Career: Radio Resource Management for Wireless Data Networks

蜂窝式电话的成功故事促使无线通信界把注意力转向其他信息服务，其中许多属于“无线数据”通信的范畴。蜂窝电话网络操作的一个教训是，有效的无线电资源管理（功率控制、信道分配和切换）对于提高系统的质量和效率至关重要。在包含高速数据信号的系统中，无线电资源管理同样至关重要。本研究的重点是在无线数据网络中开发一种新的无线电资源管理框架。这里考虑的新方法依赖于使用微观经济学理论，该理论在开发用于无线数据服务的分布式无线电资源管理算法时考虑了效用（用户满意程度）和定价（向用户收取的成本）的概念。这项研究将对目前正在发展的第三代无线系统的知识和技术产生根本性的影响。为数据用户的效用开发了一个模型，该模型基于无线终端在其电池寿命期间可以传输的有用比特数。在此基础上建立了无线数据用户使用系统的效用模型，并考虑了效用最大化的分布式功率控制算法。将这些算法表述为非合作对策，研究了这种权力控制的可行性条件以及非合作对策所达到的纳什均衡的存在唯一性。由此产生的均衡被证明是帕累托无效率的，从而激励引入定价来实现效率的帕累托改进。本研究的目标是调查将驱动分布式资源管理方案（优化算法）到帕累托高效操作点的效用和定价函数。需要考虑的具体问题包括基于功率和速率优化的非合作博弈、延迟约束下的效用最大化、效用的信息理论概念和相应的博弈以及能量约束下的动态定价。研究者的教育目标包括：(i)与电气工程、计算机科学和大众传播专业的研究生一起开设无线通信的多学科课程；（ii）无线通信技术课程，讨论无线的理论方面及其与实际标准的关系；（iii）专门的本科生研究计划，重点是无线互联网和应用。

The cellular telephone success story has prompted the wireless communications community to turn its attention to other information services, many of them in the category of "wireless data" communica5ions. One lesson of cellular telephone network operation is that effective radio resource management (power control, channel assignment and handoffs) is essential to promote the quality and efficiency of a system. Radio resource management will be equally, if not more, critical in systems that include high-speed data signals. This research focuses on developing a new framework for radio resource management in wireless data networks. The new approach considered here relies on using microeconomic theories that take into account notions utility (level of satisfaction of a user) and pricing (cost charged to a user) in developing distributed radio resource management algorithms for wireless data services. This research will make fundamental contributions that impact knowledge and technology in the currently evolving third generation of wireless systems. A model for the utility of a data user is developed that is based on the number of useful bits that a wireless terminal can transfer over the lifetime of its battery. Based on such a model for the utility wireless data user derives from using the system, distributed power control algorithms that maximize utility are considered. Formulating these algorithms as noncooperative games, conditions for feasibility of such power control as well as existence and uniqueness of the Nash equilibrium achieved by the non-cooperative game are investigated. The resulting equilibria are shown to be Pareto inefficient, thereby motivating the introduction of pricing to achieve Pareto improvements in efficiency. The goals of this research are to investigate utility and pricing functions that will drive distributed resource management schemes (optimization algorithms) to Pareto efficient operating points. Specific issues to be considered include noncooperative games based on power and rate optimization, utility maximization under delay constraints, information theoretic notions of utility and corresponding games, and dynamic pricing under energy-constrained transmission. The educational objectives of the investigator include: (i) a multidisciplinary course in wireless communications with graduate students from electrical engineering, computer science and mass communications; (ii) a wireless communications technology course discussing theoretical aspects of wireless and its relations to practical standards; (iii) special undergraduate research initiatives focusing on the wireless internet and applications.