Research On Energy-efficient Data Transmission Methods In Wireless Sensor Networks

Wireless sensor networks(WSN),especially those with large,dense nodes,are used in a wide range of applications,such as military,environmental,and industrial.But the nodes are usually arranged in remote and inaccessible areas,and the energy is very limited.If the energy is exhausted,it means that it will not work properly,causing the original function of the network to be reduced or even disappeared.Therefore,in WSN,energy constraints are an urgent problem to be solved.In WSN,the sensor data has spatio-temporal correlation,which makes the collected data repetitive.If all the data is transmitted,a large amount of node energy will be wasted,and the network lifetime will be greatly shortened.Therefore,In WSN,designing an energy-efficient wireless transmitted method is the key to increase network lifetime and improve network performance.In WSN,the advantage of compressed sensing technology is that it can effectively reduce the amount of data transmitted,because it can recover the original signal as long as it receives a small amount of data at the receiving end,and its acquisition speed is much lower than twice the original signal bandwidth;Network coding linearly or non-linearly combines data from different links and then forwards them,which enhances the anti-interference of the network and effectively increases the amount of data transmitted;In WSN,except for compressed sensing and network coding,Routing technology also plays an indispensable role in the development process.Clustered routing technology can evenly distribute the amount of data in the system,reduce resource consumption and increase network lifetime.In order to reduce network energy consumption and balance network load,this thesis explores more efficient data transmission methods,innovations and research based on existing energy-efficient wireless transmission methods based on compressed sensing and network coding and clustering routing methods.The plan is as follows:(1)This thesis proposes a new wireless data transmitted and recovered method STCNCDR(Spatio-temporal Compressed Network Coding Based Data Communication and Data Recovery,STCNCDR)by using compressed network coding technology,which mainly solves the problem of high energy consumption and limited recovery accuracy during data transmission.Specifically,in terms of spatial correlation of data,the method uses the two-hop neighbor information to select the best next hop candidate node,the intermediate node,based on the compressed sensing and network coding,to reduce the transmission of redundant data avoiding network resources wasted;in terms of time correlation,the time-sparse dictionary is trained by the dictionary training method KSVD algorithm to obtain the best time observation matrix,so that the data is better thinned,thereby improving the accuracy of data recovery;and utilizing the temporal correlation of data,It is compressed to reduce its number of transmissions,further reducing energy consumption.The method can significantly reduce the amount of information transmitted in the network,that is,compared with other methods,the number of node reception and transmission can be reduced by about 65% to 70%,which effectively reduces the network energy consumption,and the method also improves the precision of recovered data at the sink node.(2)Based on the existing cluster head selection method,this thesis proposes a dynamic cluster head selection method based on energy and density I-EEC(Improved Energy-efficient Cluster Head Selection Technique,I-EEC),which combines the residual energy of the node with the dynamic density in the cluster.In the first choice,the node dynamic density is used to avoid some nodes being selected as the cluster head,which leads to minimize the premature death of the node,and a dynamic weight factor is introduced to balance the remaining energy of the node and the dynamic density of the cluster.Specifically,as the number of transmission rounds increases,the proportion of the remaining energy in the cluster head selection process gradually increases,so that the probability that the node with lower energy becomes the cluster head minimizes the premature death of the node,thereby further prolonging the life cycle of the network.The simulation results show that the method can balance the energy consumption in the network,improve energy efficiency and increase network lifetime.