An intelligent agent-based mobile resource reservation scheme

Abstract

This paper presents an Intelligent Agent-based Resource reservation aPproach (IARSVP) to support Quality of Service (QoS) aware packet transmissions for mobile networks. Mobile Intelligent Agents (MIAs) are characterized by their ability to move across wide-area networks, operate autonomously on foreign hosts, and perform tasks on behalf of the originating hosts. With MIAs, IARSVP can pre-allocate resources in advance for anticipated location an MN may visit next. Because MIAs carries the MN's W requirement and administration specification, and mobility association together with the necessary executable codes, they can adjust themselves in accordance with the network topology and current resource usage in locating forwarding points for the anticipated locations the MNs may visit next. Therefore, IARSVP can support route optimization and regional registration naturally, avoid redundant resource reservations made in common routes, and discover alternative route dynamically. Furthermore, MIAs can also perform location updates on behalf of MNs. Simulation results show that IARSVP outperforms Mobile RSVP and Hierarchical Mobile RSVP in terms of forced termination, reservation blocking, and session completion probabilirties.