit value for this field.
• If an all-zero value is specified for extended PAN ID during JoinNetwork, the stack will use the 16-bit PAN ID specified in the Join-
Network parameters as the primary criteria for selecting a network during the join process.
• If a non-zero value is specified for extended PAN ID during JoinNetwork, the stack will use the 64-bit extended PAN ID specified in
the JoinNetwork parameters (even if 16-bit PAN ID is different) as the primary criteria for selecting a network during the join process.

5.4.2 Choosing an EPID

While the PAN ID is meant to be a randomly chosen, 16-bit value, unique to each network, the EPID is often used more like the SSID
field of WiFi networks to give them a user-friendly designation (which is often not so random and is either set by the manufacturer or the
installer). However, Silicon Labs discourages using a fixed EPID for all deployments because (unlike the conflicts of the PAN ID) EPID
conflicts cannot be resolved if they occur at runtime (because no other unique information exists to distinguish the PANs). Customers
wishing to use non-random EPIDs should, at minimum, scan the network (either at the coordinator or through some external commis-
sioning tool) and check that the new PAN's preferred EPID is not already in use by some other PAN. One approach is to use a small set
of preferred EPIDs when forming PANs so that the coordinator can have alternatives if the first choice is in conflict.

OEMs creating consumer-grade products requiring customer installation (rather than by trained installers) should take particular care to
ensure variety of EPIDs, as two customers living next door to each other may purchase the same product for their homes and would
prefer to isolate their networks from each other. If those two neighboring homes were to each use PANs with the same EPID, network
difficulties would likely arise for both users because the both homes would be considered as one network, and many network address
conflicts could occur.

silabs.com | Building a more connected world. Rev. 1.4 | 18

 UG103.2: Zigbee Fundamentals
The Zigbee Stack

5.4.3 EPID versus PID

Here is a quick overview of differences between the Extended PAN ID and the standard PAN ID:
• EPID is 64 bits; PID is 16 bits.
• EPID is usually used as stack's criteria for matching to the requested network; PID is only used for matching criteria when EPID is all
0x00 bytes.
• EPID is only present in a few kinds of packets (beacons, Network Update messages); PID is present in almost all 802.15.4 frames
(except MAC ACKs).
• EPID is used as criteria for uniquely identifying the network and for resolving conflicts of PID; PID is used for MAC destination filter-
ing at the radio receiver.
• EPID may help provide some indication of the network's identity in the scan results; PID should always be completely random, so it
is not as useful in determining which PAN is the "right one".
• EPID can be any value in range of 0x0000000000000001 to 0xFFFFFFFFFFFFFFFE (all 0's and all F's are reserved values); PID
can be any value in the range of 0x0000 - 0xFFFE (all F's constitute reserved value).

silabs.com | Building a more connected world. Rev. 1.4 | 19

 UG103.2: Zigbee Fundamentals
Zigbee Cluster Library

6. Zigbee Cluster Library

6.1 Overview

In the Zigbee Cluster Library (ZCL), a cluster is a set of messages used to send and receive related commands and data over a Zigbee
network. For example, a temperature cluster would contain all the necessary OTA messages required to send and receive information
about temperature.

To facilitate learning and management, these clusters are further grouped into functional domains, such as those useful for HVAC, Se-
curity, Lighting, and so on. Developers may also define their own clusters, if the pre-defined clusters do not meet their specific applica-
tion needs.

The Zigbee common application layer then references which clusters are used within certain applications, and specifies which clusters
are supported by different devices–some clusters are mandatory, others optional. In this way, the ZCL simplifies the documentation of a
particular application and allows the developer to understand quickly which behaviors devices support.

A more detailed overview of the ZCL, the format of messages within clusters, and a set of messages that may be used within any clus-
ter are described in the Zigbee Cluster Library Specification document (15-02017-002). Functional domain clusters are described in
separate documents, such as the Functional Domain: Generic, Security and Safety document.

