科研之友
微信
新浪微博
Facebook
分享链接
摘要
Previously, an optical code-division multiplexing (OCDM)-based network architecture was proposed to improve the wavelength utilization and to provide finer bandwidth granularities to users. By this technology, different channels using distinct optical codes (OCs) can be multiplexed onto the same wavelength, in which an OC is considered as the basic unit in lightpath provisioning. In the ideal case, multiaccess interference (MAI) inherent to the OCDM technology is assumed to be removed completely at intermediate nodes and cannot be propagated or accumulated along the lightpath. However, since no optical-electrical (O/E) or electrical-optical (E/O) conversion is allowed in transparent OCDM-based optical networks, the MAI cannot be removed completely at intermediate nodes with current all-optical regeneration techniques. As a result, the residual MAI may be propagated and accumulated along the lightpath and affect other active lightpaths carried by the same wavelength in the network. The affected active lightpaths may build unintended cycles along which the MAI is accumulated. Furthermore, this MAI keeps increasing when the lightpaths traversed by the cycle are active, which deteriorates the lightpath signal quality. Since this deterioration may eventually result in unacceptable signal quality and service disruption, the phenomenon caused by the MAI is termed as cycle attack in this paper. The explanations of the MAI propagation mechanism and the cycle attack problem are given. A depth-first search (DFS)-based algorithm is proposed to diagnose such cycle attacks under dynamic traffic conditions. The numerical results show that our DFS-based cycle attack diagnostic algorithm enables one to detect cycle attacks effectively, and the two-way resource reservation method associated with heuristic wavelength assignment is shown to mitigate the blocking performance degradation due to cycle attacks greatly with some proper wavelength and OC configuration.
- 出版日期2008-8
