Study On Target Identificaion Method With UWB-IR Based On Fuzzy Pattern Recognition And Genetic Algorithm

UWB signal has a broad spectrum of GHz, high resolution and strong detection ability; therefore UWB is potential in communication, targets identification, distance measuring and positioning areas. UWB technologies include UWB communication technology and UWB radar technology. UWB communication technology is mainly applied in the wireless sensing network, domestic high speed data transmission etc. UWB radar technology is mainly used in targets identification, such as landmines detection, through-wall imaging and etc. From the targets identification perspective based on UWB technology, there are two ways to complete it. One way utilizes UWB WSN and it can be used to detect and monitor targets, but it usually requires the wireless communication nodes as well as large and heavy sensors. The equipment is expensive and power consumption is large, so this is not reasonable for the battery-powered sensor nodes. On the other hand, UWB radar or UWB radar networks uses echo information for target detection. UWB radar can be very effective to complete the target identification, but it cannot complete forward communication and it needs the help of other communication technology to complete it.From the UWB communication perspective, to realize forward communication and target identification with no special sensors at the same time, a method of target identification with UWB-IR based on GA (Genetic Algorithm) and Fuzzy Pattern Recognition is introduced. This method is different from the traditional radar and parameters related to targets information are extracted from the received signal. According to GA and Fuzzy Pattern Recognition, the target prediction function is built based on the parameters and the maximal membership principle is used to identify the targets. Firstly four kinds of UWB targets identification scenarios are simulated in different SNRs by FDTD (Finite Difference Time Domain) and the scenarios are los scene, wood scene, glass scene, concrete scene. Parameters of signal energy and excess delay extracted from the received signals are used for targets identification. The simulation results demonstrate that the method is effective to identify targets. Finally, this method is applied to actual targets identification scenarios which include a scene with no target, a scene with wood board target and a scene with iron cabinet target. Nine parameters including the maximum amplitude of the received signal, the excess delay, the RMS delay, the number of the multipath components (MPCs) that capture85%of the received energy, the number of MPCs within10dB compared with the largest peak path, the signal energy, the standard deviation of the received signals, the kurtosis and the Skewness are extracted from the received signals and used for targets identification. The simulation results demonstrate that it is also effective in actual scenarios.