Silicon Labs provides source code to easily assemble and disassemble ZCL messages, whether they are pre-defined by the ZCL or
custom messages created by the developer. In addition, Silicon Labs provides a powerful tool called AppBuilder that allows developers
to create and configure most of their ZCL-based application. See document UG391: Zigbee Application Framework Developer's Guide
for more information.

6.2 Inside Clusters

6.2.1 Clients and Servers

Each cluster is divided into two ends, a client end and a server end. The client end of a cluster sends messages that may be received
by the server end. The client end may also receive messages that are sent by the server end. In this sense, the client and server ends
of a cluster are always complementary. In contrast to many other systems (for example, HTTP), both have the same potential for send-
ing and receiving messages: the “client” designation does not imply a subordinate or response-only status.

Because all commands have a sender and a receiver, each cluster is described in two parts – a server part and a client part, as shown
in the following figure. A device supporting the server half of a cluster will communicate with a device supporting the client half of the
same cluster.

Figure 6.1. A Zigbee Cluster

This equality complicates discussions; for clarity, this document always refers to “cluster end” when one of the client or the server end
must be used, “cluster ends” when speaking of both client and server ends, and “cluster server” or “cluster client” when a specific end is
required (usually examples).

silabs.com | Building a more connected world. Rev. 1.4 | 20

 UG103.2: Zigbee Fundamentals
Zigbee Cluster Library

6.2.2 Attributes

An attribute is data associated with a cluster end; the server and client ends of a cluster may each possess multiple attributes.

Each attribute declares a 16-bit identifier, a data type, a read-only or read/write designator, a default value, and an indicator of whether
its support by any implementation is mandatory or optional. The following table lists example data types.

Table 6.1. Data Type Quick Reference (Example Data Types)

Data Type Description

Binary data types 8, 16, 24, or 32 bits in length

Logical data type Boolean

Bitmap data type 8, 16, 24, or 32 bits in length

Unsigned Integer 8, 16, 24, or 32 bits in length

Signed Integer: 8, 16, 24, or 32 bits in length

Enumeration: 8 or 16 bits in length

Floating Point: 2-byte semi-precision, 4-byte single precision, or 8 byte double precision

String: binary octet string or character string; first byte is length

Time: time of day, date

Identifier: cluster ID, attribute ID

IEEE address type 8-byte unique IEEE address of an 802.15.4 radio

The attribute identifier is unique only within the specific cluster end: this means that the attribute 0x0002 within the cluster server does
not need to be the same as the attribute 0x0002 within the cluster client, even within the same cluster.

The data types are fully described in the ZCL Specification. The following table is an excerpt from that document.

Table 6.2. Example ZCL Data Types

Type Class Data Type ID Data Type Length of Data Invalid Analog / Discrete
(octets) Number

Null 0x00 No data 0 - -

0x01 – 0x7 Reserved - -

General Data 0x08 8-bit data 1 - D

0x09 16-bit data 2 -

0x0A 24-bit data 3 -

0x0B 32-bit data 4 -

0x0C – 0x0F Reserved - -

Logical 0x10 Boolean 1 0xFF D

0x11 – 0x17 Reserved - -

Bitmap 0x18 8-bit bitmap 1 - D

0x19 16-bit bitmap 2 -

0x1a 24-bit bitmap 3 -

0x1b 32-bit bitmap 4 -

0x1C – 0x1F Reserved - -

silabs.com | Building a more connected world. Rev. 1.4 | 21

 UG103.2: Zigbee Fundamentals
Zigbee Cluster Library

Type Class Data Type ID Data Type Length of Data Invalid Analog / Discrete
(octets) Number

