Improving the performance of wireless LAN using a new scheduling algorithm

Abstract

We propose a prediction-based longest queue first (PLQF) scheduling algorithm. The main idea behind PLQF is that of providing the information of incoming traffic to the scheduler. During scheduling, the scheduler considers not only the queue length, but also the incoming traffic. The purpose is to minimize cell loss ratio (CLR) and improve the buffer utilization. A CSD (channel state dependent) version of the PLQF algorithm is also designed for wireless error-prone channels, targeting to utilize channel status to improve overall performance. Theoretical analysis shows that the PLQF algorithm we proposed can lead to a lower CLR than the conventional algorithms, especially for bursty traffic. Simulation results prove this to be true. Therefore, the PLQF algorithm is promising as an efficient scheduling algorithm.