Research On Reliable Transmission Technology Based On Network Coding In Wireless Networks

With the development of communication technology, wireless network technology has been widely applied in people’s daily life. Since the nodes in the wireless network is easy to move and the channel quality of the wireless network is really poor etc., packet loss and error codes. will easily happen during the data packet transmission in the wireless network. In the wireless transmission process, once the packet loss occurs, it must be retransmitted to ensure the reliability of the communication link. How to improve the retransmission efficiency of wireless networks and reduce the network latency has always been a hot topic in the field of communication, and the proposed network coding points out a new direction for the research. Most existing network coding retransmission algorithms based on network coding, in the code package selection, would employ greedy coding algorithm which tries to make the number of the together encoded packet loss maximized while resulting that some encoded packet cannot be successfully decoded. Furthermore, most retransmission algorithms do not take the constraint of the data packet delay into consideration, and that will cause a number of ineffective data packets to the receiving node and a poor network service quality. This paper studies the retransmission problem of lost data packet in the wireless communication, and the main contribution of the work includes:(1)For different scenarios of each link loss rate in the wireless network, based on the network coding, the Efficient Weighted Retransmission Based on Network Coding (NCEWR) algorithm is proposed. Firstly, in the establishment of data packet reception matrix, the opposite of the packet loss rate which indicates the probability of that this receiving node can correctly receive the data packet again would be introduced into the matrix. Secondly, the weighted summation of elements in each row would be done, and this value can be viewed as the expectation of the successful reception of the receiving node of the corresponding data packet. Then according to the calculated value, a weight table corresponding to the reception can be obtained. And this table would rank in an order of value descending. Finally, the greedy algorithm encoding conditions are improved. Not only each encoded packet contains as much of the lost data packets as possible, but also ensuring that all of the nodes which have received this coded packet can be successfully decoded. To verify the performance of this algorithm on the platform of Matlab simulation software, the result shows that this improved algorithm has less packet retransmission times and lower computational complexity than the existing related algorithms.(2)For the time-sensitive scenario, the low-latency network coding retransmission algorithm (NCEWR-LD) is proposed. When creating the searching table D, the average weighted delay of each data packet which can reflect the urgent degree of the receiving node needing the data packet is calculated and then added into D. When finding the data packet to encode, the average data packet delay constraint would be viewed as the first criterion, while the corresponding weight value would be viewed as the second criterion. Finding the data packet in the way according to the sequence of the orderly increased average delay constraint and the gradually reduced weight value can guarantee that the data packet with small delay constraint may prior transmitted and the number of transmitted original data packet is as large as possible and achieve the purpose of minimizing the number of data packet exceeding the delay constraint. When doing the algorithm simulation analysis, using the exceeding delay constrained rate to evaluate the performance. The Simulation result shows that, under the same conditions, the NCEWR-LD algorithm has the lowest exceeding delay constrained rate, namely it has the best performance.