A scheduling-based event correlation scheme for fault identification in communications network

Communications network has increased dramatically in size and complexity in the past few years. A typical network may consist of hundreds of nodes fro m various manufacturers with different traffic and bandwidth requirements. The increasing complexity poses serious problems to network management and control. As faults are inevitable, quick detection, identification and reco very are crucial to make the systems more robust and their operation more r eliable. This article proposes a novel event correlation scheme for fault i dentification in communications network. This scheme is based on the bi-lev el feedback queues scheduling policy. The causality graph model is used to describe the cause-and-effect relationships between network events. The use of scheduling policy makes the correlation process simple and fast. A simu lation model is developed to verify the effectiveness and efficiency of the proposed scheme. From simulation results, we notice that this scheme not o nly identifies multiple problems at one time but also is insensitive to noi se. As the time complexity of the correlation procedure is close to a funct ion of n, where n is the number of observed symptoms, with order O(n); ther efore, the on-line fault identification is easy to achieve. (C) 1999 Elsevi er Science B.V. All rights reserved.