Scheduling algorithms for high-throughput packet data service in cellular radio systems

This paper examines the performance of a number of scheduling algorithms for the wireless packet data access evolution of third-generation cellular systems. The algorithms are analyzed using three different wireless channel models (two pedestrian, one vehicular). For each channel model, a comparison of the performance of the algorithms using outdated channel state information plus margins tuned to provide an average 1% packet error rate, as well as using perfect channel prediction in order to determine the supportable bit rate and transmission format for each user, has been carried out. The performance of the algorithms is evaluated in terms of the average throughput per sector as a function of the number of users. The average delay per packet and per user versus the number of users per sector and the distributions of allocated slots per user are also determined as a measure of the fairness of each algorithm. It is also shown that the use of outdated information and margins can be an effective substitute for prediction, provided that the outdated measurements are reasonably accurate.