The Research Of The Network Hierarchical Data Fusion Algorithm Based On Distributed Compressed Sensing Network

Wireless Sensor network (WSNs) is a self-organization, and task oriented wirelessnetwork which is composed of multiple nodes. It uses wireless communication technology,the sensor, distributed information processing and embedded processing technologycomprehensively, to real-time acquisition and monitoring the target through the sensor nodes,the node embedded processing module processes the information, and transfer data to controlcentre through a variety of ways, for example the wireless network, and complete informationacquisition and monitoring functions of the target.Compressed Sensing (CS) is the emerging signal acquisition and processing technologyin recent years, and it has brought a revolutionary breakthrough for data fusion technology ofWSNs. In the most existing CS data fusion that is based on WSNs study, the sensor nodes inthe grid sent information directly to the sink node, and only considering the correlationbetween different node information during reconstructing; this can lead to the loss of the nodeinformation and consume too much network energy. According to the characteristics oflimited energy and the processing performance of wireless sensor network node, this paperhas proposed a network hierarchical data fusion algorithm based on distributed compressedsensing: this method is combined with a hierarchical topology control protocol, using thespatial and temporal correlation between node, using the joint sparsity model, and reconstructthe compressed information of the sensor nodes. The results of the experiments show that theproposed approach in this paper is not only can reconstruct the node data informationaccurately, but also can reduce the energy consumption greatly in the process of algorithm. Inthis paper, the main research results are as follows:(1) This paper deeply analyzes the spatial and temporal correlation between wirelesssensor nodes, studies different distributed compressed sensing reconstruction algorithm basedon different sparse model systematically, explores the effective combination of hierarchicalrouting protocols and reconstruction algorithm beneficially; proposes the improveddistributed compressed sensing algorithm based on LEACH protocol and DEEC agreement,which are combined with the sparse model JMS-1and JMS-2respectively.(2) This paper proposes a distributed compressed sensing algorithm based on LEACHprotocolAccording to the distribution of wireless sensor network node which is conformed toJSM-2and the spatial correlation between the network nodes, this paper proposes an improved distributed compressed sensing algorithm based on DEEC protocol. By combiningthe Simultaneous Orthogonal Matching algorithm (SOMP) with LEACH protocol effectively,it has saved energy consumption of the network, improved efficiency of the reconstruction ofthe source information, the results of the experiments show that under the same conditions,the efficiency of distributed compressed sensing algorithm based on LEACH agreementproposed in this paper increases52.3%than the simple SOMP algorithm.(3) This paper proposes an improved distributed compressed sensing based on DEECprotocolAccording to the spatial correlation between the network nodes and the distribution ofwireless sensor network node which is conformed to JSM-1, this paper proposes an improveddistributed compressed sensing algorithm based on LEACH protocol. By combining thedistributed compressed sensing algorithm based on side-information (Siomp) with LEACHprotocol effectively, it has improved the defect of the traditional method of too much energyconsumption, improved efficiency of the reconstruction of the source information, the resultsof the experiments show that under the same conditions, the efficiency of distributedcompressed sensing algorithm based on LEACH agreement proposed in this paper increases38.7%than the simple distributed compressed sensing algorithm based on side-information;and it needs far more less nodes information for the recovery of the target, while thereconstruction of the target information is more accurate.(4) Compare the performance and the practical applicability of these two kinds ofimproved algorithmAccording to the spatial correlation between the network nodes and the distribution ofwireless sensor network node which is conformed to joint spasity models; based on the twodifferent improved distributed compressed sensing algorithms proposed in this paper; wemake a detailed comparison of these two kinds of algorithm. And the results of theexperiments show that under the same conditions, the efficiency of the improved distributedcompressed sensing algorithm based on DEEC agreement increases45.8%than the one whichis based on LEACH protocol; and extends the using range of the improved algorithm, theimproved distributed compressed sensing algorithm based on DEEC agreement apply forJMS-2while the one which is based on LEACH protocol apply to JSM-1.