A location based prediction service protocol for VANET city environment

Abstract

Location-based service is used in Vehicular Ad hoc Networks (VANETs) to locate node’s position before the start of any communication. The existing location services proposed for VANET do not distribute the load on multi servers, and some of them do not consider stable nodes for selecting location servers. Additionally, predicting node’s locations using grey prediction model accuracy is affected by nodes acceleration in VANET. This paper proposes a Vehicular Quorum Prediction-based Location Service protocol (VQPLS), which was designed for urban area topology and which utilised node’s information such as distance to intersection centre point and speed in selecting stable location servers. Formation of quorum of location servers was done by main location server by nominating some other nodes located at the intersection based on their moving directions. The quorum of location servers was used to distribute the load on multi servers and for fault tolerance. A hybrid of Grey Prediction Model and Alpha-Beta-Gamma Filter (GP-ABGF) was adopted in the VQPLS to accurately predict the next position of nodes that were moving away from the intersection. The GP-ABGF mitigated effects of VANET nodes acceleration and irregular update time intervals on grey prediction model accuracy. Prediction algorithm also filtered the noise data and produced accurate location of destination and overcame the problem of outdated locations stored by quorum systems. The VQPLS showed good performance compared with other protocols in reducing overhead of control packets and high delivery of packets to destinations, and it reduced the end-toend delay of routing packets.