Low-power design for ZigBee wireless solutions
ZigBee networks set the most stringent power requirements yet for wireless applications
Emerging wireless networks for control and monitoring applications within our homes, offices, and industrial environments target a vast number of applications and have tremendous commercial potential. Power efficiency is always a prime concern in these wireless networks in addition to low cost, high performance, and reliability. ZigBee wireless technology and the underlying IEEE 802.15.4 standard offer a wide variety of power-conservation features at implementation and operational levels providing a good fit for power-constrained applications. The technical merits and commercial potential of ZigBee wireless technology have resulted in broad industry support from both technology manufacturers and end-user application providers. With its ratification in 2004 and public release in 2005, the fundamental conditions now seem to be in place to enable ZigBee to quickly reach large market volumes already this year.
In this article, we first describe the overall characteristics of the IEEE 802.15.4 standard and how the network, security, and application layers of ZigBee build upon it. Then, with reference to low-power design, we discuss the features that affect the overall power budget of ZigBee wireless solutions and highlight implementation techniques for power optimization of the microcontroller, radio transceiver, and overall system design.
The ZigBee wireless network technology builds upon the IEEE 802.15.4 standard, as illustrated in Figure 1, and is specifically designed to provide cost-effective, standardized, and flexible wireless networks, which support low power consumption, reliability, interoperability and security for control and monitoring applications with low to moderate data rates. ZigBee is suitable for a large number of applications varying from industrial monitoring and control, home and building automation, and sensor networks to medical and automotive solutions.
The IEEE 802.15.4 standard,1 ratified in May 2003, specifies the physical (PHY) and media access control (MAC) layers at the 868-MHz (Europe), 915-MHz (U.S.), and 2.4-GHz (worldwide) ISM bands, enabling regional or global deployment. The air interface is Direct Sequence Spread Spectrum (DSSS) using binary phase-shift keying (BPSK) for the 868/915-MHz PHY and offset quaternary phase-shift keying (O-QPSK) for the 2.4-GHz PHY. The principal physical specifications of the IEEE 802.15.4 are summarized in the table. The IEEE 802.15.4 PHY also includes receiver energy detection, link quality indication, and clear channel assessment.
The IEEE 802.15.4 MAC sublayer controls the access to the radio channel using the Carrier Sense Multiple Access with Collision Avoidance (CSMA-CA) method and handles network (dis)association and MAC layer security (AES-128 encryption-based). It is also responsible for flow control via acknowledgment and retransmission of data packets, frame validation, and network synchronization as well as support to upper layers for robust link operation. The IEEE 802.15.4 standard employs 64-bit IEEE and 16-bit short addresses to theoretically support more than 65,000 nodes per network.
ZigBee wireless technology,2 ratified in December 2004 and publicly available in June 2005, specifies the network, security, and application layers upon the IEEE 802.15.4 PHY and MAC layers. The ZigBee Alliance also provides interoperability and conformance testing specifications. The ZigBee network layer is responsible for device discovery and network configuration and supports three networking topologies: star, tree, and mesh (see Figure 2).
Wireless sensor networks within the private sphere and industrial environment enable easy installation and maintenance, flexible connectivity and mobility, and an increased coverage area, and open up new monitoring and control capabilities in a cost-effective manner.
In many applications, the wireless node must be designed for a lifetime ranging from 5 to 12 years without changing or recharging the battery. This imposes drastic constraints on the power consumption of the node. Low-power design philosophy and power optimization for the targeted lifetime of the node are thus mandatory for the viability of such wireless (sensor) networks.
Low-power operation is one of the defining characteristics of wireless networks based on ZigBee. A low-power system can be achieved as the result of comprehensive power awareness across the entire hierarchy, from the standard-defined properties to architecture of the receiver and transmitter to the baseband DSP and the transistor-level circuits comprising the RF and IF analog circuits.
The IEEE 802.15.4/ZigBee technology is specified with low-power features at implementation and operational levels in order to enable broad commercial adaptation within power-constrained applications, such as sensor networks. Some operational power-saving features include:
• Very low duty cycle operation.
• Strict power management.
• Low transmission overhead.
At the implementation level, low-power devices can be made useful by exploiting the following:
• Relatively low transmit output power to limit current draw during transmission.
• Controllable output power to send just enough signal strength.
• Constant envelope modulation for efficient power amplifier design.
• Modulation scheme with low signal-to-noise requirement.
• Relatively relaxed channel spacing and blocking requirements.
Although the technical considerations in this article evolve around the 2.4-GHz PHY, most of them are also valid for 868/915-MHz systems from a power optimization perspective.
Scalable and power-friendly ZigBee system implementations can satisfy a wide range of applications with diverse complexities by a physical and logical split at or slightly above the PHY layer in Figure 1. This approach relies in principle on the combination of an integrated radio transceiver providing PHY functionalities (and in some cases, a selective small part of the MAC) and a microcontroller unit (MCU) containing the MAC network and application layers as shown in Figure 3. In this partitioning, the RF transceiver and the MCU can be selected to perfectly fit the application. No matter what the specific component choices are, low-power performance will certainly be on top of the system designer’s checklist.
