Browsing by Author "Kurunathan, Harrison"

Now showing 1 - 3 of 3
  • Thumbnail Image
    Item
    Adaptive Intrusion Mitigation in Software-Defined Vehicles Using Deep Reinforcement Learning
    (2025) Kurunathan, Harrison; Ismail Ali, Hazem; Javanmardi, Gowhar; Eldefrawy, Mohamed; Gutiérrez Gaitán, Miguel José; Robles, Ramiro; Yomsi, Patrick; Tovar, Eduardo
    Software-defined vehicles (SDVs) leverage vehicle-to-everything (V2X) communication to enable advanced connectivity and autonomous driving capabilities. However, this increased interconnectivity also exposes them to cyber threats such as spoofing, denial-of-service attacks, and data manipulation, making intrusion detection systems (IDS) essential for ensuring SDV security and reliability. In this work, we propose a novel intrusion mitigation approach that integrates Advantage Actor-Critic (A2C) reinforcement learning with a Long Short-Term Memory (LSTM) network to detect anomalies and intrusions in V2X communications. The LSTM component captures temporal dependencies in V2X data, enhancing the model's ability to identify emerging attack patterns, while the A2C framework dynamically adjusts defensive actions, including flagging, blocking or monitoring traffic, based on evolving threat levels. Experimental results demonstrate the model's effectiveness, achieving high detection accuracy and sensitivity. Additionally, we analyze how the system adapts over time, becoming more confident in its decision-making and optimizing security enforcement. This work enhances SDV cybersecurity by introducing a learning-based adaptive intrusion response system aiming at mitigating threats in highly dynamic vehicular networks.
  • Thumbnail Image
    Item
    Q-light: Q-learning enabled VLC network routing
    (ACM, 2024) Kurunathan, Harrison; Robles, Ramiro; Gutiérrez Gaitán, Miguel José; Ravichandran, Indhumathi; Tovar, Eduardo
    Visible Light Communication (VLC) is an emerging technology that uses light sources such as light-emitting diodes (LEDs) and lasers for data transmission. This is enabled by the IEEE 802.15.7 protocol, which supports deterministic communication through its guaranteed timeslot mechanism. VLC networks provide several advantages in terms of high bandwidth, security, and immunity to electromagnetic interference. Most of the works in VLC are centred around point-to-point communication and do not necessarily provide emphasis on the challenges due to blockage, deafness, or hidden node problems. In this work, we present a machine learning (ML)-optimized routing protocol (Q Light) that improves network throughput and reliability through the selection of efficient routes for the IEEE 802.15.7 protocol.
  • Thumbnail Image
    Item
    Worst-Case Response Time of Mixed Vehicles at Complex Intersections
    (2024) Reddy, Radha; Almeida, Luis; Kurunathan, Harrison; Gaitan, Miguel Gutierrez; Santos, Pedro M.; Tovar, Eduardo
    Operating autonomous vehicles (AVs) and human-driven vehicles (HVs) at urban intersections while observing requirements of safety and service level is complex due not only to the existence of multiple inflow and outflow lanes, conflicting crossing zones, and low-speed conditions but also due to differences between control mechanisms of HVs and AVs. Intelligent intersection management (IIM) strategies can tackle the coordination of mixed AV/HV intersections while improving intersection throughput and reducing travel delays and fuel wastage in the average case. An endeavor relevant to traffic planning and safety is assessing whether given worst-case service levels can be met. Given a specific arrival pattern, this can be done via the worst-case response time (WCRT) that any vehicle experiences when crossing intersections. In this research line, this paper estimates WCRT upper bounds and discusses the analytical characterization of arrival and service curves, including estimating maximum queue length and associated worst-case waiting time for various traffic arrival patterns. This analysis is then used to compare six state-of-the-art intersection management approaches from conventional to intelligent and synchronous. The analytical results show the advantage of employing a synchronous management approach and are validated with the vehicles floating car data (timestamped location and speed) and simulations carried out using SUMO.