Quality of service routing in ad hoc networks

Current generation routing protocols for ad hoc networks typically use shortest path routing. Examples include on-demand routing protocols such as DSR and AODV. All these protocols can support only the best-effort data traffic, but the quality of service (QoS) requirements, such as delay, packet loss and bandwidth requirements are not supported. In the wireline networks such as ATM, QoS routing has been studied extensively. Due to dynamic nature of ad hoc networks owing to mobility and varying radio link conditions, the available state of information is never precise, and consequently, we cannot directly apply the traditional wireline QoS routing algorithms. Soft QoS without hard guarantees is explored here. Previous work in QoS routing in ad hoc networks is reported by Chen and Nahrstedt (see IEEE journal on Selected Areas in Communication, 1999), where delay and bandwidth constrained QoS routing were studied separately. The uncertainty was handled using a deterministic model. The algorithms developed in this paper accommodate imprecise state information. Fuzzy logic is a well recognized technique for modeling imprecision. A rule based fuzzy logic model is used to describe imprecise state information. Delay constraints are considered here in finding feasible routes (paths). Among feasible paths, optimal path (least cost path) is found using hop count as a performance measure. In case of re-routing, sub-optimal paths are found using the same performance measure.