Efficient information transmission under lossy WSNs link using compressive sensing
Liantao Wu,Kai Yu,Tianxu Du,Zhi Wang
DOI: https://doi.org/10.1109/ICIEA.2014.6931214
2014-01-01
Abstract:Compressive sensing (CS) is applied to sparse signal transmission so that it can be transmitted efficiently over lossy wireless links. By exploiting the commonly sparse property of measured signal within wireless sensor networks (WSNs), we propose a CS-reconstruction based efficient information transmission framework. According to CS theory, if the sensed information has some sparsity, it can be reconstructed with only a few sensed data. In this case, we argue that, by using CS technique, information transmission can tolerate a certain degree of link lossy without requiring all of the data being successfully transmitted, thus avoiding the expensive data retransmission. Moreover, CS-based information transmission framework is established, where the lossy link transmission is modeled as compressive sampling process. Data packets are directly transmitted after signal sampling, then the sensing matrix is obtained through the original sequence of received broken data and finally signal is reconstructed through optimization algorithm. Through experimental verification, we first show the lossy link and sparsity of signal. Further, aiming at two distinct links, we make a couple of comparison tests, which shows our method achieves the same good reconstruction performance as conventional multiple data retransmission scheme does in good link. While in bad link our method outperforms conventional method even it adopts multiple retransmission. Results verify that during lossy link information transmission, the proposed CS-based method obtains high information transmission quality, also significantly reduces he energy cost and latency.