Research And Implementation Of Routing Algorithm Based On Compressive Sensing Theory For Wireless Sensor Network

Wireless sensor network is mainly used for monitoring and sensoring the objective physicalworld, has become a hot topic. Due to the limited battery power contained by the sensor nodes inthe network, the efficient use of energy is the core issue in wireless sensor network. Compressivesensing is a new theory in the field of signal processing, and it overturns the sampling frequencywhich put forward by the traditional Nyquist criterion. The theory states that the signal can bereconstructed from a small amount of the samples accurately if it is sparse or has a sparserepresentation in one basis. As most natural signals are sparse or have a sparse representation in onebasis, it provides the precondition for the combination of compressive sensing theory andwireless sensor network. The compressive sensing theory has been applied to wireless sensornetwork mainly in the aspects of date processing, location, routing and so on.This dissertation firstly applys the compressive sensing theory to the single hop wireless sensornetwork. There will be much signal interference and data redundancy when seneor nodescommunicate with the sink node. And it is not conducive to the effective use of data acquisitionand network energy. Therefore, this dissertation proposes a node scheduling scheme, using arandom probability model for chosing a subset of the network nodes to sample. The underwaterwireless sensor monitoring network built based on the scheme can own lower energyconsumption and the sparse sampling signals can be reconstructed approximate to original signalwith a large probability.Due to the limite of communication distance, sensor nodes and sink node can not bereachable via single hop in some scenes. Therefore, it applys the compressive sensing theory to themulti hop network in this dissertation. LEACH algorithm is a hierarchical routing algorithm inwireless sensor network. There exist some problems in this algorithm although the selection ofcluster heads ensures randomness and equality of opportunity. This dissertation proposes anoptimizing algorithm CS_LEACH_UC, aiming at these problems and combining with thecompressive sensing theory. A system prototype is implemented to evaluate the algorithm ofuniform clustering and sparse sampling. Performance analysis shows that the CS_LEACH_UCalgorithm is superior to LEACH in various respects, including energy consumption balance,network lifetime and so on. Thus, the proposed CS_LEACH_UC is feasible and effective.