Research On Wavelength Buffering Congestion Control Problem Under New Network Model In IONs

With the multiplying demand of bandwidth due to the explosion of data service, and the WDM technology brings new bandwidth multiplexing, Intelligent optical networks(IONs) emerged and have been developed rapidly. The network congestion problem is an important indicator to impact the performance of network. How to prevent and control the network congestion is one of the core problems in IONs. Partial-sparse-Limited network model can reduce the network blocking probability and minimize the cost of the network.This thesis studied the congestion control routing algorithm in the IONs and put forward some innovative designs. Limited wavelength resource, Wavelength Continuity Constraint(WCC) and, Link Cascade Effect(LCE) are the three classes about the causes of network congestion. First of all, In order to eliminate the wavelength continuity constraint, this thesis designed a placement of wavelength converter algorithm which is based on Partial-sparse-Limited network model(for short PCPL). Secondly, since in IONs, one of the important causes of congestion is the imbalance of network traffics, according to the characteristics of IONs, this thesis designed a congestion control routing algorithm which is based on PCPL and Wavelength Buffering(for short PCPL_WB). Thirdly, uneven utilization of network resources is also one of the causes of congestion and it is cased by the uneven spatial distribution of network traffic requests and the network topology. This thesis designed Wavelength Buffering and Soft Preemption algorithm which is based on PCPL algorithm(for short PCPL_WBSP). And the load weighted method and exchange routing method are used to solve the congestion location and reroute calculation two key issues in Soft Preemption (SP) scheme. SP scheme can protect the Quality of Service (QoS) of the existing business better. Only when the traffic which is preempted can find a new path for transmission, then the transmission will interrupt and release the resources for a new traffic.Finally, this thesis simulated these three congestion control algorithms, and then compared and analyzed the performance with other solutions. The simulation results showed that PCPL algorithm achieved an outstanding performance of the network improvement compared with the existing SPWC algorithm, and it can be up to 20%; PCPL_WB algorithm and PCPL_WBSP algorithm reduced the rate of network congestion and improved the network resource utilization effectively. Compared with PCPL algorithm, the network resource utilization of PCPL_WB algorithm and PCPL_WBSP algorithm were reduced by 13% and 10% respectively.