Research On Barrier Coverage In Directional Sensor Networks

As one of the subproblem of the converage problem,barrier coverage aims at detecting intruders that attempt to cross the region of intrest,which is widely used in military affairs,industry and agriculture automation,environment monitoringnand other fields.There are extensive number of studies in recent years about the barrier coverage problem in traditional sensor networks.Directional sensors do mainly include video sensors,infrared sensors,and ultrasound sensors are widely used in barrier coverage,especially with the increasing number of wireless multimedia sensor network(WMSN)applications.Different form traditional sensor,directional sensor is characterized by its directional sensing range,which requires specific solutions and techniques for coverage enhancement.In this thesis,several important problems in barrier coverage in directional sensor networks are studied.The main work contributions are summarized as follows:(1)The relationship between the deployment strategy and the existence of bariier coverage in direntional sensor networks is studied.There are two main kinds of deployment strategies:LNRO and Poisson Point process.First,a tight lower bound for the existence of barrier coverage under theses two deployment strategies are anlylized,by establishing mathematical models.The simulation examinations analyse the relationship between the network parameters and the the existence of barrier coverage,and compare the quality of barrier coverage of sensor deployment under LNRO with that under Poisson Point process in the simulation examination..The results demonstrate that sensor deployment strategies have direct impact on the barrier coverage of direntional sensor networks.And the simulation results will provide important guidelines to the deployment and performance of direntional sensor networks for barrier coverage.(2)It has been been shown that a large fraction of sensors do not contribute to barrier coverage,when they are redoundly deployed.An energy efficient barrier coverage algorithm is proposed to maximize the network lifetime of a directional sensor network.First,the algorithm introduces a directed graph to model barrier coverage with directional sensors.Based on this graph model,multiple barrier sets are found.Then,a Linear Programming problem is used to maximize the total work time of all the barrier sets,subjected to the constraint that all directions of any sensor work across all the barrier sets for no more than the initial lifetime of the sensor.Simulation results are presented to demonstrate the performance of the algorithm.(3)It has been shown that barrier coverage is difficult to achieve when the network are randomly deployed with stational sensor nodes,because when the stational nodes are deployed,they can't move.It studies how to efficiently improve the barrier coverage using mobile sensors to mending the barrier gaps.Two barrier coverage mending algorithm is proposed to slove this problem:The Minimum Gaps Barrier Repair Algorithm and The Minimum Moving-distance Barrier Repair Algorithm.The Minimum Gaps Barrier Repair Algorithm is divided into two steps:first,it finds the minimum gaps barrier by modeling a directed graph for the static sensors in the network;then the mobile sensor which has the shortest moving distance is chosen to repair the gap which is found before.The Minimum Moving-distance Barrier Repair Algorithm is also divided into two steps:first,it compute the maximum flow of the flow graph;then it choose the flow which has the minimax moving distance to work for the sensor network.The extensive simulations demonstrate that the algorithms improve thee barrier coverage and plong the network lifetime.(4)The barrier coverage problem when there are no sufficient mobile sensor nodes to guarantee full barrier coverage is considered.Mobile sensors are able to antonomously improve barrier coverage performance by their mobility and motility.Two algorithms are designed to solve this problem:Distributed Dynamic Patrolling(DDP)algorithm and Dynamic Monitoring with High Probability(DMHP)algorithm.First,the target area is devided into multiple subareas,and each area can be covered by only one sensor.In DDP algorithm,each subarea along the barrier line is monitored periodically by mobile sensors.In DMHP algorithm,the probability of intruder arrival at each subarea is calculated dynamically,and the mobile sensor moves to the subarea which has high probability.The simulation results demonstrate that in DDP algorithm,the detect probability only has impact on the number of mobile sensors,and the DMHP algorirhm performance better than the DDP algorithm.