Research On Data Transmission Cross-layer Optimization Algorithm Based On Compressive Sensing

In recent years,with the development of semiconductor technology,microsystem technology and communication technology,cyber-physical systems(CPS)have become important research field.Wireless sensor networks(WSNs),which is an important part of CPS,has been attracting much attention.Because the sensor node has the characteristics of high communication energy consumption and low energy consumption of data processing,the data transmission becomes the research hotspot of wireless sensor networks.In a network,data flow through large and not timely processing can lead to congestion,which can increase packet delay.The allocation of link capacity is also a key technology.Too much resource can not be fully utilized,which will cause unnecessary waste of energy.Data cannot be effectively transmitted over little of resources,and network throughput will be severely reduced.Routing selection directly affects network load balancing,and energy management plays a vital role in prolonging the network lifetime.It has become an important aspect of data transmission technology to effectively reduce data transmission,reasonably allocate link capacity,furthest select route and reduce energy consumption and achieve network load balance in WSNs.The thesis mainly studies data transmission problem based on compressed sensing technology of WSNs,emphatically from the sampling,routing protocol design,node scheduling strategy and network performance analysis for further research.The new mechanism of data processing and transmission is discussed.The main research contents are summarized as follows:Firstly,the compressive sensing and matrix completion theory are introduced into data transmission for WSNs,and the optimizations of power,route selection,link capacity and control signal problems are studied.The signal is compressed by matrix completion and compressed sensing before sampling,which can effectively reduce the sampling frequency,eliminate redundant data and decrease data transmission.Based on the protocol limitation of physical layer,MAC layer,network layer and transport layer in network,a cross-layer optimization model of data transmission is established.Congestion ratio is the main parameters.Saving energy and improving network transmission efficiency are objective functions.Service ability and service requirements are constraint conditions.The model designs the optimal power control,routing and link capacity allocation and control input signal.The performance analysis and simulation experiments show that the algorithm can not only reduce data transmission,but also save energy and improve data transmission.Secondly,in order to reduce data transmission delay,improve the efficiency of data transmission,a rapid compression-reconstruction algorithm is proposed,at the same time the optimization protocols are designed in physical layer,MAC layer,network layer and transport layer.Firstly,the fast sensing matrix is constructed so that the original signal can be quickly mapped from high dimensional space to low dimensional space.Then,the optimal power,routing,link capacity and rate are given with the minimum congestion ratio as objective.In order to solve the slowing convergence of nonlinear iteration in signal reconstruction,a linear projection method is proposed,which greatly reduces the signal reconstruction time.Thirdly,the problem of data retransmission caused by signal temporal correlation is solved based on the time entropy.The dynamic programming model is established according to the actual situation of the network,and the data transmission algorithm is optimized in time.The algorithm adjusts transmission rate,link capacity,routing selection and power distribution scheme in real time.The good effectives of the proposed algorithm are demonstrated on the accuracy of transmission and the stability of transmission rate.Fourthly,a data transmission optimization algorithm is proposed to solve the problems of energy limitation of EHWSNs and the spatiotemporal correlation of original data.By Markov random theory,a sensing matrix which excludes spatiotemporal redundancy of data is constructed,so that the data collected by the sensor contains all useful information.According to the objective of maximum transmission probability,a cross-layer optimization model is established based on the constraints of energy consuming-harvesting,link capacity and transmission rate.Lagrangian multiplier method is used to solve the optimal scheme of energy harvesting,routing,link capacity and transmission rate.The algorithm not only simplifies the complexity of data processing,reduces the amount of data transmission,and overcomes the congestion caused by unbalanced data processing problems in traditional algorithm.