Authors : R.M. Bhavadharini, S. Karthik and N. Karthikeyan
Abstract: Congestion in the network is a severe problem that affects the performance of MANET applications. When, a packet loss occurs in the wired medium, it clearly signs congestion in the network and suitable measures are taken to alleviate the effect of congestion. In a wireless environment, a packet can be lost for the factors such as wireless link error probability, link breakages, energy depletion, MAC contention fading and interference, etc. Due to various reasons contributing to the packet loss, Transmission Control Protocol (TCP) is unable to efficiently exploit the available network capacity in MANET. Across-layer congestion control algorithm with efficient Bandwidth Allocation and Dynamic Congestion Window Adaptation (CC-BADWA) is proposed in this study that uses TCP receiver’s prediction of bandwidth availability in the network and the receiver advertised window size back to the sender which then adjusts sender window. The timeout and the delayed acknowledgement can differently able the sender to predict the level of congestion in the network. When the acknowledgement is not received due to congestion in the network, the sender can estimate the transmission probability, adopt the dynamic congestion window size and transmit the packets. This improves the network throughput since, the sending rate is adjusted consistent to the currently prevailing network status. The bandwidth allocation uses the currently available bandwidth based on the present active connections and the route optimization proposed in this research selects the optimal path for data forwarding. The receiver’s timeliness helps to improve the network performance parameters throughput, end to end delay, packet delivery ratio of the existing method. The proposed algorithm is analyzed through ns 2 simulation tool.
R.M. Bhavadharini, S. Karthik and N. Karthikeyan, 2016. Cross-Layer Congestion Control with Efficient Bandwidth Allocation and Dynamic Window Adaptation in Mobile Ad Hoc Network. Asian Journal of Information Technology, 15: 455-466.