Study On Power Allocation And Adaptive Transmission Of Network Coding

Due to the scarce spectrum resource, how to use the space resource to increase thethroughput becomes a research focus where MIMO (Multiple-Input Multiple-Output)and relay technology are effective solutions. In traditional relay technology, informationis forwarded by the relay to enlarge the transmission radius. However, the simpleforwarding can not achieve the maximum flow transmission, which is the maindrawback of network communication. To solve this problem, wired communicationbased network coding (NC) was proposed. Although proposed in wired network, NC ismore useful in wireless network because of its broadcast nature. Applying NC inwireless network improves the spectrum efficiency while reducing transmission powerand delay. Therefore, wireless NC has become one of the recent research focuses.The research of this thesis includes the following points:①First, the fundamental theory of NC is explained. Then, the application of NCin wireless network and its latest progress are introduced. In addition, the relaycooperative technology is studied. Finally, the application of NC in the relaycooperative technology is discussed.②To enhance the performance of bi-directional multi-hop relay collaborativenetwork, two PNC based optimal power allocation schemes are proposed in the rayleighfading channel. By using the convex optimization theory, power allocation schemes areobtained to minimize the system outage probability and system power comsumptionrespectively. Numerical results show that the outage probability and power consumptionof the proposed schemes are reduced compared with the traditional uniform powerallocation scheme. Furthermore, with the increasement of node number, theperformance enhancement is more obvious.③To increase the throughput of bi-directional relay collaborative network, anadaptive transmission scheme based on NC is proposed. Rayleigh fading is consideredin all links. First, the throughputs of several network coding schemes are computed interms of the channel state information. Then, the scheme with the maximum throughputis selected as the transmission scheme. Numerical results show that the throughput ofthe proposed scheme is the highest among these network coding schemes. In the lowSNR region, the BER curve of the proposed scheme approaches that of traditionalscheme with the lowest BER; in the high SNR region, though being higher than the BER of the traditional scheme, the BER of the proposed scheme is lower than that ofANC scheme.