中文摘要：

MIMO（multi-inputmulti-output）作为一种有效提高无线信道可靠性和带宽的新兴技术，已在无线网络中得到广泛应用．但是，如何利用MIMO在多跳无线网络中为多用户提供高质量视频服务，尚未得到广泛关注．多跳无线链路之间的共信道干扰是需要解决的关键问题．提出一种面向多跳无线网络的多用户视频传输方法，利用链路选择、MIMO的空间复用和空间分集等特点减少链路间的共信道干扰，最大化多用户的平均视频传输质量．通过对链路选择和天线分组进行建模，将上述传输策略抽象成一个最优化问题，该问题是一个NP-hard问题．为了降低求解复杂度，引入遗传算法来求解链路选择问题．该算法采用基因遗传“优胜劣汰”的特性，在保证性能的同时，大幅度降低了求解复杂度．另外，由于遗传算法中每条“染色体”的“优劣”与天线分组策略有关，因此结合可伸缩视频的失真模型，将天线分组问题转化为一个标准的0／1背包问题，并在搜索时采用深度优先和分支限界技术，进一步降低算法复杂度．实验结果表明，所提出的链路选择算法和天线分组算法均能显著提高用户接收视频的质量．

英文摘要：

MIMO （multi-input multi-output）, a promising technology that hopes to improve reliability and bandwidth of wireless networks, has been widely deployed. However there has been little literature that describes how to provide high-quality video service for multi-user over MIMO multi-hop wireless networks. The co-channel interference among multiple wireless links is one of the critical issues. This paper proposes a scheme for multi-user video transmission over multi-hop wireless networks and combats with interference by employing link selection, spatial multiplexing and spatial diversity gain of MIMO. Furthermore, the study formulates this scheme as an optimization problem, which is a NP-hard problem. To make it suitable for practical implementation, the study employs a modified genetic algorithm to solve the link selection problem. Moreover, according to the characteristics of scalable video and MIMO, the antenna grouping problem is transformed into one 0/1 knapsack problem. The depth first search and branch bound techniques are used to further reduce the searching range. The study demonstrate the effectiveness of the proposed scheme by extensive experiments.