Research On Path Planning Algorithm Of Wireless Rechargeable Sensor Network Based On UAV

Wireless sensor network is a kind of self-organized network composed of several sensor nodes,which is widely used in smart home,military application and environmental monitoring and many other files.Each node of the network needs enough energy to complete the task of data collection,calculation,storage and transmission.Once the energy is exhausted,the node will not work as usual,and then even the whole sensor network will be out of service.Because of the small size of sensor nodes and the low capacity of batteries,energy supplement has been an important issue for the development of wireless sensor network.In order to solve this problem,on the one hand,the existing work studies the high-power-capacity materials of the battery and the methods to reduce the energy consumption speed of nodes.On the other hand,considering energy supplement for sensor nodes from the outside is also a hot issue.With the progress of wireless charging technology,the strategy of using mobile charger to replenish energy of sensor nodes has been paid more and more attention.Taking the need of charging all nodes successfully as a prerequisite,the direction of our research is to study and improve the appropriate charging path plan of mobile charger,just to make the best performance.In this paper,a single mobile charger is used to charge wireless rechargeable sensor networks of different scale with different methods.This process is of moving also summed up as a Traveling Salesman Problem.In particular,the paper points out,different from the "one-to-one" charging mode in the previous research,a single mobile charger can charge multiple sensor nodes at the same time.The paper limits the research on the path planning algorithm to be applicable for the "one to multiple" mode,and uses UAV to replace the original car as the mobile charger.First of all,this paper will build a model and calculate the "optimal height" of UAV when only considering "charging" model in the previous study.Secondly,in order to solve the NP hard problem of the best charging path,the method of calculating the fitness value in the original genetic algorithm is adjusted to make it suitable for the "one to multiple" charging mode and find a suitable charging spot.As a result,the algorithm is improved.Then,in this paper,limited energy of the mobile charger is a question,and we will also take it into consideration for large scale of networks.Instead of the distance between two nodes,we take the number of charged nodes during per unit time as apriority,we implement a feasible charging path plan for large-scale sensor networks.Finally,considering the "move-charge" mode,the experiment is used to explore the charging performance of UAV at different heights,and the performance is compared with the previous " optimal height ".Finally,after performance evaluation,the improved algorithms in the paper has better performance.Compared with the time spending on charging task of original methods,the improved genetic algorithm can reduce about 9.4%,and the improved greed algorithm can reduce about 11.2%.At optimum height of experimental result,the task completion time can also reduce 8.6%.