异构无线网络中的能效改善机制研究

The explosive increasing of energy consumption resulted from wireless networks not only deteriorates the greenhouse gas problem, but also increases the cost of network operation; thus, it is necessary to reduce the wasted energy consumption and improve the energy efficiency during the process of information transmission. Although a lot of work has been done on the improvement of the energy efficiency in recent years, there are lots of open issues, such as how to enhance the energy efficiency under the constraint of user's quality of service, the limited improvement of energy efficiency and so on. Considering that many kinds of wireless networks can be integrated into a framework where the coordination ability can be flexibly exploited, the energy efficiency may be significantly improved. In this proposal, cognitive learning, network utility maximum model and dynamic differential game theory are introduced to handle the problems of energy efficiency problem in heterogeneous wireless networks. Firstly,a cognitive learning based scheduling policy of silent base station or access point also needs to be taken into account. Secondly, an energy efficiency model with the constraint of quality of service is required to establish in order to evaluate the performance of the proposed schemes in the future. Finally, dynamic differential game theory based energy-efficient wireless network access scheme is considered. By comprehensively analyzing in theory and network simulation in computer, all the proposed schemes will be validated and assured to apply to the practical wireless networks.

无线网络能源消耗的增长不仅导致温室气体排放增长而且带来网络运营成本增加，因而有必要降低不必要的能量消耗和提升无线网络传输的能量效率。已有研究尚存许多问题，如不必要能量消耗的下降有限，能量效率提升以牺牲用户服务质量为代价等。考虑到异构无线网络能够将多种不同特性的无线网络融合在统一框架以实现不同无线网络之间协作的能力，这将有可能极大提高无线网络的能效，而且能保证不影响用户的服务质量。本课题拟结合认知学习、网络效用最大化模型和动态微分博弈，研究异构无线网络环境下基于认知学习的基站（或AP接入点）休眠调度机制，具有服务质量限制的能量效率模型和基于能量效率的无线网络接入等关键技术，为实现绿色通信提供理论依据。通过深入的理论分析、计算机仿真，使所提出算法在将来能够应用于实际的无线网络中。

无线网络能源消耗的增长不仅导致温室气体排放增长而且带来网络运营成本增加，因而有必要降低不必要的能量消耗和提升无线网络传输的能量效率。已有研究尚存许多问题，如能量消耗的下降有限，能量效率提升以牺牲用户服务质量为代价等。考虑到异构无线网络能够将多种不同特性的无线网络融合在统一框架以实现不同无线网络之间协作的能力，这将有可能极大提高无线网络的能效，而且能保证不影响用户的服务质量。本课题重点是：针对上述问题，考虑利用认知技术对异构无线网络中业务的动态变化进行分析预测，从而更好地管理无线资源，实现能量效率提升。另外，针对异构无线网络特征，如多网络协同通信，研究认知异构无线网络场景下的中继选择、信道分配和功率分配问题。另一方面，考虑到异构异构（混合）无线网络环境下不同网络实体之间的对资源竞争会影响到能量效率的问题，考虑利用博弈论和数学优化等方法研究网络的动态选择、干扰管理和保障服务质量下的无线资源分配等问题。最后，通过构建仿真平台，验证所提出的新算法新机制。研究成果能够适应未来异构无线网络环境下能量效率的需求，从而促进无线通信工业达到节能减排目的。