TELECOM: Latency-Aware Load-Balanced Task Offloading in Vehicular Fog Computing

Abstract

Task execution latency in Internet of Things (IoT) frameworks remains a critical challenge. In this regard, fog computing was introduced as a suitable strategy to bring data processing centers closer to the edge and reduce latency. One application of fog computing is vehicular fog computing (VFC), whose primary objective is to help mitigate computational limitations of vehicles connected to this structure and their lack of full environmental awareness. With the constant expansion of communication networks, the widespread implementation of VFC still faces various challenges. One of these challenges is the increased distance of vehicles from the execution location of assigned tasks, which leads to the migration phenomenon. Another challenge is managing the significant workload offloaded to the fog nodes while meeting the execution time requirements and offering better availability levels. In this research, we have proposed TELECOM, a task offloading algorithm that takes a unique approach and introduces a communication-aware interzone service model to reduce the dependency on explicit task migration and maintain service continuity under mobility-induced zone changes. The mobility model for edge and vehicular fog nodes is obtained using real-world scenarios. The goal is to reduce latency and balance the task execution so that tasks are distributed proportionally between nodes without causing overload. TELECOM improved the network’s latency by an average of 37.72% and load balancing by an average of 17.94% compared to other methods. In addition, availability was also improved by an average of 33.73% and 32.35% when considering 2- and 4-s thresholds, respectively.