Research On Energy Optimization For Data Gathering In UAV-aided Networks

The development of the Unmanned Aerial Vehicles(UAV)make it possible to collect data from Wireless Sensor Networks by the UAV.The WSNs which using the UAV as the sink node is UAV-aided Networks,sensor nodes and the UAV is powered by the batteries in the UAV-aided networks,energy efficient is critical for improving the lifetime of the network.Data transmission and flight distance contributes majority of energy consumption of sensor nodes and the UAV.An efficient data gathering scheme and a UAV trajectory planning scheme which is aimed to data gathering are critical for improving the lifetime of the network.Both clustering and Compressive Sensing can be applied for saving energy of the network.Only the cluster head are able to communicate with the UAV in the UAV-aided networks,the heterogeneity of the sensor nodes requires the network must be clustering.Most of the existing works focus on the selection of cluster head,the clustering process only considers the distance,and the character of the data has been ignored.Although few works have integrated CS with clustering,they ignored the sparsity difference in temporal and spacial domain,and cannot decrease the energy consumption of the network efficiently.The object of the data gathering oriented trajectory planning for the UAV is minimum the energy consumption while maximum the amount of collected data,and the special mobility of the UAV requires the curvature of the trajectory must be continuous and bounded,and it is different the trajectory planning of other applications.In order to solve the above problem,this thesis researches on the data gathering techniques of the UAV-aided networks and achieves the following results:(1)The data gathering scheme that integrates CS and clustering.The problem that integrates CS and clustering for data gathering has been formulated to a mixed integer programming problem,the NP-hardness of the problem has been proved.By jointly consider the effect of the compressive ratio variation of a cluster and the distances variation between cluster members and the CH in the proposed greedy algorithm.And an advanced scheme has proposed to maintain the low energy consumption as well as decrease the computational complexity.The simulation result on the real data trace shows that the proposed schemes have good performance in energy consumption and computation complexity.(2)The data gathering oriented trajectory planning for the UAV.Theoretical analysis of the trajectory planning of the UAV is carried out,and has proposed Particle Swarm Optimization based trajectory planning(PSOTP)scheme.Although the PSOTP can find the near-optimal solution of the problem,the computational complexity of this scheme is too high for the practical application.A computational efficiently heuristic trajectory planning scheme(HTP)has been proposed to decrease the computational complexity.By consider the curvature and amount of the collected data in the generating and choosing process of the trajectory,the constraints of the UAV trajectory planning is satisfied.Simulation results show PSOTP and HTP have similar performance in trajectory length,but the running time of HTP is far less than PSOTP.