Figure 1

In this project, the use of intelligent agents was explored for the purpose of managing networks. The research was conducted at Bell Laboratories that at that time served as the research arm of Lucent Technologies [1].


Figure 2

The purpose of the research was to expand the known paradigms for network management that employed entities called agents (in SNMP, CMIP, for example; shown in the Figure 1), but these agent were just communicating processes. The goal was to design an architecture that would provide another layer on top of existing frameworks (that also included distributed middleware like CORBA) that would provide intelligent tools to minimize outages of communication networks (Figure 2).

The approach taken included utilization of the emerging multi-agent standards (FIPA) that targeted intelligent communication layer. The Bell Laboratories team was working at the higher level that provided plug ability for intelligent agents that could detect and potentially fix emerging issues (Figure 3).


Figure 3


  1. Bieszczad, A., Biswas, P, Buga, W., Malek, M. and Tan, H., (1999) Managing Heterogeneous Networks with Intelligent Agents, Bell Laboratories Technical Journal, October 1999 (best paper of 1999 award).