Unsigned Integer 0x20 Unsigned 8-bit integer 1 0xFF A

0x21 Unsigned 16-bit integer 2 0xFFFF

0x22 Unsigned 24-bit integer 3 0xFFFFFF

0x23 Unsigned 32-bit integer 4 0xFFFFFFFF

0x24 – 0x27 Reserved - -

Signed Integer 0x28 Signed 8-bit integer 1 0x80 A

0x29 Signed 16-bit integer 2 0x8000

0x2A Signed 24-bit integer 3 0x800000

0x2B Signed 32-bit integer 4 0x80000000

0x2C – 0x2F Reserved - -

Enumeration 0x30 8-bit enumeration 1 0xFF D

0x31 16-bit enumeration 2 0xFFFF

0x32 – 0x37 Reserved - -

Floating Point 0x38 Semi-precision 2 Not a Number A

0x39 Single precision 4 Not a Number

0x3A Double precision 8 Not a Number

0x3B – 0x3F Reserved - -

String 0x40 Reserved - - D

0x41 Octet string Defined in first octet 0xFF in first octet

0x42 Character string Defined in first octet 0xFF in first octet

0x43 – 0x47 Reserved - -

Array 0x48 – 0x4F Reserved - - -

List 0x50 – 0x57 Reserved - - -

Reserved 0x58 – 0xDF - - - -

Time 0xE0 Time of day 4 0xFFFFFFFF A

0xE1 Date 4 0xFFFFFFFF

0xE2 – 0xE7 Reserved - -

Identifier 0xE8 Cluster ID 2 0xFFFF D

0xE9 Attribute ID 2 0xFFFF

0xEA BACnet OID 4 0xFFFFFFFF

0xEB – 0xEF Reserved - -

Miscellaneous 0xF0 IEEE address 8 0xFFFFFFFFFFF D

FFFFF

0xF1 – 0xEF Reserved - - -

Unknown 0xFF Unknown 0 - -

Attributes may be accessed over the air by use of the attribute commands described later in this document.

silabs.com | Building a more connected world. Rev. 1.4 | 22

 UG103.2: Zigbee Fundamentals
Zigbee Cluster Library

Global attributes were introduced in revision 6 of the ZCL. These attributes are defined for every cluster instance on a device and func-
tion as normal ZCL attributes. Cluster Revision (attribute ID 0xFFFD) is one example of a global attribute. This attribute is a way to
version each of the individual cluster specifications, to improve backwards compatibility, and allow for testing of specific versioned clus-
ter behavior. One device may read the Cluster Revision attribute on another device to determine how to communicate certain cluster
behavior to them over the air. This Cluster Revision attribute is incremented for every major change to each cluster’s specification.

6.2.3 Commands

A command is composed of an 8-bit command-identifier and a payload format. Like attributes, the 8-bit identifier is unique only within
the specific cluster end. The payload format is arbitrary to the command type, conforming only to the general packet format guidelines
as described in the ZCL Specification.

Commands are divided into two types: global and cluster-specific. Global commands are defined in the ZCL Specification and are not
specific to any cluster. These global commands were originally referred to as Profile-Wide, but have changed name to fall in line with
the Zigbee 3.0 common application layer. Cluster-specific commands are defined inside the cluster definitions in the ZCL functional do-
main documents, and are unique to the cluster in which they are defined.

Global Commands

Global commands are not unique to a specific cluster; they are defined in the ZCL General Command Frame (see ZCL Specification
075123r02, Chapter 7). The following table lists example profile-wide commands:

Table 6.3. Example Profile Wide Commands

Messages Sent to the Cluster End Supporting the Attribute Messages Sent From the Cluster End Supporting the Attrib-
ute

