A Distributed Multipath Routing Algorithm To Minimize Congestion

Xin Guo,Jun Zhang,Tao Zhang

DOI: https://doi.org/10.1109/DASC.2009.5347425

2009-01-01

Abstract:Unlike traditional routing schemes that route all traffic along a single path, multipath routing strategies split the traffic among several paths in order to ease congestion. It has been widely recognized that multipath routing can be fundamentally more efficient than the traditional approach of routing along single paths. Yet, most studies in the context of multipath routing focused on centralized methods, which need the information of whole network and not suitable for mobile networks. In this paper, we demonstrate the significant advantage of distributed multipath routing solutions for congestion minimization. Our work exhibits several important properties that include: 1) routing decisions are made locally by network nodes without the centralized information of the entire network topology; 2) routing decisions minimize network congestion factor, which is the max value of proportion between link flow and link capacity in all links in networks; and 3) routing decisions address link congestion avoidance via topology aggregation and maximum flow optimization. We formalize the problem that incorporates two requirements of multipath routing: distribution and congestion minimization. Then, we establish the detail routing algorithms and protocols to solve the problem. Finally, under different network topologies, we approve the better performance of our proposed algorithm (D-CMM) by decreasing the congestion factor and increasing the throughput compared to other multipath routing algorithms, and D-CMM algorithm is more suitable for long distance communication and link-dense networks.