Analysis On Zigbee Link Wireless Channel State Based On Software Defined Radio

Compared with wired communication, a wireless link is more easily affected byenvironment changes or surrounding wireless activities. In previous studies, most ofworks used metrics like Received Signal Strength Indication (RSSI), Signal-to-NoiseRatio (SNR), and Signal-to-Interference plus Noise-Ratio (SINR) which werecalculated at the packet level to reflect the channel conditions. However, such metricsare based on the entire packet, which may not be capable of capturing the variationduring the whole packet transmission. It is of great significance to analyze morefine-grained information from the lower layer of the network protocol stack.Zigbee standard specifies physical layer (PHY) and medium access control(MAC) sublayer protocols for low-rate and low-power communication applications.In this protocol, every4-bit symbol is encoded into a sequence of32chips that areactually transmitted over the air. The32chips as a whole is also called a pseudo-noisecode (PN-Code), Due to complex channel conditions such as attenuation, multipath,and interference, the transmitted PN-Code will often be received with some PN-Codechips corrupted. Since the chip duration is much shorter, traditional measurementmetrics, such as RSSI,SNR and SINR could not reflect the change of the erroroccurring at chip-level. The chip error information can provide a new basis for thelink state determination.With the realization of Zigbee communication protocol based on open sourcesoftware radio (GNU Radio) device, this paper records the PHY chip error collectedin different channel conditions and conducts a systematic analysis on the relationshipbetween the chip error and the channel conditions with the help of artificialintelligence classification algorithm. According to the characteristics of different chiperrors, the paper could determine the channel condition, such as attenuation, multipathand interference, and those judgments could be instructions for the designs and operations of the upper-layer protocols.The main work of this paper is as follows:1) With the help of software radio equipment, the communication process ofZigbee protocol is implemented. PHY information of Zigbee and the chip errorpattern are obtained by utilizing the USRP device, and thereby the percentages anderror indicators of different chip error patterns are calculated.2) Attenuation, multipath and interference communication conditions aresimulated to obtain the chip error through USRP device, and a systematic analysis isconducted by using artificial intelligence classification algorithm, such as J48Random Forests and Naive Bayesian. By experiments, it is found that whencalculating with error indicators of different chip error patterns, J48is the bestmethodology and the precision is86.67%. Meanwhile, if calculating with percentagesof different chip error patterns, Naive Bayesian is best and the precision is around80%.