Introduction to Wireless Sensor Network Hardware Design
Because of its huge application prospect, wireless sensor network has drawn more and more attention from academia and industry . This paper introduces the architecture of wireless sensor network nodes, analyzes and compares the current typical hardware platforms at home and abroad, focuses on the advantages and disadvantages of processors, radio frequency chips, power supplies and sensors commonly used in wireless sensor network nodes, and compares them in detail Wireless communication technology currently used in wireless sensor networks .
Key words wireless sensor network hardware platform low power consumption wireless communication
introduction
WSN (Wireless Sensor Network) is a network composed of sensor nodes, which can monitor, perceive and collect various information (such as light intensity, temperature, humidity, etc.) , noise and harmful gas concentration and other physical phenomena), and the information is processed and sent wirelessly, and finally sent to the observer through the wireless network. Wireless sensor networks have broad application prospects in military reconnaissance, environmental monitoring , medical care, smart home , industrial production control, and commerce.
In the sensor network, the sensor node has the functions of end nodes and routing: on the one hand, it realizes data collection and processing; on the other hand, it realizes data fusion and routing, and integrates the data collected by itself and the data sent by other nodes. , forward the route to the gateway node. The number of gateway nodes is often limited, and often the energy can be supplemented; the gateway usually uses multiple methods (such as the Internet, satellite or mobile communication network , etc.) to communicate with the outside world . However, the number of sensor nodes is very large, and batteries that cannot be replenished are usually used to provide energy; once the energy of a sensor node is exhausted, the node cannot perform data collection and routing functions, which directly affects the robustness and life cycle of the entire sensor network. Therefore, the sensor network mainly studies the sensor network nodes. The specific application is different, the design of the sensor network node is not the same, but its basic structure is the same. A sensor network node generally consists of four parts: a processor unit, a wireless transmission unit, a sensor unit and a power module unit , as shown in Figure 1.
Figure 1 Typical composition of wireless sensor network nodes
1 Typical nodes of wireless sensor network
As a miniaturized embedded system , the sensor network node constitutes the base layer support platform of the wireless sensor network. Because most of the wireless sensor networks are powered by batteries, the working environment is usually harsh, and the number is large, and it is very difficult to replace the batteries. Therefore, low power consumption is one of the most important design criteria for wireless sensor networks. From the hardware design of wireless sensor network nodes The protocol design of each layer of the entire network regards energy saving as one of the design goals, and prolongs the life of the wireless sensor network as much as possible.
Due to different specific application backgrounds, there are many hardware platforms for wireless sensor network nodes at home and abroad. Typical nodes include Mica series, Sensoria WINS, Toles, μAMPS series, XYZnode, Zabranet, etc. In fact, the main difference between each platform is the use of different processors, wireless communication protocols and different sensors related to applications. Commonly used wireless communication protocols include 802.11b, 802.15.4 ( ZigBee ), Bluetooth, UWB and custom protocols; processors range from 4-bit microcontrollers to high-end processors with 32-bit ARM cores. There is another type of node that uses a single-chip microcomputer integrated with a wireless module , typically WiseNet. Typical wireless sensor network nodes are listed in Table 1.
This paper introduces the concept and characteristics of wireless sensor networks and the composition of wireless sensor network nodes, focuses on analyzing and comparing the characteristics of various commonly used chips in each component unit of the node, and always regards low power consumption as one of the important criteria for comparison.
2 Comparison of Typical Wireless Sensor Network Nodes
At present, researchers at home and abroad have developed a variety of wireless sensor network nodes. In fact, the components of these nodes are similar, but their application backgrounds are different, and the requirements for node performance are also different. Therefore, the hardware components used are different. big difference.
2.1 Processor unit
The processor unit is the core of the sensor network node, and completes data collection, processing and sending and receiving together with other units. The EM6603 is a 4-bit microcontroller with very low power consumption but also very limited processing power. Most of the Mica series nodes developed by Berkerly University adopt Atmel 's microcontroller. Among them, the Mica2 node adopts Atmel enhanced microcontroller ATmega128L. The microcontroller has rich on-chip resources, including 4 timers , 4 KB S RAM , 128 KB Flash and 4 KB EEP ROM , and has UART , S PI , I2C , JTAG interfaces to facilitate the access of wireless chips and sensors; There are 6 power saving modes, which are convenient for low power consumption design. Another advantage of using this processor is: There are many compilers, among which GCC (WINAVR) is completely free and open software. Due to the above advantages and the influence of Mica2 nodes, it is widely used in actual wireless sensor design. But from the point of view of low power consumption, the chip is not the best choice.
