A case-based reasoning approach in network traffic control

This paper presents a methodological approach based on knowledge acquisition (KA) and case-based reasoning (CBR) to design a decision support system based on solved incident cases reusing in continuous control domain. The specific domain to which we are applying KA and CBR is the control of traffic flow in the standard urban public switched telephone network of the Ile de France. Main research objective is to place at the network managers disposal an intelligent tool designed to capitalize and reuse past solved incident cases in order to assist them in their diagnosis and decision making activities. This methodological approach is developed within an interdisciplinary research cooperation supported by the CNET (France Telecom Research Center) and the CNRS (French National Board for Scientific Research)