Title :
Adaptive Load-balancing in WDM mesh networks with partial traffic information
Author :
Dai, Rui ; Li, Lemin ; Wang, Sheng ; Zhang, Xiaoning
Author_Institution :
Key Lab. of Broadband Opt. Fiber Transm. & Commun. Networks, Univ. of Electron. Sci. & Technol. of China, Chengdu
Abstract :
Valiant load-balancing [1] (VLB) has attracted considerable research interest in recent years, as it guarantees 100% throughput for any traffic matrix satisfying hose model[2] constraints without obtaining precise traffic matrix. However, when the traffic variation tends to be gentle, the measurement of traffic demands between certain node pairs becomes feasible. Then it is likely to adjust the VLB scheme to access significant reduction in propagation delay. This paper studies adaptive load-balancing (ALB) in WDM networks provided that the traffic matrix is partially known during the operation time, which can be solved in two steps: 1) construct VLB architecture for a given WDM network; 2) route all the traffic demands in the architecture with a partially known traffic matrix. We first propose a novel heuristic termed MNC (Maximizing Network Capability) to structure a VLB WDM network. Then a heuristic ADT (Adding Direct Traffic) is proposed to minimize average propagation delay with traffic grooming [3][4][5]. Simulation results show that MNC costs less than the previous heuristics and ADT reaches notable improvement in propagation delay.
Keywords :
resource allocation; telecommunication network reliability; telecommunication traffic; wavelength division multiplexing; WDM mesh networks; adaptive load-balancing; adding direct traffic; maximizing network capability; partial traffic information; traffic matrix; Costs; Hoses; Mesh networks; Propagation delay; Robustness; Routing; Telecommunication traffic; Traffic control; WDM networks; Wavelength division multiplexing;
Conference_Titel :
Communications, Circuits and Systems, 2008. ICCCAS 2008. International Conference on
Conference_Location :
Fujian
Print_ISBN :
978-1-4244-2063-6
Electronic_ISBN :
978-1-4244-2064-3
DOI :
10.1109/ICCCAS.2008.4657836