宏蜂窝/小蜂窝网络的高能效基础问题研究

This project intends to advance the state-of-the-art of both theory and technology in the area of green information and communication technology (ICT), specifically the green macro-small cellular networks, a typical multi-tier heterogeneous wireless network (HetNet), which has been widely considered as an emerging market of ICT industry. This proposal mainly aims at advancing energy-efficient information theory of green HetNets with a novel concept of "green efficiency" proposed, and developing radio resource management approaches for green HetNets along with the proposal of a novel method of "multi-dimensional resource optimization". And the potential research results will provide a solid theoretical and technological basis for the energy saving and emission reduction of ICT industry. Compared with current homogeneous macro-cells, spectrum efficiency of green small cells could be increased by 5 times, and up to 20bps/Hz. Energy efficiency of green small cells can be raised up to 150 times what it is in homogeneous macro-cells, and up to 30bps/Hz/W. The detailed objectives are as follows: 1) Mathematical Fundamental of Green HetNets: In order to identify the key problems of green ICT, through analyzing the variable efficiency metrics, we shall propose a novel concept of green efficiency and its optimization criterion, i.e., maximizing green efficiency upon guaranteeing QoS of users and system's resource restriction, and investigate the existence, uniqueness, solvability, and practicability of its optima. Based on the Shannon's and Kumar's information theory, we are motivated to develop a set of mathematical tools to model the efficiency metrics of HetNets. In this case, we would be empowered to investigate their theoretical upper bounds, and the variation of the bounds with different types of radio resources as well as various QoS requirements. 2) Radio Resource Management of Green HetNets: In order to approach the theoretical upper bounds of efficiencies in real communications, we shall provide game-theoretic multi-dimensional resources optimization methods. Firstly, we shall design energy-efficient utility functions, which should have both the strict mathematical characteristics and explicit physical meanings. Secondly, In order to reduce the computational complexity of joint resource optimization algorithms, distributed optimization and iterative searching will be used to approach the suboptimal solution. Moreover, Genetic algorithm and Particle Swarm Optimization algorithm will be introduced into the distributed optimization to get a rapid convergence. As a senior research scientist, the applicant has a rich scientific background in green wireless communications. His scientific experience, enthusiasm and proven ability to work as a team-leader, has allowed him to closely link his research team to implement this project effectively.

针对绿色无线通信的应用需求，研究宏蜂窝/小蜂窝网络的高能效传输机理和无线资源管理方法： 首先，提出"绿色效率"理论：1）从高能效的原始内涵出发，分析频谱效率、能量效率与绿色效率的内涵、外延与优化准则，探索其最优解的存在性、唯一性、可解性与可行性。2）依托经典信息论与网络信息论，探索分层异构蜂窝网络的能效理论极限及其成因关系；阐明能效极限与无线资源、用户服务质量之间的内在联系。 其次，提出"多维资源联合优化"方法，以逼近能效极限：1）依托博弈论，构造基于绿色效率的效用函数；该函数既具有严格的数学特征，又具有明确、直观的物理意义。2）为了提高多维资源联合优化算法的收敛速度，拟采用分步搜索、迭代优化，逼近次优解，同时引入遗传算法和粒子群优化算法，以实现分步优化的快速收敛。 与宏蜂窝比较，绿色小蜂窝的频谱效率拟提高5倍，达到20b/s/Hz；能量效率拟提高150倍，达到30b/s/Hz/W。

针对绿色无线通信的应用需求，本项目研究了宏蜂窝/小蜂窝网络的高能效传输机理和无线资源管理方法：.首先，提出高能效传输的信息理论：1）从高能效的原始内涵出发，分析频谱效率与能量效率的内涵、外延与优化准则，探索其最优解的存在性、唯一性、可解性与可行性。2）依托经典信息论与网络信息论，探索分层异构蜂窝网络的能效理论极限及其成因关系；阐明能效极限与无线资源、用户服务质量之间的内在联系。.其次，针对异构网络多维资源联合优化问题：1）依托博弈论，构造基于能量效率的效用函数；该函数既具有严格的数学特征，又具有明确、直观的物理意义。2）对异构蜂窝网络中的时域、频域、码域、空间域、协作域、功率域、能量域等多种资源进行认知和表达，并提出了多维资源优化算法。3）为了提高多维资源联合优化算法的收敛速度，采用了分步搜索、迭代优化，逼近次优解，同时引入遗传算法和粒子群优化算法，实现分步优化的快速收敛。4）研究异构网络的虚拟资源管理方法，虚拟化网络的时延能效制约机制，和联合资源虚拟化映射和管理的方案。.最后，针对多点多用户网络的时频偏补偿问题，本项目研究了基于可容忍同步误差限的收发联合补偿。该策略可建模为有约束的优化问题，由基站与终端的优化准则，分别计算频偏预补偿量和后补偿量。