Mobile Intelligence

15.8 CONCLUSIONS

We have proposed an agent-oriented architecture for the provision of enhanced services for users of personal communications systems and ubiquitous computing, as well we have focused on the use of AgentSpeak(L) for the double role of implementation and policy language, following important developments in the area of agent-oriented programming and multiagent systems. Communication and telephony are areas of application where, to the best of our knowledge, the BDI architecture has not been used, yet their complexity requires the sophisticated control that can be readily expressed in AgentSpeak(L). More than other BDI models, AgentSpeak(L) has an exact notation, as well as clear and precise logical semantics, which resulted in the successful implementation of its abstract interpreter [27]. AgentSpeak(L) enables an elegant specification of the BDI agents and allows to model how agents can search and expand planned actions in order to select alternatives. In addition, it allows interagent negotiation for the resolution of interactions.

Another contributions of our work is our concept of “Consolidated Presence Information” [2]. Consolidating all the information that is available for users and their environment provides a unified view of the status of the user at a given time. The architecture allows real-time use of this information, thus simplifying the implementation of a large spectrum of complex services.