As listed in Table 1, in terms of low power consumption, the MSP430 F1xx MCU family offers the industry’s lowest current consumption with an operating voltage of 1.8 V, real-time clock standby current consumption of only 1.1 μA, and run mode current as low as 300 μA (1 MHz), the entire wake-up process from sleep to normal operation takes only 6 μs. PIC series microcontrollers also have products with low power consumption. The Toles node and the ZebraNet node use MSP430 series microcontrollers with very low power consumption. In some applications with a large amount of data, high-end processors are also used. For example, the μAMPS1 node adopts the StrongARM processor SA1110, and the power consumption is 27-976 mW. The processor supports DVS energy saving, which can reduce power consumption by about 450 mW; turning off the wireless module can reduce power consumption by 300 mW. The processor used in μAMPS2 is DSP . The processor used by XYZnode is ML67Q5002 with ARMTDMI core of OKI Company, which also supports DFS (Dynamic Frequency Adjustment), the working current is 15-72 mA, and the frequency is 1.8-57.6 MHz.
Table 1 Typical wireless sensor network nodes
From the perspective of processors, wireless sensor network nodes can basically be divided into two categories: One category uses high-end processors represented by ARM processors. The energy consumption of this type of node is much higher than that of a microcontroller, and most of them support energy-saving strategies such as DVS (Dynamic Voltage Scaling) or DFS (Dynamic Frequency Scaling), but their processing capabilities are also much stronger, suitable for applications with high data volume services such as images ; In addition, it is also a good choice to use a high-end processor as a gateway node. The last three processors in Table 2 are processors with ARM cores, and their power consumption is significantly higher than that of low-end microcontrollers. The other category is nodes represented by low-end microcontrollers. The processing capability of this type of node is weak, but the energy consumption power is also very small. When choosing a processor, you should first consider the system's needs for processing power, and then consider power consumption.
Table 2 Performance comparison of various common microcontrollers
2.2 Wireless transmission technology and chip
Available transmission media include air, infrared , laser, ultrasonic, etc. Commonly used wireless communication technologies include: 802.11b, 802.15.4 (ZigBee), Bluetooth, UWB, RFID , IrDA, etc.; Defined by the user, these chips generally work in the ISM free frequency band, as listed in Table 3. Using laser as a transmission medium has lower power consumption than electromagnetic waves and is safer. The disadvantages are: It can only be transmitted in a straight line; it is easily affected by atmospheric conditions; the transmission is directional. These disadvantages make it not an ideal transmission medium. The transmission of infrared rays is also directional, the distance is short, and no antenna is needed . The chip 83F88S is a wireless transceiver chip conforming to the IrDA standard. UWB has the advantages of low power spectral density of transmitted signals , low system complexity, insensitivity to channel fading, good security, high data transmission rate, and positioning accuracy of several centimeters; the disadvantage is that the transmission distance is only about 10 m, and the partition wall Penetration is not good. 802.11b is not widely used because of its high power consumption. Bluetooth works in the 2.4 GHz frequency band, and the transmission rate can reach 10 Mbps; the disadvantage is that the transmission distance is only about 10 m, and the complete protocol stack is 250 KB, which is not suitable for low-end processors. It is used in the family personal wireless local area network, and it is also used in the wireless sensor network. ZigBee and common radio frequency chips are the most widely used in wireless sensor networks. ZigBee is a short-distance, low-complexity, low-power, low-data-rate, low-cost two-way wireless communication technology. The complete protocol stack is only 32 KB, which can be embedded in various devices and supports geolocation. The above characteristics determine that ZigBee technology is very suitable for application in wireless sensor networks. Chip manufacturers currently on the market that support the ZigBee protocol include Chipcon and Freescale Semiconductor .Company, Figure8 company also specially developed the ZigBee protocol stack. Chipcon's CC2420 chip is widely used, and both the Toles node and the XYZ node use this chip; Chipcon provides a complete development kit including the ZigBee protocol developed by Figure8. The 2.4 GHz wireless transmission chips of ZigBee provided by Freescale Semiconductor Company include MC13191, MC13192, and MC13193; the company also provides supporting development kits.
