Energy and delay constrained scheduling of real time traffic over OFDM communication system with non-causal channel information

Abstract

A point to point wireless communication system based on Orthogonal Frequency Division Multiplexing (OFDM) is considered, wherein R bits of information have to be scheduled within a hard delay deadline of T time slots over a frequency selective slow fading channel. The scheduling objective is to minimize the energy expenditure with non-causal knowledge of channel state over a limited time horizon, number of bits to be served and delay constraints. We assume a convex monomial energy-bit cost function and propose novel efficient two stage heuristic algorithm that allocate the R bits on the time slots, followed by greedy optimal bit loading over the OFDM subcarriers. An analytical model is developed and numerical simulations are performed to validate the proposed approach as well as the results show that our algorithms outperform existing approaches in published literature by up to 10 dB in power margin on an average in high load conditions while meeting the deadline of T timeslots.