Read Attributes Read Attributes Response

Write Attributes Write Attributes Response/No Response

Write Attributes Undivided Write Attributes Response/No Response

Configure Reporting Configure Reporting Response

Read Reporting Configuration Read Reporting Configuration Response

Discover Attributes Discover Attributes Response

Report Attributes Report Attributes

Default Response Default Response

silabs.com | Building a more connected world. Rev. 1.4 | 23

 UG103.2: Zigbee Fundamentals
Zigbee Cluster Library

• Read Attributes: Requests one or more attributes to be returned by the recipient; replies with Read Attributes Response.
• Write Attributes: Provides new values for one or more attributes on the recipient; the reply will contain a Write Attributes Response
portion indicating which attributes were successfully updated, and/or a Write Attributes No Response portion for attributes that were
not successfully updated.
• Write Attributes Undivided: Updates all attributes and replies with Write Attributes Response; if any single attribute cannot be upda-
ted, no attributes are updated and this command replies with Write Attributes No Response.
• Configure Reporting: Configures a reporting interval, trigger events, and a destination for indicated attributes. Replies with Configure
Reporting Response.
• Read Reporting Configuration: Generates a Read Reporting Configuration Response containing the current reporting configuration
sent in reply.
• Discover Attributes: Requests all supported attributes to be sent; replies with a Discover Attributes Response.
• Report Attributes: A report of attribute values configured by Configure Reporting command.
• Default Response: A response sent when no more specific response is available (and the default response is not disabled by the
incoming message).

Since attributes are always tied to a cluster, the commands affecting attributes specify which cluster and which attributes are to be ac-
cessed or modified. Additionally, each cluster defines which attributes support which commands – for example, an attributes may be
declared READ ONLY, in which case it will not support the Write Attributes command. Thus, while the command format is not cluster-
specific, the attributes it describes and its result on the receiving system are both cluster-specific.

Readers interested in more detail about the format or specific behaviors of these messages should review the ZCL Specification
(075123r02).

Cluster-Specific Commands

The payload format, support requirements (mandatory, optional), and behavior on receipt of a cluster-specific command are all defined
in the cluster definition. Typically, these commands affect the state of the receiving device and may alter the attributes of the cluster as
a side-effect.

For example, the ZCL defines three commands (OFF, ON, and TOGGLE) that are received by the On/Off cluster server. It further de-
clares that each of these commands is mandatory and the payload format for each command (in this case, none of them have pay-
loads). The ZCL defines that the On/Off cluster client is responsible for generating the commands received by the server.

silabs.com | Building a more connected world. Rev. 1.4 | 24

 UG103.2: Zigbee Fundamentals
Zigbee Cluster Library

6.3 Example: Temperature Measurement Sensor Cluster

As an example, consider a portion of the Temperature Measurement Sensor cluster that is fully described in Chapter 4 of the Zigbee
Cluster Library Specification (15-02017-002).

The following table lists some of the attributes from the ZCL specification.

Table 6.4. Temperature Measurement Sensor Server Attributes (incomplete)

Identifier Names Types Range Access Default Mandatory/

Optional

0x0000 MeasuredValue Signed 16-bit MinMeasuredValue to Read only 0 M

Integer MaxMeasuredValue

0x0001 MinMeasuredValue Signed 16-bit 0x954B – 0x7FFE Read only - M

Integer

0x0002 MaxMeasuredValue Signed 16-bit 0x954C – 0x7FFF Read only - M

Integer

0x0003 Tolerance Unsigned 16- 0x0000 – 0x0800 Read only - O

bit Integer

0xFFFD ClusterRevision Unsigned 16- 0x0001 – 0xFFFE Read only 0x0001 M

bit Integer

1. Overview of Attributes: The cluster server supports at least those five attributes, four of which must be supported by any imple-
mentation (MeasuredValue, MinMeasuredValue, MaxMeasuredValue, and ClusterRevision), and one of which may be optionally
supported (Tolerance). All of these attributes are read-only, indicating that any write attempts to them will fail. Note the ClusterRevi-
sion global attribute functions as a regular attribute.
2. Implications for Cluster Server, Cluster Client: Clearly, the cluster server should be implemented by the device that contains the
temperature sensor. Meanwhile the cluster client should be implemented by any device that wishes to receive temperature sensor
information, either actively (through a read attributes command) or passively (through first a configure report attributes command
and then from report attributes).
3. Further Information: The cluster description also provides useful information about the actual format of the data (for example, the
range of the signed 16-bit integer for MaxMeasuredValue) and the mandatory supported operations – not all attributes support all
basic commands. For example, MaxMeasuredValue is a read-only command and cannot be written.
Note: While the incoming write attribute commands are supported, in this case MaxMeasuredValue always generates a Write At-
tributes No Response reply.

