Compressed Sensing Multiuser Detection Techniques For Underwater Sensor Networks

In recent years, with the continuous development of marine economy and gradual progress of natural and engineering science, exploration and utilization of marine resources have been increasing dramatically and underwater sensor network has thus become a hot area of research. In the underwater sensor network, communication environment is significantly different from that in the air mainly in two following aspects. First，due to rapid decay of electromagnetic waves, acoustic communication is widely used in underwater environment while it suffers from specific problems such as limited bandwidth. Second, as most underwater sensor networks comprise a large number of sensor nodes, which are deployed at the bottom of ocean or suspended in the deep sea, it is really difficult to supply sustainable energy or frequently recharge new batteries for them. Therefore, for underwater sensor networks, efficient use of bandwidth, conservation of energy and enhancing lifetime of network become extremely important. Encouragingly, the revolutionary theory of compressive sensing technology provides an innovative solution to optimizing underwater sensor network.The compressive sensing theory exploits the fact that as long as the original signal has a sparse representation in a transform domain, it can be measured utilizing some kind of sensing matrix and the original signal can be reconstructed exactly from a few measurements. Conditions are that the sensing matrix is limited to satisfying the RIP criteria and it does not correlate with the sparse matrix. The reason why compressive sensing theory can be applied in the underwater sensor network is that most natural signals are sparse or compressible in a sparse matrix. Applying compressive sensing theory, the original signal can be recovered from a very small number of samples yet with high probability and precision，which differs from traditional method use.Two main aspects are included in this paper. One is data acquisition scheme based on compressive sensing theory in the two-dimensional underwater sensor network; another is mainly focused on data recovery and multiuser detection technology. For the proposed scheme, compressed sensing theory combined with multiple access and multiuser detection technology respectively, greatly reduced the required number of sensor nodes, and the number needed to obtain and transmit in the underwater sensor network, thus reaching the goal of saving energy and bandwidth. The proposed compressive sensing multiuser detection scheme mainly considers the situation that the number of active users in a practical multiuser system is far less than the total number. According to the sparsity level of the signals being detected, a clever switching between compressive sensing reconstruction algorithm and classical multiuser detection technique contributes to the decrease of BER in the multiuser system. Theoretical analysis and simulation results demonstrate that compared with conventional solutions the proposed scheme possesses great performance gains.In this paper, the research provides theoretical supports and technical references for the support of large-scale applications and development of marine exploration.