Table 3 Wireless communication technologies applied to wireless sensor networks
Ordinary RF chips are also an ideal choice, and the communication protocol can be customized. Representative MAC protocols include TMAC, SMA, CWiseMAC, BMAC, DMAC , etc. Routing protocols include Gossiping, SPIN protocol, LEACH protocol, TE EN protocol, etc. Considering performance, cost, and power consumption, RF M's TR1000 and Chipcon's CC1000 are ideal choices. These two chips have their own strengths. TR1000 has lower power consumption, while CC1000 has higher sensitivity and longer transmission distance. WeC, Renee and Mica nodes all use TR1000 chip; Mica2 uses CC1000 chip; Mica3 uses CC1020 chip of Chipcon Company, the transmission rate can reach 153.6 kbps, and supports OOK, FSK and GFSK modulation modes; Micaz node uses CC2420 ZigBee chip. There is also a type of wireless chip that integrates the processor itself. For example, CC2430 is a single-chip microcomputer integrated with 51 cores based on CC2420; CC1010 is a single-chip microcomputer integrated with 51 cores based on CC1000, which further improves the integration of the chip. WiseNet nodes use the CC1010 chip. Common wireless chips include Nordic's nRF905, nRF2401 and other series chips. Because of high power consumption, low receiving sensitivity, and difficult development, they are rarely used in actual wireless sensor networks. The main parameters of commonly used wireless chips are compared as listed in Table 4.
Table 4 Comparison of main parameters of commonly used wireless chips
2.3 Power Module
There are many types of batteries, and the size of the battery energy storage is related to factors such as shape, diffusion speed of active ions, and selection of electrode materials. The battery of the wireless sensor network node is generally not easy to replace, so the selection of the battery is very important, and the efficiency of the DC DC module is also very important; in addition, the energy of the nature can also be used to supplement the energy of the battery.
According to whether it can be charged, batteries can be divided into rechargeable batteries and non-rechargeable batteries; according to electrode materials, batteries can be divided into nickel-chromium batteries, nickel-zinc batteries, silver-zinc batteries, lithium batteries, lithium polymer batteries , etc. Generally, non-rechargeable batteries have higher energy density than rechargeable batteries. If there is no source of energy supply, non-rechargeable batteries should be selected. Among rechargeable batteries, lithium batteries and lithium polymer batteries have the highest energy density, but the cost is relatively high; nickel-manganese batteries and lithium polymer batteries are the only rechargeable batteries that are not toxic. The performance parameters of common batteries are listed in Table 5. Wireless sensor network nodes generally work outdoors, and can use natural energy to replenish battery energy. The energy available in nature includes solar energy, electromagnetic energy, vibration energy, nuclear energy and so on. Since the number of rechargeable batteries is limited, and most rechargeable batteries have a memory effect, the use of natural energy cannot frequently charge the battery, otherwise the service life of the battery will be greatly shortened.
Table 5 Performance parameters of common batteries
2.4 Sensor Module
There are many types of sensors, which can detect physical quantities such as temperature and humidity, light, noise, vibration, magnetic field, and acceleration. US Cross sbow company has developed a series of sensor boards based on the Mica node. The sensors used include photoresistor Clairex CL94L, temperature-sensitive resistor ERTJ1 VR 103J ( Panasonic Electronics ), acceleration sensor ADI AD XL202, magnetic sensor Honeywell HMC1002, etc. The SHTxx series of temperature and humidity sensors can support low power consumption mode, and automatically switch to sleep mode after collecting data, and the current is less than 1μA.
The power supply circuit design of the sensor power supply is very important for the energy consumption of the sensor module. For sensors working with small currents (hundreds of μA), they can be directly driven by the processor I/O port; when the sensor is not used, set the I/O port as an input mode. In this way, the external sensor has no energy input, so there is no energy consumption. For example, the temperature sensor DS18B20 can adopt this method. For the sensor module working with high current, the I/O port cannot directly drive the sensor, and a field effect transistor (such as Irlm16402) is usually used to control the energy input of the subsequent stage circuit. When multiple high-current sensors are connected, an integrated analog switch chip is usually used to realize power control, and MAX4678 is such a chip.
3 Conclusion
Due to different application backgrounds, there are many hardware platforms at home and abroad, and there are many wireless communication technologies used. This paper mainly summarizes the current common wireless sensor network hardware platforms, analyzes and compares commonly used processors, wireless chips, wireless communication technologies, sensors and power supplies, and always takes power consumption as one of the important comparison factors considered. Through the detailed analysis of the wireless sensor network hardware platform, it is expected to play a positive role in the research and development of wireless sensor network in our country.