Adaptive and intelligent path discovery on-demand for wireless networks using service composition

Route setup and discovery require low communication overhead in portable computing and wireless environments. We propose an adaptive and intelligent route discovery on-demand approach that introduces a service composition concept. Since service composition along a network is like a path in a multicast tree, path discovery can be regarded as service composition. Hence, a route-composition is proposed to combine existing services (paths) into candidate complex services (paths) called on-demand route compositions. The proposed route setup mechanism has the following characteristics: (1) on-demand route setup to avoid periodical multicast overhead for member discovery; (2) a directed service graph to represent a transparent route-composition; (3) candidate routes are found adaptively and intelligently; (4) the best route among candidate routes is determined by the max-flow and min-cut algorithm. The simulation shows that for a newly joining multicast receiver, the proposed method can avoid periodical multicast-related control overhead for member discovery, and improve the throughput for multicast communication from 29% to 74% as compared with the nearest neighbor algorithm.