A unified approach to scheduling, access control and routing for ad-hoc wireless networks

An integrated method to solve the capacity allocation (at both the node level and the flow level) and the routing problems in wireless ad-hoc networks is proposed. The algorithms manage delays due to localized congestion at nodes and packet loss due to error prone wireless links to provide improved end to end delay/throughput. A hierarchical scheme for node level access control and flow level scheduling is defined. At each level, the scheduling discipline is a modified weighted fair queuing scheme that uses a weight estimate derived from the link quality of the route (called link error adjusted rate (LEAR)) and a route assignment determined by the routing algorithm. Routing is based on a shortest path algorithm that uses a delay estimate as a link distance metric, which in turn depends on capacity allocations from the hierarchical scheduling. The capacity allocations to the nodes and flows and the route assignments for the flows are iterated ever frame. The adaptability and interactions among protocol layers in this method allows the network to respond to random channel errors and congestion arising from flows. The scheduling and routing algorithms are distributed