EV Assisted Resource Sharing Framework Through Stream Processing

Abstract

The rapid growth of the Internet of Things (IoT), and the influx of real-time data streams from diverse sources have catalyzed the need for efficient and scalable stream data pro cessing techniques. This thesis introduces an innovative paradigm that harnesses the untapped potential of electric vehicles (EVs) to enhance the scalability and efficiency of stream data processing systems. In resource-constrained environments such as edge devices and IoT deployments, conventional computing infrastructures often face chal lenges in meeting the demands of processing continuous and high-velocity data streams. EVs, equipped with robust computational capabilities and rechargeable batteries, offer a unique opportunity to alleviate these challenges by acting as supplementary process ing nodes. The research begins by comprehensively reviewing the landscape of stream data processing, edge computing, and vehicular networks. Existing stream processing frameworks and their challenges are analyzed, paving the way for the exploration of an innovative approach to address these limitations. The proposed framework integrates EVs into the stream data processing architecture, presenting a novel approach for task offloading and collaborative stream processing. By leveraging V2V (vehicle-to-vehicle) communication, EVs within the network can dynamically share computational work loads, thereby enhancing the scalability and responsiveness of the entire system. A detailed network model is formulated to illustrate the interactions and resource sharing mechanisms among EVs, further substantiated by an in-depth system model that quanti fies energy consumption, task execution times, and offloading strategies. Through metic ulous simulation and performance evaluation, the novel approach is compared against conventional methods. Results reveal a substantial improvement in task completion rates, marked reduction in failure rates, and optimized energy utilization when leverag ing EVs for stream data processing. The study not only underscores the efficiency of the proposed approach but also highlights its potential to revolutionize real-time data analysis in dynamic vehicular environments. This thesis contributes to the fields of stream data processing, edge computing, and vehicular networks by introducing a pioneering solution that leverages EVs to enhance the scalability, efficiency, and resilience of stream data processing systems. The findings underscore the viability of incorporating EVs as computational resources, opening new avenues for addressing the challenges posed by the burgeoning influx of real-time data streams in the era of IoT.