SWAY034 April   2021 CC1312R7 , CC1352P , CC1352P7 , CC1352R , CC2642R , CC2642R-Q1 , CC2652P , CC2652P7 , CC2652PSIP , CC2652R , CC2652R7 , CC2652RB , CC2652RSIP

 

  1. At a glance
  2. 2
  3. Introduction
  4. Design considerations
  5. Wireless protocols
  6. Zigbee
  7. Thread
  8. Similarities Between Zigbee and Thread
  9. Differences Between Zigbee and Thread
  10. 10 Project CHIP
  11. 11Conclusion
  12. 12References
  13. 13Important Notice

Similarities Between Zigbee and Thread

Zigbee and Thread technologies are standard-based protocols that primarily operate in the worldwide 2.4-GHz ISM band. These technologies provide a built-in mesh networking security and application infrastructure for embedded, low-power and low-cost devices.

The Zigbee Alliance and Thread Group both have a process for member companies to enact changes to the specification.

Both protocols leverage a common underlying data-link communication layer designed and maintained by the IEEE.

Figure 8-1 shows Zigbee and Thread protocol layering.

GUID-A7E68220-D360-4919-81EF-9B0B0BBEBF58-low.svg Figure 8-1 Zigbee and Thread Protocol Layering

The IEEE 802.15.4 standard specifies the Media Access Control (MAC) and physical (PHY) layers of the Open Systems Interconnection communication model. Both Zigbee and Thread implement a personal area network that guarantees a reliable hop-to-hop link for the transfer of upper-layer data frames at very low-power operation.

Since higher, less timing-sensitive protocol layers are implemented in software, it’s possible to implement 802.15.4-based standards like Zigbee and Thread as different software variants that run on the same silicon (as is the case for the TI SimpleLink multistandard CC2652R wireless microcontroller [MCU]). With a single unique hardware design, the corresponding firmware can be loaded at the factory or upgraded in the field, providing a simplified and futureproof solution.

Both Zigbee and Thread implement an asynchronous mode of operation within the IEEE 802.15.4 standard. This transmitter-initiated profile enables the efficient exchange of small packets in a low-power wireless network. Devices that do not generate data often can wake up and reliably send packets with extremely short latency.

Regardless of the data’s destination in the network (one or multiple hops away), the battery-powered devices wake up from sleep, send the data to their one-hop relay node, and then quickly go back into a standby state. Between instances when the device is active and sending or receiving data, the radio can be off and operating in the realm of microamperes. For instance, the CC2652R device can sleep while retaining full random-access memory contents and consume only 0.9 µA.

Both Zigbee and Thread use a distance vector algorithm to build routing tables between routers. Zigbee uses ad-hoc on-demand distance vector routing, and Thread uses a modified Routing Information Protocol. Having the routers of each network generate and store the routing information rather than the end devices minimizes network maintenance traffic to the end nodes, conserving radio time.

This efficiency is significantly advantageous for devices that typically generate data triggered by sporadic alarm events (such as door and window sensors) or user actions (such as switches/key fobs, alarm panels or shade systems). The battery-powered devices can sleep most of the time, only waking up for the occasional application-initiated data or periodic data poll messages. The periodic data polls are needed for unsolicited down-link messages and to maintain connection with the end device’s parent router.

With peak current levels around the single-digit microamperes, Zigbee and Thread enable devices in the home and building automation space to operate for years off of coin-cell batteries. Header compression and reuse make communication in both Zigbee and Thread efficient by creating smaller over-the-air packets. Thread leverages 6LoWPAN compression, fragmentation and link-layer forwarding. Zigbee was designed from the ground up, with binary-data optimization in the networking protocol for the underlying 802.15.4 frames.

The headers and networking management operations necessary to maintain and establish routes are short and reliably enable a 20-byte application frame (for a lighting control command or an alarm event) in a single 802.15.4 packet instance of 50 to 80 bytes, with a turnaround time of a few tens of milliseconds per hop. In most systems, with four to five hops as the mesh branch’s biggest length, this speed still provides less than 100 ms of latency for actuating device-to-device communication.

Low-power operation and network scalability are both important requirements in residential systems with tens of nodes interoperating, such as lights, environmental sensors and thermostats. But these factors are an even bigger priority in commercial and industrial building automation systems, where the number of devices may reach hundreds or even thousands of nodes.

Both the Zigbee and Thread protocols implement an efficient routing algorithm to minimize over-the-air traffic and broadcasts. The receiver in these nodes is always on (they are usually mains-powered, like a light bulb/fixture or a thermostat) and store next hops to the final destination by building a small and lean routing table. The networks don’t relay packets by flooding the network through broadcasts, which ultimately can hinder scalability.

The routing nodes exchange only small intermittent broadcast messages, minimizing overall housekeeping traffic to maintain the mesh. Routing nodes in the network also have the important role of buffering the data for the downlink communication of their sleeping “children,” which can be configured to extract packets efficiently depending on the downlink requirements (which in many cases are latency-insensitive).

Both Zigbee and Thread technologies have been demonstrated successfully in large commercial deployments that reach hundreds of nodes within the same network. TI has deployed Breaking the 400-Node ZigBee Network Barrier TI's ZigBee SoC & Z-Stack Software, and the technology can enable even larger networks depending on node density, amount of traffic generated and application profile.