A Fuzzy Approach for Achieving Proportional Fairness in Data Networks

Proportional fairness criterion, which has proposed the first time by F.P.Kelly, has outstanding properties in allocating users’ rates. For example, it resembles the Jacobson’s AIMD method in rate allocation to users and there exists a well- established stability analysis relating to stability of rate allocation algorithm in the Kelly’s work. The Kelly’s algorithm uses a form of scaled gradient ascent projection method for converging to the equilibrium point, but there is not any method by which users can adjust their gain parameter in the rate allocation algorithm intelligently. In this paper, we have proposed a novel Fuzzy method by which users can adjust their gain parameter based on some measure that is derived from congestion information that is fed back from network to each user. As our simulation results show, our rate allocation method outperforms that of Kelly in rate of convergence.

کلیدواژه ها:

Proportional Fairness ، Fuzzy Inference System ، penalty function ، time-varying scaled gradient method ، elastic traffic ، best-effort.