Abstract
As networks become all-pervasive the importance of efficient information gathering for purposes such as monitoring, fault diagnosis, and performance evaluation can only increase. Extracting information out of largescale, dynamic networked systems is becoming increasingly difficult. Distributed monitoring systems based on static object technologies such as CORBA and Java-RMI can cope with scalability problems only to a limited extent. They are not well suited to monitoring systems that are both very large and highly dynamic because the monitoring logic, although distributed, is statically pre-determined at design time. The paper presents an active distributed monitoring system based on mobile agents. Agents act as area managers which are not bound to any particular network node and can sense the network, estimate better locations, and migrate in order to pursue location optimality. Simulations demonstrate the capability of this approach to cope with large-scale systems and changing network conditions. The limitations of our approach are also discussed in comparison to more conventional monitoring systems.