Local Area Discovery

Discovery is a mechanism by which agents and infrastructural services can find each other in ad-hoc networks. Discovery enables agents/services/servers to auto-configure with reference to each other, thus facilitating interactions. With Discovery, agent systems no longer require the tedious manual configuration of system locations, host names, and port numbers, before being able to operate.

Agents and infrastructural services that use Discovery can implement advanced behaviors to respond to their changing environments. Servers can locate each other, share information and coordinate their operations. Client agents can learn of new servers/services as they become operational, and automatically re-register/re-advertise with the new system components, to ensure registration availability and increased system fault tolerance. Agents can also be programed to survive initialization within incomplete environments, and to finish initialization once all needed peer systems are available, online, and active. This Discovery feature reduces the problems associated with previous implementations that relied on strict, order-specific startup sequences.

Our current implementation of Discovery is based on the Simple Service Discovery Protocol (SSDP) that is part of the Universal Plug-n-Play (UPnP) specification.

Discovery in ANS

Discovery has been integrated into the RETSINA Agent Name Service (ANS) to help agents locate ANS servers. On startup, both the ANS client and server load a list of ANS servers from a file. They use the entries in this memory-cached-list to connect to and conduct conversations with ANS servers. Before implementation of Discovery, the list was static. Discovery causes the list to be dynamically expanded and contracted as ANS servers come and go from the network. Agents are now registered on multiple servers, and servers share registration and unregistration information about agents.

Publications

K. Sycara, M. Paolucci, A. Ankolekar and N. Srinivasan, "Automated Discovery, Interaction and Composition of Semantic Web services," Journal of Web Semantics, Volume 1, Issue 1, September 2003, pp. 27-46.

M. Paolucci, K. Sycara, T. Nishimura, and N. Srinivasan, "Using DAML-S for P2P Discovery," in Proceedings of the First International Conference on Web Services (ICWS'03), Las Vegas, Nevada, USA, June 2003, pp 203- 207 .

B. Langley, M. Paolucci, and K. Sycara, "Discovery of Infrastructure in Multi-Agent Systems." In Agents 2001 Workshop on Infrastructure for Agents, MAS, and Scalable MAS.

Robotics Institute Project Page