The MCU device provides the processing capability for execution of protocol stack, network, and application software. The fundamental idea of achieving low power for the data processing tasks in a ZigBee system is simple: Use an MCU designed for ultralow power. Key power-saving features at the MCU part are briefly described below.
Low current consumption in active and sleep modes. Obviously, the MCU must provide the necessary processing bandwidth at minimum power consumption while it is active. Although the ZigBee/IEEE 802.15.4 specification has been tailored to minimize the active power consumption of compliant devices, achieving months and years of operation from one single battery must be supplemented by very low duty cycle operation. In such ultralow-power designs, the average current consumption determines battery lifetime. Having the system in sleep most of the time (low duty cycle) means that the average power consumption will gravitate toward the sleep-mode power value of the system. The system should therefore employ an MCU with the lowest current consumption also in sleep mode.
There are two different sleep modes: shutdown and standby. During shutdown, the MCU is completely turned off with no timers or clocks operating; the MCU will wake up by an external triggering event. For a wireless system, the external event can come from the RF transceiver waking up the MCU when a received message needs to be processed. The standby mode can be a real-time clock operation, in which the MCU utilizes a timer to accurately track the time (commonly done with an external 32.768-kHz crystal or a recurring wakeup typically implemented with an internal oscillator).
Zero-power brownout protection. To provide stable and robust operation without wasting a lot of power, the MCU should support a low-power brownout reset protection, which enables the MCU to continuously monitor and reset itself to protect valuable data and to avoid damage to the controller when the battery voltage drops below the threshold specification. Brownout protection must be enabled all the time to be useful because brownouts cannot be predicted. The brownout’s current draw adds directly to the bottom line of the power budget.
Short wakeup period. The MCU should have very short wakeup time to minimize wasted power during the transition from idle to full-speed, active state.
Self-managed peripherals. Low-power MCUs operate peripherals without CPU intervention to minimize active CPU current consumption. Self-managed peripherals are preconfigured so the MCU does not need to be utilized. For example, an MCU with a direct memory access controller (DMA) can automatically sample the A/D converter and move the data to memory. All MSP430 peripherals are self-managed and can be preconfigured for power saving. For example, an A/D conversion can be event-triggered using a timer, and the data is automatically stored in a memory buffer. When the conversion and data storing are ready, the A/D converter triggers an interrupt so the CPU can process the data. The power optimization of the MCU subsystem must be accompanied by a low-power radio transceiver with high RF performance.
The RF transceiver must be carefully selected for low power consumption, high functional performance, and minimum overall system cost. Commonly used power-saving techniques at the radio transceiver part are briefly reviewed.
Low power consumption in active mode. The radio transceiver is by far one of the most dominant contributors to the active power budget in a ZigBee implementation. Low power consumption by the RF, analog, and modulation/demodulation circuitries while transmitting and receiving data is therefore a must. Driving ZigBee silicon to the power performance and cost level necessary for mass marketing of sensor networks requires pushing the channel filtering function on-chip using a single-conversion receiver architecture at a conveniently low or even zero IF.
Efficient generation of the transmit signal according to the IEEE 802.15.4 PHY can be achieved by using single-step I/Q-upconversion, which provides excellent performance and is extremely flexible with respect to data rates and modulation formats. The O-QPSK with half-sine pulse shaping modulation selected for the 2.4-GHz PHY is a constant envelope modulation scheme allowing the use of simple, low-cost, and relatively nonlinear power amplifier designs.
Enhanced link margin and minimum retransmission. With the relatively unpredictable wireless space, low-power networking can be further improved by increasing the receiver sensitivity and selectivity for robust signal link and minimum amount of data retransmission.
Early discard of unwanted data. Early detection of false data packets, for example, by packet address and length filtering or CRC, can reduce the overall power consumption by minimizing the useless active operation time (quickly power down the receiver and go back to sleep mode).
Minimum accessory periods. Short startup and calibration times of the on-chip circuitries are also effective means to avoid unnecessary power usage.
The proliferation of wireless communication devices shows no sign of stagnation. Low-power wireless networks will emerge in countless applications with the ability to increase security and comfort in our daily life and entail significant commercial potential. Power efficiency is one of the decisive factors for the viability of these networks. ZigBee, with its foundation in the robust IEEE 802.15.4 standard, has the technical qualities and industry support to cost-effectively deliver low-power wireless solutions. Design techniques and commercial semiconductor devices are available today to realize the envisioned ubiquity of low-power wireless networking.
1. IEEE std. 802.15.4-2003: Wireless Medium Access Control (MAC) and Physical Layer (PHY) Specifications for Low Rate Wireless Personal Area Networks (LR-WPANs).
2. See www.zigbee.org.
Khanh Tuan Le is an RF-IC systems architect at Texas Instruments (www.ti.com), Low Power Wireless Division, working on CMOS radio transceiver and system-on-chip devices for wireless applications. He can be reached at k.t.le@ti.com.
Juan Alvarez is marketing manager for catalog MSP430 microcontrollers at Texas Instruments. In this capacity, he is responsible for strategic product definition and positioning, new customer design engagements, product and business development, and customer support.
