The main focus of this work is a method development of efficient use of scheduling algorithm in wireless packet transmission. For the next generation wireless access systems, wireless transmission needs efficient QoS control in respect to throughput and fairness consideration. However the state of the wireless channel can change dynamically. With this fact, the wireless system performance will not be stable and the efficient scheduling is not easily achieved. In this work, we propose the wireless packet scheduling algorithm including wireless link state prediction method for the fixed wireless environment. The one-step prediction is implemented using LMS criteria for OFDMA sub-channels. With the predicted sub-channel states, scheduling algorithm can work to balance data service transmission rate for fairness and furthermore throughput. Simulation results are shown that the prediction accuracy rate is increased using the predictive sub-channel allocation than the simple adaptive sub-channel allocation for OFDMA system with adaptive modulation. Furthermore, we show the fairness is better achieved using weighed fairness factor in the simulation.
