Congestion control based on optimal problem for long range dependence network

Article type: Research Article

Authors: Guan, Xinping | Yang, Bo | Long, Chengnian | Liu, Zhixin

Affiliations: Institute of Electrical Engineering, Yanshan University, Qinhuangdao, 066004, P.R. China Tel.: +86 335 807 4722; Fax: +86 335 807 4881; E-mail: {xpguan, yangbo, dylcn, lzxauto}@ysu.edu.cn

Abstract: This paper studies the congestion control in ATM network as a stochastic optimal control problem. Based on the first equivalent controller presented in Automatica 35(12) (1999), 1937–1950, the modified linear quadratic guass (MLQG) control is introduced to control over long-range dependence networks. In addition, the finite impulse response filters employed here compensate the round trip delays as well as smooth transmission rate. Further, we draw a sufficient condition for system stability and analyze the robustness margin along with the maximum tolerable delay from frequency domain technique. With lag compensators satisfactory steady state performance and scalable robustness can be achieved. Also, we expound the choices of parameters from robustness point of view, which has been related with fairness. Through simulations, the effectiveness of the proposed algorithm is shown as compared with the LQG algorithm.

Keywords: ATM networks, ABR, FARIMA, Filter, time delay, fairness

Journal: Journal of High Speed Networks, vol. 14, no. 3, pp. 215-226, 2005