Neighbor predictive adaptive handoff algorithm for improving mobility management in VANETs

Abstract A vehicular ad-hoc network (VANET) was employed in commercial, road-safety, and entertainment applications due to its accessibility. Applications and services were shared with the service providers (SP) over mobile nodes to any destination without special infrastructure. The mobility pattern of the nodes was independent, and the acceleration remained unpredictable, which led to service failures and information drop-outs. Resuming communication requires the flooding of additional control messages, which exploits the network resource in a shorter period. This paper introduces the neighbor predictive adaptive handoff (NPAH) algorithm for ensuring seamless communication, regardless of the application service time. NPAH discovers weak communication links in the service, which persist through the least resource dependent distance-based neighbor discovery. The selected neighbors are characterized by distance and minimum cost exploitation using the Q-learning technique. The process of learning decides the handoff of a vehicle based on storage utilization and cost factors. The results demonstrated the effectiveness of the NPAH algorithm by achieving less packet loss, shorter outage times, and improving the delivery factor.