Driver’s Emotional State-Based Data Anomaly Detection for Vehicular Ad Hoc Networks

Abstract

The effectiveness of data for safe driving in the Vehicular Ad Hoc Network is the basis to improve vehicle safety. Unlike traditional information systems, data anomaly analysis of vehicle safety driving faces the diversity of data anomalies(associated with virus invasion, failed sensors, and noise jamming plague), and the randomness and subjectivity of the driving behavior. How to combine the characteristics of the vehicular network data and the sentiment analysis of the driver to provide effective real-time data anomaly detection has become an important issue in the application of the vehicular network. Aiming at the cooperative vehicular networks, this paper adopts a method based on driver's emotional state, and proposes Anomaly Detection Based on Driver's Emotional State (EAD) algorithm to realize real-time detection of data related to safe driving. Firstly, this paper defines the emotion recognition coefficient R_de and designs a driver's emotion quantification model to realize the characterization of the driver's driving style. Secondly, based on the emotion quantization model and vehicle driving state information, the data anomaly detection algorithm is designed based on Gaussian mixed model(GMM). Finally, combined with the application scenarios of cooperative vehicular networks, we conduct extensive experiments on real data set(NGSIM) to prove the EAD algorithm has good performance.