4. Commands: No custom commands are supported by this cluster (either the server or the client). For an example of a cluster con-
taining custom commands, see the Thermostat Cluster in the ZCL (15-02017-002).

silabs.com | Building a more connected world. Rev. 1.4 | 25

 UG103.2: Zigbee Fundamentals
Zigbee Cluster Library

6.4 Functional Domains

As of this writing, the Zigbee Cluster Library defines the following functional domains:
• General
• Closures
• HVAC
• Lighting
• Measurement & Sensing
• Security & Safety
• Protocol Interfaces
• Smart Energy
• Over-the-Air Upgrading
• Telecommunications
• Commissioning
• Retail
• Appliances

Each domain defines a number of clusters that are then used by the Zigbee Application Layer to describe the OTA behavior of devices
(see the following figure).

Figure 6.2. Cluster Library Functional Domains

6.5 Manufacturer Extensions

The ZCL allows extension of the existing library in two ways: users may add manufacturer specific commands or attributes to existing
clusters, or they may define entirely new clusters that are manufacturer specific.

Manufacturer-specific commands are identified by setting a special bit in the ZCL header and including the manufacturer code (received
from the Zigbee Alliance) in the ZCL header. This guarantees that manufacturer-specific extensions do not interfere with other manu-
facturer-specific extensions or existing ZCL clusters, commands, or attributes.

Manufacturer-specific clusters must use the cluster ID range of 0xFC00-0xFFFE.

silabs.com | Building a more connected world. Rev. 1.4 | 26

 UG103.2: Zigbee Fundamentals
Zigbee Compliance

7. Zigbee Compliance

Zigbee compliance is based on a building block of compliance testing used to ensure that each layer is tested.

Products that meet all compliance requirements may be branded "Zigbee Compliant" and can display the Zigbee logo.

The Zigbee radio and MAC are required to have passed the applicable parts of IEEE 802.15.4 certification testing. Parts of the MAC not
required for Zigbee operation are not required for this testing.

Zigbee stacks are required to be built on certified IEEE 802.15.4 platforms. Zigbee stack providers are required to have the stack tested
against a standard Zigbee test plan known as Zigbee Compliant Platform Testing. This testing validates the basic network, security, and
Zigbee Device Object (ZDO) operations for the stack.

Zigbee products are required to be built on a Zigbee Compliant Platform. The developer can choose to build a public or manufacturer-
specific application. Those applications which are manufacturer-specific are tested against a specific set of tests to validate basic oper-
ation. Products based on the common Zigbee application layer are more extensively tested using a standard Zigbee test for the applica-
tion layer commands. Testing scrutiny has been heavily increased in Zigbee 3.0; there are now many negative test cases that products
must pass in order to pass certification.

The Zigbee Test Harness is new in Zigbee 3.0. The Zigbee Test Harness is provided to Zigbee members for testing their products and
also is used for the final compliance testing for a Zigbee 3.0 product. The Zigbee Test Harness involves both a test harness Zigbee
software stack running on a compliant IEEE 802.15.4 platform and a desktop application offering a friendly user interface to the tester.

Each shipping product has to be certified and substantive changes to the product after certification can require recertification. Silicon
Labs sends all chips and stack releases through compliance testing.

All of the Zigbee test plans are developed and tested using at least three members’ products to ensure interoperability between differ-
ent implementations.

