Research On Energy Optimization Algorithms For Wireless Sensor Networks Based On Swipt

Nowadays wireless sensor network system(WSN)has broad application prospects,but one of the main bottlenecks restricting its development is that the working life of sensor nodes is restricted by energy.This is mainly because sensor nodes are usually very small in size,and the corresponding batteries are relatively small,the battery power is limited,and charging is difficult.SWIPT(simultaneous wireless information and power transfer)is to use radio frequency signal to transmit energy while carrying information.After receiving radio frequency signal,the sensor nodes can decode part of the signal through certain conversion,and the other part can be stored by energy harvesting,which can greatly prolong the life of the sensor nodes and fundamentally solve the problem of energy limitation of sensor nodes.Nowadays,the researches on SWIPT mainly focuses on the downlink beamformers and PSFs(po we splitting factors).A small number of works have to date taken the uplink design into account and only some HTT(harverst-then-transmit)systems optimized the time allocation where only energy is transferred in the downlink.In order to meet the actual situation,this paper also considers the joint optimization of downlink and uplink in wireless sensor network system.However,with the rapid development of wireless communication,the available spectrum resources of wireless channel are becoming less and less.Therefore,in this paper,we also consider how to improve the spectrum utilization of the wireless channel.On this basis,we build a charging system of the MISO(Multiple-Input Single-Output)wireless sensor network,which is of great significance.In this paper,a wireless sensor network system with an energy harvesting is mainly discussed.A total mean-square-error(Total-MSE)of both downlink and uplink minimization problem with multiple constraints is formulated,since it considers both downlink and uplink Total-MSE and solve more coupled variables when optimizing the reader beamformer.Then to optimize the beamformer,power allocation and time allocation jointly.In order to solve the original non-convex optimization problem,two different alternative iterative algorithms are proposed:1)optimization algorithm based on semi-definite relaxation;2)sub-optimization algorithm based on Lagrangian duality.The two algorithms have their own advantages.The first one has higher performance,but its computational complexity is high.The another one has a slightly worse performance,but its computational complexity is low.Finally,this paper uses the simulation software of MATLAB to simulate the optimization problem and the corresponding solving process.Then,the simulation results are analyzed.The corresponding energy optimization algorithm proposed in this paper is verified to improve the system performance of sensor networks.The combined simulation results summarize the effectiveness of each algorithm in the end.