Research On Sparse Basement Based On Compressed Sensing In Multi-parameter Wireless Sensor Network

Wireless sensor networks can perceive and collect kinds of information with high dense micro sensors random or fixed deployed in the monitoring region. Based on self-organization networking and multi-hop relaying communication, the data is effectively sent to the users. Wireless sensor network has high application value. However, the limited energy supply for wireless sensor nodes seriously affects the development of large-scale and high-density wireless sensor networks. Compressed sensing theory provides a new solution to solve the problems about nodes’limited energy and unbalanced energy consumption.After in-depth study of wireless sensor networks and compressed sensing theory, this paper establishes a Multi-parameter Transmission Model. In this model, every cluster-header node only sends a signal with same smaller dimension after data fusion of all sensor information in its cluster based on compressed sensing theory. As a result, it makes load balanced in the global network. In addition, choosing new cluster-header periodically by energy optimal selection algorithm so as to ensure load balanced in clusters.Furthermore, it proposes a new method to construct a multi-parameter joint sparse basement used in Multi-parameter Transmission Model, which can greatly relieve storage burden of base station. It uses Daubechies wavelets function as the atom model and gets coefficient vectors by Matching Pursuit algorithm to obtain the joint sparse basement.Judging from the simulation by Matlab, the Multi-parameter Transmission Model and multi-parameter joint sparse basement are both feasible and stable. The research will greatly promote compressed sensing theory applied in wireless sensor networks, which has extremely important significance.