Optimal and Approximate Mobility Assisted Opportunistic Scheduling in Cellular Data Networks

This paper considers the problem of scheduling of multiple users in the downlink of a time-slotted cellular data network. It introduces optimal and approximate opportunistic scheduling algorithms, which combine channel variations and user mobility information in the decision rule. The proposed algorithms modify opportunistic scheduling algorithm of Liu et al. with dynamic fairness constraints that adapt according to the user mobility. The optimum algorithm is an offline algorithm because it pre-computes constraint values for all mobility states according to a known mobility model. The approximate algorithm is an on-line algorithm, and it relies on the future prediction of user mobility locations in time. These predicted values are used in computing constraint values. Simulation results illustrate the usefulness of the proposed schemes for elastic traffic and restrictive constraints. The use of mobility information in opportunistic scheduling also increases channel capacity.