Research On Subset-Based Node Scheduling Algorithm For Wireless Sensor Networks

In recent years, with the rapid development of wireless network technology, wireless sensornetworks have attracted wide attention. The network is composed of vast sensors with limitedcapacity of perception and transmission. Because of the energy constraints, how to schedulenodes to prolong the lifetime of network as long as possible becomes the most important thingsin the design of wireless sensor networks. It means that the networks don't need all the sensorsworking on the entire lifetime of the system, some nodes can be converted to a sleep mode, solong as to ensure normal operation of sensor networks. So, the life of network can be extendedby scheduling nodes by turns.In this paper, the architecture and typical application of wireless sensor networks are firstlydescribed, the difficulties of the design of the node scheduling strategy are discussed, and thecurrent research situation of node scheduling strategy is introduced. Then, we proposed two nodescheduling solutions: a grid-based node scheduling algorithm and an order-based nodescheduling solution based on quantum genetic algorithm. The grid-based node schedulingapproach is based on the relationship of coverage, sensor range, number, and so on. On this basis,the maximal number of cover sets is decided. Then, the approach controls the nodes to work byturns using grid. Experimental results and theoretical analysis demonstrate that the proposedalgorithm has the characteristics of less calculation and short runtime, and can extend thenetwork lifetime effectively. The approach based on quantum genetic algorithm is studied for thepoint coverage problem in wireless sensor network. The situation of cover sets is expressed by anorder in the proposed algorithm, so that the number of cover sets doesn't need to be pre-set.Besides, the chromosome of algorithm is encoded by quantum bit, the population has thecharacteristic of better variety. The population is updated by quantum revolving door to avoid theconvergence rate of the algorithm is too fast, and find the better value of k. Experiment resultsshow that the proposed algorithm has the advantage of lager search range, better adaptability,higher efficiency and better effect.