Scheduler Design for Heterogeneous Traffic in Cellular Networks with Multiple Channels

The design of an efficient base station scheduler with the ability to support different kinds of IP traffic, ranging from conventional data to real-time IP services plays a crucial role in the all-IP convergence goal of next-generation cellular systems. In this context, we first consider the basic network utility based data (rate) scheduler and extend it to a more generic, unified scheduler, capable of handling heterogeneous (data and voice) traffic types and their respective parameters (rate, delay and jitter). More importantly, we then consider the case where both the base station and mobile users are equipped with multiple sub-channels as in OFDM systems. The introduction of multiple sub-channels exponentially increases the search space for scheduling decisions as well as significantly increases the feedback overhead. To reduce the complexity of scheduling and the feedback overhead without sacrificing appreciably on performance, we propose a parameter-based optimization. While such an optimization does not reduce the complexity and overhead in single channel systems, we show that it has the potential to significantly reduce complexity and overhead in multiple channel systems. This is achieved in the form of a unified scheduling algorithm that identifies and exploits specific transmission strategies that optimize individual parameters. This is verified through comprehensive evaluations in a packet-level event-driven network simulator.