An OpenMP-Based Algorithmic Optimization for Congestion Control of Network Traffic

Abstract

The last decade being a web revolution in the field of electronic media, data and information exchange in various forms has significantly increased. With the advancement in the technological aspects of the communication mechanism, the textual form of data has taken the shape of audiovisual format, and more and more content over internet is being shared in this form. Data sharing in this form calls for the need of high bandwidth consumption which may slow down the network resulting in performance degradation of content delivery networks due to congestion. Several attempts have been made by the researchers to propose various techniques and algorithms to achieve optimal performance of the network resources under high-usage circumstances. But due to high-dense network architectures, the performance implementations of suggested algorithms for congestion may not be able to produce the desired results in real time. In this paper, we have presented an optimized multi-core architecture-based parallel version of two congestion control algorithms—leaky bucket and choke packet. The experimental results over a dense network show that optimized parallel implementation using OpenMP programming specification gets the network rebalancing in a very short span of time as compared to its serial counterpart. The proposed approach runs 60% faster than the serial implementation. The graphical map for the speed up continues to increase with the size of the network and routers. The paper throws the light on the implementation aspects as well as result analysis in detail along with some existing algorithms for the problem.