Research Of Wireless Sensor Network Acoustic Target Detection Algorithm In Noise Environment

Ground reconnaissance is an important part of military reconnaissance.Ground reconnance has the ability of detecting the target continually in complexterrain. Ground reconnaissance system can work effectively in complex terrain,which is the things that radar, optics, Infrared reconnaissance and other moderndetection technology can not accomplish. Wireless Sensor Network has a bigapplication value in ground reconnaissance, especially in some long-termunattended areas, which is unfit for human (for example, environmental damagedareas, high risk areas and desolate field). Terrestrial Wireless Sensor Networkscan play its strengths to them. It can generate sound when the object moving inthe ground. According to sound physical quantity, ground reconnaissance systemcan determine whether the target appears or not.In this paper, Acoustic target detection in Wireless Sensor Networks is inlarge noisy environments, and it uses Wireless Sensor Network platform, throughnodes collect sound signal datas, and these datas are processed, which canachieve the detection of specified target (car) in certain regain accurately. Thisapplication will make a contribution to regional alert.The topic analyzes acoustic signal detection algorithms on a single node indetail. Study the performance of past detection algorithm based on signal’senergy, and give an improved energy detection algorithm based on theirweakness. The improved method can decide the threshold easily, and it can adjustthe threshold according to monitoring changes of environment automatically. Inaddition, the improved method sets the secondary threshold, which suppressesthe interference effect of the verdict effectively. In Wireless Sensor Networks,there are a large number of wireless sensor nodes, and multiple nodes canimprove the detection accuracy through a joint judgment. At this point, we havestudied joint decision method in detail, and we give an improved approach basedon the conventional method of joint decision. Another important aspect of thisthesis is building a platform of the wireless sensor detection system and theselection of functional implementation. The system consists of a control center(PC), a coordinator node and a number of common monitoring nodes. Monitoringcenter and the coordinator node use the USB to communicate, while coordinatornode and the common monitoring nodes of area use a short-range wirelesscommunication technology to achieve data exchange.To assess the performance of the algorithm, we use experimentalplatform to do target detection for specific areas in a real environment.Compared the experimental results with the conventional method results, we can test the effectiveness of the algorithm presented in this paper.