Research On Data Transmission Technology Based On Network Coding

In the traditional mobile communication systems, the intermediate node is only responsible for storing and forwarding the received information. By the application of network coding technology, the intermediate node encodes the received data in order to achieve the capacity of multicast networks. Since the wireless channels have a natural property of broadcasting, and the radio links suffer from fading and interference, if the network coding is applied in the wireless networks, it will not only increase the transmission efficiency, enhance the system throughput, but also improve the transmission performance of the wireless links and guarantee the safety of the information. In this paper, the data transmission technology based on network coding is studied. Apart from research on the application of network coding in the wireless data broadcasting and multicasting systems and the Device-to-Device communication scenarios underlaying cellular networks, effective transmission schemes are proposed as well. Theoretical analysis and experimental simulations verify and prove the performance advantages and feasibility of the proposed schemes.To begin with, the basic theory of network coding is introduced. Since the importance of the wireless data broadcasting and multicasting technologies in the next generation of wireless communications, the basic model of network coding is presented, and the index coding problem in the packet networks is introduced further. Opportunistic XOR network coding and random linear network coding are the network coding methods that will be used later.Secondly, the retransmission problem in the wireless broadcasting system is studied. When the base station broadcasts the same block of data to a group of user equipments (UEs), some UEs may fail to receive all the packets successfully due to the diversity of all the UE’s channel conditions. Therefore, a retransmission is requested, and the base station will retransmit certain packets according to the feedback information. Traditional retransmission schemes are inefficient, so network coding is introduced in the system. In this paper, the key problem to be solved in the retransmission issues is analyzed, and it is converted into a coloring problem in graph theory. Then a Network Coding algorithm based on Index Coding and Vertex Coloring, namely NCICVC algorithm, is proposed. Compared with the classic Genetic algorithm and the traditional retransmission schemes, the proposed algorithm can achieve ahigher transmission efficiency when the implementation is not quite complex. It is also a heuristic algorithm.Next, a brief introduction about the D2D technology surpported by the cellular mobile network architecture is presented. D2D technology is based on proximity discovery. When a D2D group is requesting the same service from the base station, the base station will transmit the data packets via broadcast or multicast links. However, the phenomenon that several D2D UEs can not receive all data packet successfully also exists. In order to offload the base station, and make more efficient use of spectrum resources, a retransmission scheme is considered that UEs who successfully receive all of the data packets will retransmit packets to the D2D UEs through D2D links. In this paper, a cooperative communication scheme based on the opportunitic XOR network coding is proposed and it will work within a D2D group. D2D transmitter (D2D TX) is selected by a certain criteria, and it encodes selecting the data packets efficiently according to the feedback information of the D2D group members. Analysis and simulations about the encoding schemes and signals show that the proposed scheme improves the throughput and performance of the system.Finally, a cooperative communication scheme based on random linear network coding is presented in brief. Application of random linear network coding contributes to the reduction of the signaling overheads, and to further improve system performance, the selection method is optimized.