Scholarlink

<p>Identifying and localizing rogue nodes in Vehicular ad hoc network (VANET) is acquiring traction, as it has the potential to significantly increase the network’s lifespan and value. Several localization approaches have been developed for precisely estimating unknown nodes. It is challenging to discover acceptable network settings for node localization during the network configuration procedure with the needed accuracy in a short amount of time. The core objective of the suggested technique is to locate the source of routing attacks in VANETs by using a binary ant colony intelligence-based discriminative deep belief network (BAC-IDDBN) for improving cyber security. These methods are used to determine the optical sensor location and its distance from others calculate the average localization accuracy and identify malicious nodes. Collect various attacks in wireless sensor network-detection system (WSN-DS) datasets, including Time Division Multiple Access (TDMA) scheduling, Black hole, Flooding and Gray hole. Applying the Z-score normalization approach help to reduce noise in the WSN-DS dataset. Extracting features from the pre-processed data using Linear Discriminant Analysis (LDA). The proposed method is used to identify anomalies or intrusion detection in VANET localization data. This integration enables real-time identification and response, enhancing VANET security. The finding shows that the suggested BAC-IDDBN approach accuracy (98%), recall (96.65%), precision (97%) and f1- score (97.37%). The proposed method run in python platform. The outcomes of the simulation that the suggested method can effectively localize the unknown node with a better degree of precision, the ability of BAC-IDDBN to protect VANET s against cyber attacks, developing an opportunity for its use in real-world implementations.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Intrusion Detection and Localization Using Deep Learning Approaches in VANET Environments

  • 期刊论文

  • 发表日期：

    2026年2月25日

查看全文
  • M. Vijayakumar,
  • Elangovan Muniyandy,
  • Gayatri Mirajkar,
  • Krovvidi S. B. Ambika,
  • S. P. M. Kannan,
  • R. Venkatesh,
  • R. Mohan Kumar

摘要

Identifying and localizing rogue nodes in Vehicular ad hoc network (VANET) is acquiring traction, as it has the potential to significantly increase the network’s lifespan and value. Several localization approaches have been developed for precisely estimating unknown nodes. It is challenging to discover acceptable network settings for node localization during the network configuration procedure with the needed accuracy in a short amount of time. The core objective of the suggested technique is to locate the source of routing attacks in VANETs by using a binary ant colony intelligence-based discriminative deep belief network (BAC-IDDBN) for improving cyber security. These methods are used to determine the optical sensor location and its distance from others calculate the average localization accuracy and identify malicious nodes. Collect various attacks in wireless sensor network-detection system (WSN-DS) datasets, including Time Division Multiple Access (TDMA) scheduling, Black hole, Flooding and Gray hole. Applying the Z-score normalization approach help to reduce noise in the WSN-DS dataset. Extracting features from the pre-processed data using Linear Discriminant Analysis (LDA). The proposed method is used to identify anomalies or intrusion detection in VANET localization data. This integration enables real-time identification and response, enhancing VANET security. The finding shows that the suggested BAC-IDDBN approach accuracy (98%), recall (96.65%), precision (97%) and f1- score (97.37%). The proposed method run in python platform. The outcomes of the simulation that the suggested method can effectively localize the unknown node with a better degree of precision, the ability of BAC-IDDBN to protect VANET s against cyber attacks, developing an opportunity for its use in real-world implementations.