Channel modeling and network control have long been separate in wireless research. This is mainly because the existing channel modeling techniques are neither measurement-based nor network-oriented. They cannot characterize the great diversity of wireless channel dynamics, which are time-varying, frequency-selective, and highly dependent on many system factors such as noise, distance, mobile speed, channel bandwidth, power control, coding, fast and slow fading, etc. Also, they have largely ignored many important network dynamics, such as buffer capacity, data arrival statistics and quality of services. In consequence, the significant impact of wireless channel dynamics on network packet-level performance cannot be evaluated, which impedes the development of effective network/channel control technologies to cope with wireless channel dynamics for performance improvement and quality of services.
The proposed modeling technique is measurement-based and network-oriented, which provides a unique approach toward the integration of channel modeling and network control. The study will start with single-channel modeling to identify what are important channel statistics to end-to-end packet-level data performance, given a variety of channel dynamics, arrival data statistics and buffer allocations. Basic guidelines will then be developed for channel modeling to capture such important channel statistics. The study will help us to evaluate the effectiveness of new channel control technologies such as adaptive coding and power control to overall data performance improvement. The research will be further extended to multi-channel modeling at network level, where multiple channels are shared by a pool of mobile users at different geographic locations with different driving patterns. Both CDMA and TDMA networking technologies will be evaluated. Other advanced network design issues such as buffer congestion control, priority scheduling and multiple access control will be examined. The fundamental understanding of interactive roles played by channel dynamics, traffic behavior and network control will greatly help improve the performance of next generation wireless multimedia networks.
Throughout the proposed study, a sophisticated wireless channel modeling tool will be developed, which can assist wireless engineers for network analysis, control design and service planning.
