Research On Network Coding Based On Distributed Compressed Sensing

The International Telecommunication Union (ITU) published a "ITU Internet Report2005:Internet of Things "in2005, and the concept of the Internet of Things(IOT) was first stated formally.As the extension network of IOT terminal, Wireless Sensor Network (WSN) plays a key role as abridge of the physical world and the virtual information world. The research of transmission andprocessing technologyof information in WSN has achieved a lot of achievements, however, theconcept of IOT covers a wider range, greater network size, more diversifiedtypes of network andmore massive amount of sensing information. Thus, collecting efficiently and processing promptlyof diverse data at the front end of IOT become the most critical step from theory to application.Both WSN and Communication Transmission Networks transmit signals which carryinformation, so the research of signal processing technology and the development ofcommunications technology are always inextricably linked. Distributed Compressed Sensing (DCS)is an effective technology of source processing in consideration of the sparse of intra-signal and thecorrelation of inter-signal. Given that there is a dense distribution of wireless sensors and a strongcorrelation of collected data from the neighboring nodes, it is possible to compress and aggregaterelative data. Therefore, on the background of the wireless sensor network in IOT, this thesis makesuse of the distributed compressed sensing technology to collect information at the source nodes, thenetwork coding technology to transmit multiple packets at the aggregation nodes, thus improvingtheefficiency of data compression and information transmission in WSN. The main innovations of thethesis are summarized as follows:(1) This thesis proposes a transmission method of Compressed Sensing(CS) as a sourceprocessing technology combined with network coding. In wireless sensor networks, usingcompressed sensing and network coding technology jointly, not only can reduce the times of packettransmission, but also improves the efficiency of data transmission and the throughput of thenetwork. The scheme in this thesis takes full account of the intra-signal sparsity and inter-signalrelativity in wireless sensor networks. Simulations show that the joint compressive sensing andnetwork coding or joint distributed compressed sensing and network coding, compared to onlynetwork coding scheme, improves the compression ratio of the data transmission.(2) Based on the multipath random network coding, this thesis proposed a transmission methodnamed randomized network coding based on distributed compressed sensing for multipath network. For general network, this thesis, on the base of the sparsity rate of distributed compressed sensing,consider the intra-signal sparsity and inter-signal relativity in wireless sensor networks, andcompressed the data at the source node, then transmit the data after applying the multipath randomnetwork coding. Both theoretical studies and simulation results show that this method eliminatesredundant data and further improves the efficiency of data transmission in the network, furthermore,the method has a wide range of applications.