Companies are required to become members of the Zigbee alliance to ship product containing the Zigbee stack. There are specific
rules for use of the Zigbee logo on product that are detailed on the Zigbee web site (www.zigbee.org).

silabs.com | Building a more connected world. Rev. 1.4 | 27

 UG103.2: Zigbee Fundamentals
Zigbee 3.0

8. Zigbee 3.0

Zigbee 3.0 marks the beginning of a new chapter in the Zigbee Alliance. Zigbee 3.0 unites all of the different application profiles into
one common application layer. Furthermore, it introduces greater test coverage for product certification, ensuring better interoperability
for Zigbee devices in the field. The Zigbee 3.0 document suite contains both revised and completely new material for the Zigbee appli-
cation specification.

The Base Device Behavior Specification (BDB) is perhaps the most important new document in the Zigbee 3.0 document suite. It speci-
fies mandatory and optional application layer behavior for any Zigbee product. These behaviors include, but are not limited to:
• Network commissioning
• Network security
• ZDO command handling
• Reporting
• MAC data polling
• Persistent data

See document 13-0402 for more specific information on these items. Because there is now one specification for all Zigbee applications,
two different Zigbee 3.0-compliant products in different application domains can interoperate on the same network (for example, a light
switch and a garage door opener can perform separate functions on the same network).

At the heart of the BDB are the commissioning methods that a Zigbee 3.0 device can use. The methods follow an order of execution,
but they are all optional for a device. For example, it may only makes sense for a Zigbee End Device switch application to performing
touchlink commissioning, classical joining, and then finding and binding. On the other hand, a Zigbee Coordinator gateway application
may only wish to form a network. Lastly, maybe a Zigbee Router light application could perform classical joining, and then fall back to
classical network formation if it does not find any networks to join. The commissioning methods, and order, are as follows.
• Touchlink commissioning
• Classical network joining
• Classical network formation
• Finding and binding

Touchlink commissioning is the first step in the Zigbee 3.0 network commissioning process. A device may choose to perform the tou-
chlink process for an initiator to try and find a touchlink target. A typical example of this interaction may be a switch and a light. The
switch may perform touchlink commissioning to establish itself on the same network as the light. See 13-0402 for a more in-depth de-
scription of touchlink commissioning.

The next step in the BDB defined process is classical joining. This means finding open networks and joining one using the original Zig-
bee joining process. Finding a network involves performing IEEE 802.15.4 active scans on a product-defined set of channels. Once a
device has found a network that it wishes to join, it may join that network using an install code derived link key, a default link key for a
centralized security network, and a default link key for a distributed security network. For more information on the security mechanisms
involved in the Zigbee 3.0 application layer, see AN1233: Zigbee Security.

A Zigbee 3.0 application may then choose to form a classical Zigbee network. A product may form a centralized security network as a
trust center or form a distributed security network as a router. This is different from many legacy application profiles where only a Zig-
bee coordinator could form a network. A light may wish to form a network as a router so that it can act as a touchlink target, whereas it
makes more sense for a gateway to form a centralized network as a trust center. Trust centers may choose to only allow only certain
devices on the network by using install code derived link keys for joining. Again, see AN1233: Zigbee Security for a thorough discussion
of Zigbee 3.0 security.

Once devices are on a network, they can perform finding and binding, where devices create bindings to establish application layer links.
For example, an On/Off switch may perform finding and binding to create a binding to an On/Off Light. Again, as all commissioning
steps, this step is optional.

As mentioned previously, Zigbee 3.0 also introduces more stringent testing on products. This means products will go through more neg-
ative test cases and specific cluster functionality tests before they are allowed on the market. Included in the Zigbee 3.0 document suite
are the new Cluster Test Specifications. These documents contain test cases for Zigbee products implementing certain clusters. For a
product using a certain cluster to be Zigbee 3.0 certified, it must pass the test cases contained within this document for that specific
cluster. There is an individual Zigbee document for each cluster test spec. For an example, see document 15-0310, the On/Off Cluster
test specification.

