Research On Barrier Coverage Algorithms In Wireless Sensor Networks

Barrier coverage is a typical coverage model in wireless sensor networks which aims at detecting intruders that attempt to penetrate the monitored area.In barrier coverage model,sensor nodes are deployed in a trip area to form a chain of overlapping sensing ranges.Barrier coverage has critical applications of intruder detection in military and homeland security.The current researches on barrier coverage can be categorized into three aspects.(1)How to build barriers efficiently and effectively.(2)How to schedule the barriers to prolong the lifetime of the network.(3)How to refill the gaps with the minimum cost after a period of running.The main contributions of this thesis are:(1)An ant colony based barriers constructing algorithm is proposed to achieve k-barrier coverage with only stationary sensor nodes in the monitored area.The whole monitored area is divided into k sub-regions in which the improved ant colony algorithm is applied to form a barrier.Besides,there is overlapping between the adjacent two sub-regions.The nodes in the overlapping area can be utilized by both the two sub-regions.(2)An intrusion track prediction model based segmented scheduling method is presented.Only specific segments of the barriers are activated and the other parts are in the sleep state to save energy.This method can significantly prolong the lifetime of the network with a certain predefined detecting ratio.(3)An maximum flow based barrier gap refilling algorithm is proposed.Mobile sensor nodes are utilized to refill the gaps with a minimized moving distance.Some experiments are conducted,and the results show the good performance of the proposed methods.The iterations in the proposed barrier construction algorithm are only 1/4 of the traditional ant colony algorithm.The lifetime can be promoted by 60% after applying the proposed segmented scheduling algorithm with the detecting ratio of 90%.Much more gaps can be refilled with the proposed gaps refilling algorithm than the traditional greedy algorithms.