Selection of PCEs' location in multi-domain optical networks  

作　　者：XIE Wen-jun HUANG Shan-guo ZHAO Yong-li LIU Xin GU Wan-yi 

机构地区：[1]State Key Laboratory of Information Photonics and Optical Communications, Beijing University of Posts and Telecommunications [2]Huawei Technologies Co. Ltd

出　　处：《The Journal of China Universities of Posts and Telecommunications》2013年第6期62-68,共7页中国邮电高校学报（英文版）

基　　金：supported by the National Basic Research Program of China (2010CB328202, 2010CB328204, and 2012CB315604);the Hi-Tech Research and Development Program of China (2012AA011302);the Beijing Nova Program (2011065);the RFDP Project (20120005120019);the Fundamental Research Funds for the Central Universities (2013RC1201);the Fund of State Key Laboratory of Information Photonics and Optical Communications (BUPT)

摘　　要：Path computation elements （PCEs） are employed to compute end-to-end paths across multi-domain optical networks due to the advantages of powerful computation capability. However, PCEs＇ location selection is still an open problem which is closely related to the communication overhead. This paper mainly focuses on the problem of PCEs＇ location selection to minimize the overall communication overhead in the control plane. The problem is formulated as a quadratic integer programming （QIP） model, and an optimal decision rule is gained from the solution of the QIP model. Then based on the decision rule, a distributed heuristic algorithm is proposed for dynamic network scenario. Simulation results demonstrate the benefit and the effectiveness of our proposed approach by comparing it with random selection policy.Path computation elements （PCEs） are employed to compute end-to-end paths across multi-domain optical networks due to the advantages of powerful computation capability. However, PCEs＇ location selection is still an open problem which is closely related to the communication overhead. This paper mainly focuses on the problem of PCEs＇ location selection to minimize the overall communication overhead in the control plane. The problem is formulated as a quadratic integer programming （QIP） model, and an optimal decision rule is gained from the solution of the QIP model. Then based on the decision rule, a distributed heuristic algorithm is proposed for dynamic network scenario. Simulation results demonstrate the benefit and the effectiveness of our proposed approach by comparing it with random selection policy.

关 键 词：multi-domain optical networks path computation element PCE location selection QIP model 

分 类 号：TN929.1[电子电信—通信与信息系统]