Research On Key Technologies Of Wireless Sensor Network Deployment In Restrict Environment

The extending application of the wireless sensor network (WSN) in restricted environments, suchas indoor space, pipeline, road network, underwater, battleground and so on, has drawn plentyscholars’ attention to the research of the deployment technology of WSN in such restricted scene. Dueto the differences between the influences in imparity environments, it’s necessary to analyze theenvironmental impact on sensors first while considering locations to deploy the WSN in order tosatisfy the certain requirement of coverage and connectivity.On the basis of existing research, we particularly studied the deployment problem of WSN inthree-dimensional indoor space and pipeline monitoring application. The main results of our workachievements are summarized as follows:(1) We briefly analyze the defects of applying existing research on deployment problem of WSNbased on two-dimension space to three-dimensional indoor space and modeling environmentalimpacts by LoS. We design a static WSN deploy mechanism distinguishing the importance ofdifferent area by weight division and modeling the environmental impacts by log-normal shadowingmodel with dynamic variance (LNSM-DV) instead of LoS. A greedy-based heuristic algorithm isproposed to solve the problem. And then, a weight ware genetic algorithm is presented to get theglobal optional solution.(2) We point out the defects of existing research on static WSN deployment in pipeline monitoringapplication and applying traditional dynamic research on it. Then we propose an ICLA-baseddynamic deploy mechanism for pipeline monitoring application to make sensors spread autonomouslywithout knowing the structure of the pipeline network. The algorithm separates pipeline network intosegments called sub-area. Then the voted host node will guide the movements of sensors incorresponding sub-area. Simulator results show that the algorithm won the best k-coverage rate whenthe nodes could not cover the pipeline completely.(3) We introduce the core idea of NHTrack system which designed for symbolic positioning andbehavior detection of indoor moving objects. The specific model is presented to describe how todeploy sensors in given indoor scene. Then we describe the key data structure and functional designof the implementation of NHTrack. Experiments on real nodes prove the availability of this system.