A Holistic Approach to Reconstruct Data in Ocean Sensor Network Using Compression Sensing

Abstract

In the complex marine environment, a large-scale wireless sensor network (WSN) is often deployed to resolve the sparsity issue of the signal and to enforce an accurate reconstruction of the signal by upgrading the transmission efficiency. To best implement, such a WSN, we develop a holistic method by considering both raw signal processing and signal reconstruction factors: a node re-ordering scheme based on compression sensing and an improved sparse adaptive tracking algorithm. First, the sensor nodes are reordered at the sink node to improve the sparsity of the compression sensing algorithm in the discrete cosine transformation or Fourier transform domain. After that, we adopt the matching test to estimate sparse degree Kis. At last, we develop a sparse degree adaptive matching tracking framework step-by-step to calculate the approximation of sparsity, and ultimately converge to a precise reconstruction of the signal. In this paper, we employ MATLAB to simulate the algorithm and conduct comprehensive tests. The experimental results show that the proposed method can effectively reduce the sparsity of the signal and deliver an accurate reconstruction of the signal especially in the case of unknown sparsity.

Keywords

• Compression sensing
• sparsity
• ocean sensor network
• data reconstruction