【作者】 宋振峰；

【导师】 王晓湘；

【作者基本信息】 北京邮电大学 ， 信号与信息处理， 2010， 博士

【摘要】 协作通信技术是近几年成为研究热点的能够有效对抗衰落、提高频谱效率和通信可靠性的无线链路增强技术,不仅可以改善信道状况、解决高速传输与网络覆盖之间的矛盾,而且可以促进MIMO技术实用化,因此受到来自学术界与产业界的广泛关注。本文围绕基于协作的无线网络中多用户协作MAC层关键技术展开讨论,研究的侧重点选择在上行链路中协作MAC协议的设计和下行链路中可靠的协作广播MAC层机制。首先,针对时变的无线信道和动态的网络拓扑,研究了协作时机和协作伙伴的选择问题,提出一种基于信道状态信息反馈的自适应协作MAC协议ACoopMAC。该协议中,站点根据瞬时信道信息反馈,对传输方式进行二次判定,选择协作时机,提高了协议的鲁棒性。针对典型网络分配向量(NAV)存在的不足,改进了NAV更新机制,通过缩短节点的信道预约时间,提高系统资源的利用率。仿真结果证明：自适应协作MAC协议在系统吞吐量、接入时延、平均能量消耗、支持节点移动性等方面都优于现有的协作MAC协议。其次,针对现有协作MAC协议为了支持多用户间的协作传输,改变传统控制信令的帧结构而导致的协议不兼容性,深入挖掘传统控制帧的潜力,充分利用帧头中预留的比特位信息,在控制信令握手交互过程中“捎带”传输信道状态信息和帧头中设置协作标志位,减小了系统的控制开销。然后,针对相关研究中缺乏有效的协作MAC协议理论分析模型的现状和MAC协议的传统二维马尔科夫链模型的复杂性,研究了协作MAC协议的性能分析方法,以条件概率的简易模型为基础,结合几何概率的相关理论,提出一种扩展的协作MAC协议的理论分析模型,推导出系统饱和吞吐量和平均接入时延的闭式解。在此基础上,开发了一套行之有效的协作MAC仿真平台,并通过计算机仿真验证了该理论分析模型的科学性。最后,针对多跳无线网络中,终端与基站或中继站间“跳数”的不确定性、网络拓扑的动态性以及无线信道条件的不理想所导致的分组传输冲突和系统效率降低的情况,研究了下行链路的协作广播技术,提出一种基于节点聚簇的MAC层可靠协作广播机制；以条件概率和一维Markov链为基础,对提出的新型协作MAC层广播机制进行了数学建模和理论分析并通过仿真验证了算法的有效性。论文重点研究了基于协作的无线网络的MAC层协议,为上行链路多用户协作传输设计了一种自适应协作MAC协议,为下行链路的广播业务提出了一种可靠的MAC层协作广播机制。论文对协作通信从理论走向应用进行了探索性的研究,具有一定的理论前瞻性和实用性的价值。更多还原

【Abstract】 In recent years, cooperative communications is an enhanced method, which can overcome negative multi-path effects, effectively increase the spectrum efficiency and robustness of wireless systems. Meanwhile, cooperative communication technology can improve the channel situation, promote the practical application of MIMO technology and solve the contradiction between high-speed transmission and network coverage, which caused widespread concern from academe and enterprises, has been recognized as a candidacy of the future mobile communication system.This dissertation studies on MAC layer key technologies of multi-user cooperative transmission in future wireless networks. My research focuses on cooperative protocol design for uplink transmission and reliable MAC layer cooperative broadcast mechanism for downlink transmission.Firstly, to adapt to variable wireless channel and dynamic network topology, I investigate cooperative choice and partner selection mechanism, and propose an adaptive cooperative MAC ptotocol, called ACoopMAC, based on channel state information (CSI) feedback. In ACoopMAC protocl, the station selects transmission mode based on the instantaneous channel measurements and will determine again whether it directly transmits to the receiver or forwards through a helper station, which can enhance the robust of cooperative protocol. Meanwhile, a new network allocation vector (NAV) mechanism is introduced to reduce access delay of stations caused by collision, which is beneficial to throughput improvement. Simulation results show that ACoopMAC is superior to the existing cooperative MAC protocols when it comes to saturated throughput, access delay, mean energy consumption, and node mobility.Sencondly, the existing cooperative MAC protocols change more or less the frame formats of legacy RTS/CTS in order to support the cooprative communications among multi-users, which causes the control signalings of cooperative transmission is not comptabile with those of directly transmission. By fully utilizing the reserved bits in the head of control frame, the information of CSI is piggybacked in control signalings and a cooperation flag is introduced to indicate whether coopertive transmission is adopted. Thus, the adaptive cooperative MAC protocol is completely compatible with the legacy 802.11 DCF.Thirdly, in order to overcome absence of efficent analysis model of cooperation at MAC layer and complexity of typical 2-dimensional Markov chains analysis model for MAC protocols, I investigate performance analysis of cooperative MAC protocl. A simple analytical model based on elementary conditional probability and geometrical probability is proposed to evaluate performance of cooperative MAC protocol and closed-form solutions of throughput and average delay are given. On the basis of this, we develop an efficent system level simulation platform for cooperative MAC protocols and verify its scientificalness by simulation.Lastly, in order to greatly decrease the probability of collisions and improve system efficiency caused by uncertain hops, dynamic network topology and variable wireless channel, I propose a rate-adaptive cooperative broadcast mechanism based on node clustering. Furthermore, I investigate an analysis model based on elementary conditional probability and 1-dimensional Markov chains to model cooperative broadcast MAC scheme and evaluate its performance.The main works of this dissertation are investigating MAC layer protocols for cooperative wireless networks, an adaptive cooperative MAC protocols for uplink transmission and a reliable cooperative broadcast MAC scheme for downlink broadcast services are proposed. The research is helpful for study on cooperative communication from theroy to practice and has forward-looking in theroy and practicability in reality.更多还原