Opportunistic Scheduling for Delay Sensitive Flows in Wireless Networks

We present an "opportunistic" scheduling policy with the objective of improving delay performance for time-sensitive users in wireless networks. Since packet delay depends on both resource allocation and time-varying capacity of wireless channels, we introduce a search radius (SR) into the framework of packet fair queueing (PFQ) and employ maximum relative SNR (Max-rSNR) as scheduling rule with the purpose of providing short-term temporal fairness guarantee and improving user throughput. We make theoretical analysis of the delay performance of each user and find that each user\´s delay is directly related to SR, moreover the value of SR should be restricted within a limited range in order to provide better delay performance to each user. Based on this, we propose a feasible iterative algorithm to achieve an appropriate SR. We conduct an extensive set of simulations, which characterizes the performance of our scheduling scheme.