Study On Loss Recovery Using A Network Coding Buffering Strategy In WSN

Currently, with the continuous development of MEMS, sensor technology, wireless communication and modern network etc., wireless sensor networks (WSNs) have been greatly developed and widely used in military, environmental, medical, home and other fields. WSNs are multi-hop self-organizing networks that consist of a large number of sensor nodes distributed in a specific monitoring area communicating with each other through wireless channel. Their assignment is to perceive, collect and process the information of the specific objects in the coverage area, and then send it to the user. The sensor nodes usually only have limited storage space, limited computational capability and non-renewable energy mostly consumed in the wireless communication. Hence, how to utilize the energy of nodes more efficiently to maximize the network lifecycle is a primary challenge of WSNs.Network coding proposed in2000is a new data transmission mode. Compared with the traditional store and forward mode, it allows intermediate nodes in the network to encode the received information before forwarding. The receivers can decode the received encoded packets to the original data as long as they receive enough encoded information. Network coding can significantly improve network performance, such as increasing network throughput, reducing network delay, balancing network load and improving the reliability of data transmission, etc. Because of the broadcast nature of physical layer and the unreliability of links, wireless networks become the hottest filed of network coding applications. Related works show that by using network coding, WSNs can ensure the reliability of data transmission and decrease the energy consumption.This dissertation focuses on related issues of the reliable data transmission based on network coding in WSNs and the main research works and contributions are as follows:1. This dissertation proposes a buffering strategy based on network coding to utilize the buffer of nodes in WSNs effectively and reduce the retransmission number for loss recovery. In a multicast application with the given collection of receiver nodes in WSNs, hop-by-hop loss recovery can be done by retransmitting the packets stored at buffers of network intermediate nodes and the source node. However, the storage space of the sensor node in WSNs is generally small and intermediate nodes in the network need to allocate some storage space for other tasks, which results in limited buffer used for data recovery. This dissertation introduces network coding theory to the buffering strategy of intermediate nodes. The intermediate nodes store encoded packets in their buffers and retransmit them when necessary. Network coding based buffering strategy weakens the dependence of data recovery on the buffer size and can effectively reduce the retransmission number of lost packets.2. This dissertation proposes a hop-by-hop automatic repeat request (ARQ) loss recovery mechanism based on the network coding buffering strategy, for the purpose of minimizing the retransmission number for loss recovery. In WSNs, when sensor nodes are required to update their software, it is necessary to make sure that all the data for updating is transmitted reliably. However, the unreliable wireless links may lead some packets lost and the retransmission for loss recovery to be the main cause of the increase of network energy consumption. Therefore, how to reduce the retransmission number for loss recovery becomes the most effective approach to decrease the network energy consumption and extend the network lifetime. Under the condition that the buffer of nodes in WSNs is limited, this dissertation proposes a hop-by-hop ARQ loss recovery mechanism based on the network coding buffering strategy, gives the mathematical model of this mechanism and the hop-by-hop ARQ loss recovery mechanism based on the traditional FIFO buffering strategy respectively, and compares the expect number of retransmission needed for loss recovery based on each mechanism.3. This dissertation makes the numerical simulation after mathematical analysis. It chooses the expect number of retransmission and the reduced expect retransmission ratio (RERR) as the performance measures, and sets different packet loss rate, buffer size, and number of child nodes to compare the performance of hop-by-hop ARQ scheme based on the buffering strategy proposed by this dissertation and that based on the traditional FIFO buffering strategy. Simulation results show that network coding can significantly reduce the expect number of retransmission of lost packets in multicast applications of WSNs, and thus can decrease the energy consumption needed for loss recovery.4. This dissertation studies the problem of reliable data transmission in WSNs systematically and proposes a hop-by-hop ARQ loss recovery mechanism based on the network coding buffering strategy, and then compares this mechanism with the hop-by-hop ARQ loss recovery mechanism based on the traditional FIFO buffering strategy comprehensively. Furthermore, on the basis of the works mentioned above, future research work is illustrated.