Zigbee 3.0 also mandates that devices should support proxy functionality for Green Power, a networking and application-level specifica-
tion designed for energy-harvesting end devices. Essentially, this means that every Zigbee 3.0 device will be able to process and for-
ward Green Power messages to other devices with more application-layer Green Power functionality. The Green Power Basic specifi-
cation and test specification are included in the Zigbee 3.0 document suite. See UG103.15: Silicon Labs Green Power Fundamentals
for more information.

silabs.com | Building a more connected world. Rev. 1.4 | 28

 UG103.2: Zigbee Fundamentals
Zigbee 3.0

It is worth mentioning that the 6th revision of the Zigbee Cluster Library (document 15-02017) is also included in the Zigbee 3.0 docu-
ment suite. See section 6. Zigbee Cluster Library for more information about this specification.

silabs.com | Building a more connected world. Rev. 1.4 | 29

 UG103.2: Zigbee Fundamentals
Applying Zigbee

9. Applying Zigbee

A set of design decisions must be made in any Zigbee implementation. For more information, see UG103.3: Software Design Funda-
mentals.

silabs.com | Building a more connected world. Rev. 1.4 | 30

 Smart.
Connected.
Energy-Friendly.

Products Quality Support and Community

www.silabs.com/products www.silabs.com/quality community.silabs.com

Disclaimer
Silicon Labs intends to provide customers with the latest, accurate, and in-depth documentation of all peripherals and modules available for system and software implementers using or
intending to use the Silicon Labs products. Characterization data, available modules and peripherals, memory sizes and memory addresses refer to each specific device, and "Typical"
parameters provided can and do vary in different applications. Application examples described herein are for illustrative purposes only. Silicon Labs reserves the right to make changes without
further notice to the product information, specifications, and descriptions herein, and does not give warranties as to the accuracy or completeness of the included information. Without prior
notification, Silicon Labs may update product firmware during the manufacturing process for security or reliability reasons. Such changes will not alter the specifications or the performance
of the product. Silicon Labs shall have no liability for the consequences of use of the information supplied in this document. This document does not imply or expressly grant any license to
design or fabricate any integrated circuits. The products are not designed or authorized to be used within any FDA Class III devices, applications for which FDA premarket approval is required
or Life Support Systems without the specific written consent of Silicon Labs. A "Life Support System" is any product or system intended to support or sustain life and/or health, which, if it fails,
can be reasonably expected to result in significant personal injury or death. Silicon Labs products are not designed or authorized for military applications. Silicon Labs products shall under no
circumstances be used in weapons of mass destruction including (but not limited to) nuclear, biological or chemical weapons, or missiles capable of delivering such weapons. Silicon Labs
disclaims all express and implied warranties and shall not be responsible or liable for any injuries or damages related to use of a Silicon Labs product in such unauthorized applications.

Trademark Information
Silicon Laboratories Inc.® , Silicon Laboratories®, Silicon Labs®, SiLabs® and the Silicon Labs logo®, Bluegiga®, Bluegiga Logo®, ClockBuilder®, CMEMS®, DSPLL®, EFM®,
EFM32®, EFR, Ember®, Energy Micro, Energy Micro logo and combinations thereof, "the world’s most energy friendly microcontrollers", Ember®, EZLink®, EZRadio®, EZRadioPRO®,
Gecko®, Gecko OS, Gecko OS Studio, ISOmodem®, Precision32®, ProSLIC®, Simplicity Studio®, SiPHY®, Telegesis, the Telegesis Logo®, USBXpress® , Zentri, the Zentri logo and Zentri
DMS, Z-Wave®, and others are trademarks or registered trademarks of Silicon Labs. ARM, CORTEX, Cortex-M3 and THUMB are trademarks or registered trademarks of ARM Holdings.
Keil is a registered trademark of ARM Limited. Wi-Fi is a registered trademark of the Wi-Fi Alliance. All other products or brand names mentioned herein are trademarks of their respective
holders.

Silicon Laboratories Inc.

400 West Cesar Chavez
Austin, TX 78701
USA

http://www.silabs.com