Rank Criteria Improved Confidence-based centroid scheme for Non Line of Sight node localizations in vehicular networks

Abstract

The location verification of the vehicles interacting during the process of communication needs to be cooperatively determined under Non Line Of Sight (NLOS) situations for facilitating risk free environment with the least degree of congestion. The vehicular nodes in NLOS conditions possess the possibility of introducing channel congestion and broadcast storm into the network either intentionally or unintentionally during the event of emergency message dissemination in the vehicular network. Thus, the NLOS vehicular nodes of the network need to be detected through direct interaction and neighbourhood collaboration such that the rate of emergency message dissemination is sustained to the maximum degree. In this paper, a Rank Criteria Improved Confidence-based Centroid Scheme (RCICCS) is proposed for potential localization of NLOS nodes. This proposed RCICCS scheme uses an integrated cost, computed using the primitive cost and penalty cost in order to increase the effectiveness in localizing NLOS nodes. Thus effective location of NLOS nodes is determined based on rank criterion-based neighbor confidence measure that are iteratively enhanced during the process of perturbation using the method of gradients. The experimental analysis of the proposed RCICCS scheme confirmed a remarkable enhancement rate of emergency message delivery and neighbourhood awareness under NLOS conditions.

Keywords

• Neighbour confidence
• Non Line of Sight nodes
• Stochastic Approximation
• Simultaneous Perturbation
• Adaptive Penalty Function