Performance Improvement by Dynamic Queue Management in Mobile Ad hoc Network

Mobile Ad hoc network [MANET] consists of mobile nodes without any fixed infrastructure, communicating through wireless medium. In MANET, Congestion occurs in any intermediate nodes when data packets travel from source to destination and they incur higher packet loss rate and long delay, which cause the degradations in performance of a network. Congestion can be controlled by Active Queue Management [AQM] like Random Early Detection [RED]. But RED is very sensitive to traffic load and parameter configuration, and its average queue length varies greatly with the congestion degree and parameter settings. To resolve the above problems, this paper proposes Performance Improved RED (named PIRED) by using auto-tuning RED Queue thresholds parameters and packet dropping probability. An adaptive mechanism is designed to adjust the packet dropping probability and RED Queue thresholds parameters for stable average queue length. The key idea is that when the traffic load changes and therefore the queue length deviates from the target value, the algorithm adjusted RED Queue thresholds parameters and packet dropping probability to improve the network performance over the conventional RED.