A Study Of Network Coding Based Multimedia Transmission Technology

Multimedia technology is one of the most rapidly develop and active technologies, and is the focus of the evolution and competition in new generation electronic technologies. With the development of network infrastructure, multimedia (including data, video and so on) services take more and more important role in network communications. However, the reliable data transmission can't be provided by the existing communication networks, especially, the bit-error or packet loss occurs when delivering the multimedia information over networks because of the random and burst errors in wireless channels caused by fading and interference. So the robust multimedia transmission scheme and effective error resilient mechanism must be designed to solve the problem of the multimedia transmission. Furthermore, the apprearance of the hot applications, such as the Internet of things and ubiquitous network, makes more and more terminals and devices with different performance of bandwidth and packet loss rate integrated. Thus how to design the effective multimedia transmission schemes for heterogeneous networks is a problem to be solved urgently. In conclusion, how to improve the data transmission efficiency and ensure the QoS requirement of forthcoming multimedia services is a key research content in the field of multimedia communication.Network coding has been proved to be very effective to make the peer to peer information flow over networks achieve the maximum capacity of that of Shannon channel. On the other hand, scalable video coding is applied to the video coding and transmission over heterogeneous networks. The combination of network coding and scalable video coding can be used to solve the multimedia transmission problems discussed above. In this thesis, based on the multimedia transmission technology with network coding, we investigate several key problems of data and video transmission in network scenario and obtain many practical results, which will help the realization of the multimedia transmission in the next generate wireless network.The main contribution and innovation of the thesis are as follows:1. A realiable scalable video (SVC) transmission system based on random linear network coding (RLNC) is proposed. By analyzing the scalable characteristics of SVC, we setup a state model to represent the transmission order of SVC streaming and present a data partition method for scalable bit streams, which helps the scalable video streaming transmission efficienctly over networks. By combining the advantage of distribution transmission feature of RLNC and scalable approach of SVC, we analyze the packetization, distortion and latency characteristic of SVC streaming, and deduce their expressions respectively. The packetization algorithms, bit stream organization algorithm and bit stream transmission algorithm derived from the characteristics are proposed to improve the reconstructed quality of video and the transmission latency.2. A novel unequal error protection algorithm based on random linear network coding (RLNC) is proposed for scalable video coding (SVC) streaming. Based on the proposed state model of scalable bit streaming, we study the rate-distortion feature of scalable bit stream. By using the RLNC-based packet transmission method, we put forward an effective distortion computation method for SVC streaming, based on which an unequal error protection algorithm for scalable video streaming is presented with the local optimal rate-distortion optimization method. The proposed algorithm effectively improves the video transmission efficiency and the quality of reconstructed video sequence. The simulation results show the effectiveness and feasibility of the new algorithm.3. Opportunity network coding-based retransmission algorithms are proposed for wireless networks. We study the opportunistic network coding, on the basis of which two retransmission algorithms are proposed for single-hop wireless networks. The proposed algorithms select the encoding packets at the base station with the constrain of the minimum number of retransmissions and the maximum combined packets in one combination retransmission packet respectively, and transmit the combination retransmission packet generating by using opportunistic network coding, then the terminals can recover their packets via one retransmission. The the proposed algorithms can effectively reduce the number of transmissions and improve the transmission efficiency of wireless packets.4. Practical broadcast transmission algorithms are proposed for wireless networks. We introduce a single-hop wireless broadcast transmission model, on the basis of which three opportunistic network coding-based broadcast transmission algorithms are presented, namely random algorithm, hash algorithm, and direct algorithm. The proposed algorithms construct the combination retransmission packet with different strategies, and improve the throughput efficiency by recovering the lost packets of different terminals from one retransmission packet. Random algorithm takes the advantage of low complexity, thus practice. Hash algorithm achieves the better performance and low complexity. Direct algorithm is designed to improve the packet transmission delay and is good for video streaming transmission in broadcast scenario. The proposed algorithm can make the average transmission bandwidth achieve the best one of network coding-based transmission schemes under certain conditions. The proposed algorithms can be applied to the existing wireless networks and the 4th generation wireless networks, such as WiMAX or LTE networks.5. A network coding-based hybrid ARQ (HARQ) algorithm is proposed for video broadcast over wireless networks. The sender applies NC technique to combine the unretrieved packets from different terminals so that multiple terminals can recover their lost packets rapidly. By combining the advantage of FEC scheme and NC-based ARQ scheme, the computation of the optimal redundancy of FEC mechanism and the construction of video streaming packets are presented. The proposed algorithm is designed so as to maximize not only wireless throughput efficiency but also reconstructed quality for video broadcast. The proposed algorithm is suitable for the transmission of the video streaming over wireless networks.All of the proposed methods in this thesis are simulated and compared with related schemes in order to verify their advantages and validities.