基于缓存辅助的双向多中继网络调度方案

发布时间：2018-02-13 13:16

  本文关键词： 缓存辅助中继 双向多中继网络 平均和速率 功率分配 KKT条件　出处：《西南交通大学》2017年硕士论文　论文类型：学位论文

【摘要】：中继技术,即信源节点不直接发送数据给目的节点,而是通过引入的中继节点进行数据的转发,从而完成信源节点和目的节点之间的数据交换。中继技术的引入可有效对抗衰落,提升系统的传输可靠性。缓存辅助的中继技术,即给中继节点配备缓存,使得中继能够储存接收的数据,而不用立即转发出去,从而打破传统的一收一发的固定传输模式,系统可以根据当前时隙不同的链路信道质量好坏以及缓存中的数据状态,自适应地选择链路进行信息的传输。同时,缓存技术也为多中继网络选择最佳中继提供更多机会。缓存的引入对提升中继系统吞吐量、分集增益以及信噪比等方面都有重要贡献。本文主要研究缓存技术在双向多中继网络中的应用,即基于系统平均和速率最大化的缓存辅助双向多中继网络优化问题。论文首先简单介绍几种基本的中继网络传输模型,给出不同的中继转发策略以及基于不同通信指标的中继选择策略。其次,研究现有的缓存辅助的中继技术,即针对单中继网络的缓存辅助链路选择策略,针对多中继网络的缓存辅助中继选择策略以及缓存辅助技术在双向单中继网络中的应用。接着,研究缓存辅助技术在双向多中继网络中的应用,即针对缓存辅助的双向多中继系统,以缓存空间无限大为前提,在缓存收发信息平衡的情况下,以系统平均和速率最大化为目标,提出联合优化中继选择、传输模式选择和传输速率选择的网络调度方案]。具体地,运用KKT条件求解的优化方法,在每个时隙根据信道状态信息,自适应地选择最合适的中继、传输模式及传输速率。本文对所提方案Ⅰ的性能进行了仿真分析,并和传统的中继处没有配备缓存的TDBC协议、MABC协议以及点到点的双向中继协议进行了比较,证明由于缓存的引入导致了双向中继网络系统性能的极大提升。最后,针对所提出的缓存辅助的双向多中继网络调度方案Ⅰ中所有节点都采用固定相等的传输功率,并未考虑功率限制问题,提出根据不同链路状态,自适应地给各个节点分配功率进行信息传输,考虑在各个节点平均功率受限制的情况下,对所提出的调度方案Ⅰ进行优化,即缓存辅助的双向多中继网络调度方案Ⅱ。本文对优化方案Ⅱ的性能进行了仿真研究,并和以固定功率进行传输的方案Ⅰ的性能进行了比较,证明功率分配可以更大地提升系统的整体性能,实现系统的平均和速率最大化的优化目标。
[Abstract]:Relay technology, that is, the source node does not send data directly to the destination node, but transmits the data through the introduced relay node. Thus the data exchange between the source node and the destination node is completed. The introduction of relay technology can effectively resist fading and enhance the transmission reliability of the system. It enables the relay to store the received data without forwarding it immediately, thus breaking the traditional fixed transmission mode of one receive and one transmission, and the system can vary the link channel quality and the data state in the cache according to the current time slot. At the same time, cache technology also provides more opportunities for multi-relay network to choose the best relay. The introduction of cache can improve the throughput of relay system. Diversity gain and signal-to-noise ratio (SNR) play an important role. In this paper, the application of cache technology in bidirectional multi-relay networks is studied. That is to say, the buffer-assisted bi-directional multi-relay network optimization problem based on system average and rate maximization. Firstly, several basic relay network transmission models are introduced briefly in this paper. Different relay forwarding strategies and relay selection strategies based on different communication indexes are given. Secondly, the existing cache assisted relay technology is studied, that is, cache assisted link selection strategy for single relay network. The cache assisted relay selection strategy for multi-relay network and the application of cache auxiliary technology in bidirectional single-relay network are discussed. Then, the application of cache auxiliary technology in bidirectional multi-relay network is studied. That is to say, in the case of buffer receiving and sending information balance and system average and rate maximization, a joint optimization of relay selection is proposed for buffer-assisted bi-directional multi-relay systems with infinite buffer space as the premise. The network scheduling scheme of transmission mode selection and transmission rate selection]. Specifically, using the optimization method of KKT condition solving, adaptively selecting the most suitable relay in each time slot according to the channel state information, In this paper, the performance of the proposed scheme I is simulated and analyzed, and compared with the traditional TDBC protocol without buffer and the point to point two-way relay protocol. It is proved that the performance of bidirectional relay network system is greatly improved due to the introduction of cache. Finally, all nodes in the proposed buffer-assisted bi-directional multi-relay network scheduling scheme I adopt fixed and equal transmission power. Without considering the power limitation problem, this paper proposes an adaptive allocation of power to each node for information transmission according to different link states, and optimizes the proposed scheduling scheme I under the condition that the average power of each node is limited. In this paper, the performance of optimization scheme 鈪，

本文编号：1508249