QoS-Aware Multi-Hop Routing Framework for Reliable Wireless Communication in Dynamic Ad Hoc Networks

Abstract

Mobile Ad Hoc Networks (MANETs) enable infrastructure-free wireless communication by allowing mobile nodes to cooperate as routers over multiple hops. Their flexibility is valuable in emergency response, tactical communication, temporary field networks, disaster recovery, and mobile sensing, but reliable Quality of Service (QoS) remains difficult because topology, link quality, queue occupancy, available bandwidth, and residual energy vary continuously. Conventional shortest-path routing can select a route that is topologically valid but unsuitable for delay-sensitive or bandwidth-demanding traffic, while multipath protocols improve resilience without necessarily coordinating QoS admission, path quality, and fast recovery. This paper proposes a QoS-Aware Multi-Hop Reliability Framework (QAMHRF) that extends on-demand routing with multi-metric path discovery, QoS feasibility filtering, adaptive path scoring, and a low-overlap backup-route mechanism. Each candidate path is characterized by bottleneck bandwidth, accumulated delay, queue load, residual energy, hop count, and link-stability estimates. Hard QoS constraints are checked before optimization; feasible paths are then ranked using traffic-class-dependent weights so that real-time flows emphasize bandwidth and delay while best-effort traffic gives greater importance to energy and stability. A backup path is retained to reduce the cost of mobility-induced route breaks. A reproducible Python-based stochastic MANET evaluation with 60 mobile nodes, 20 flows, and maximum speeds of 5, 10, and 20 m/s compares the proposed approach with AODV and AOMDV. Across the three mobility levels, QAMHRF achieves an average packet delivery ratio of 93.31%, average delay of 28.93 ms, throughput of 694.10 kbps, QoS satisfaction of 91.22%, and energy expenditure of 10.23 mJ per delivered packet. Relative to AODV, the proposed framework improves average packet delivery by approximately 9.76 percentage points and reduces average end-to-end delay by approximately 29.25%. These preliminary results support the value of combining QoS admission, multi-objective path selection, and rapid backup-route switching for reliable communication in dynamic ad hoc networks.