| 摘要 |
<p>Difficulties with prior disrupted communications restoration systems are overcome by employing for each node of a communications network a restoration processor in which predetermined functions of the restoration process are managed by respective ones of a plurality of manager modules. The manager modules are arranged to be directly interconnected only with a so-called restoration manager that coordinates the operations of the modules and provides for their interconnection needs. In a preferred embodiment, four manager modules, a) a user interface manager module, b) a digital cross-connect system (DCS) manager module, c) a distributed communications manager module and d) a core algorithm manager module, are interconnected and coordinated by the restoration manager. Additional robustness to the restoration process is provided by way of enhancements to the functionality of the core algorithm manager module. These enhancements include: a) improved glare processing (resolution of contention for a particular resource), b) "broadcast" of takedown messages, c) elimination of the need for time-outs during automatic restoration, d) a "peek ahead" routine that immediately takes down a needed resource that is already on a list to be taken down, e) use of a "minimax factor" and f) a "quick fix" technique for use in cases of limited failure of circuits between two nodes. The restoration process can also be cascaded so that it sequentially restores one level at a time, the levels being defined by the hierarchy of digital transmission rates employed in the network. <IMAGE></p> |
