Radio Resource Management For Broadcast/Multicast In Future Wireless Communication Systems

Along with the popularization of Internet and the development of communication technologies, high-rate data services required by large amount of users will be the main traffic in future mobile communication systems. In some kinds of multimedia applications, e.g. mobile TV, file download and local news report, the same data content would be received by many users, and such applications are typically transmitted through broadcast/multicast mode, so that the radio resources are shared among a group of users, and the network capacity can be fully utilized. As pointed in many studies, efficient and flexible radio resource management (RRM) plays an important role for improving the spectral efficiency of mobile systems, achieving certain fairness criterions among users, and maintaining QoS of different services. Although resource allocation problems in unicast mode have been extensively and intensively investigated, there are still many unsolved issues for multicast mode.The dissertation sets up a complete optimization model for the resource allocation problems in ideal multicast systems, and provides sub-optimal solutions with assistance of mathematical tools. Many considerations in realistic multicast systems, such as the deployment, link adaptation, scheduling, tradeoff between throughput and coverage, overhead of control signaling and feedback, are evaluated and optimized in the dissertation, using a powerful system level simulator, to which we have devoted lots of efforts. The major work of the dissertation and corresponding conclusions are summarized as follows:Chapter 2 deals with resource allocation in ideal multicast systems. Unlike the traditional model where the data rate of the group is limited by user with worst channel condition, we explicitly introduce the concept of special user. The problem discussed is to maximize weighted sum rate of all users, considering multiple multicast groups within the framework of multi-carrier OFDMA, where sum power constraint is imposed. An iterative-separate algorithm and a PSO-based power searching algorithm are proposed to solve the problem. Simulation results show that the tadeoff between sum rate and user fairness can be achieved by flexibly configuring the special users. Futhermore, a modified version of the iterative-separate algorithm is developed, when considering minimum rate constraint per group.Chapter 3 first evaluates and compare different deployments of LTE MBMS, namely p-t-p (unicast), p-t-m (multicast in single cell) and SFN (coordinated multicast in multiple cells), with the effort of building a powerful system level simulator. In particular, a modeling method for SFN is porposed. We then analyze the application of AMC and HARQ in p-t-m deployment, and the results show that with the same coverage, link adaptation leads to remarkable throughput gain. We further propose a feedback reduction scheme for adaptive p-t-m, in which a set of feedback users are selected according to their G-factors, and results show almost no loss in terms of throughput and coverage, compared to the case with all users send feedback.Chapter 4 focus on scheduling and optimized RRM of LTE MBMS. Based on RRM of adaptive p-t-m, we develop a max min-rate scheduling algorithm, where a multicast group is represented by the worst user in resource allocation. Since the algorithm is based on CQI feedback instead of perfect channel state information, considering only the user with worst channel cannot bring any gain, so some modifications are made in case there are large number of users per group. We analyze the impact of several system parameters, e.g. number of users within a group, number of available resource blocks and the accuracy of CQI feedbacks, on the scheduling gain through simulations. A G-factor limiting mechanism is proposed, so that users with G-factors lower than a threshold will not be considered in MBMS RRM, and considerable throughput gain can be obtained at the cost of slight and controllable coverage loss. Furthermore, we propose to combine the G-factor limiting mechanism and MBMS scheduling, and the scheduling gain is proven to be increased with effective G-factor of the service. Simulation results verify that the mechanism and scheduling can boost each other, significantly improving the throughput for adaptive p-t-m.Chapter 5 is concerned about control channel in LTE. We first revisit the system level modeling and corresponding link adaptation algorithm for control channel, and then propose to use OLLA control factor to adjust control channel performance. The control overheads of different transmission schemes, as well as the impact on the throughput of data transmission are analyzed. We then propose the principle that the performance of different data transmission schemes should be compared under same or similar control channel performance, and LTE MBMS deployments are re-evaluated. Futhermore, we introduce the tradeoff between RRM gain and control overhead in multicast systems.In a summary, the research productions of the dissertation could be used as meaningful references for the RRM of broadcast/multicast transmission in future mobile communication systems.