A Framework of Effective Resource Allocation with Privacy Preservation in Connected Vehicles

Abstract

A vehicular fog network is an emerging paradigm adopted to facilitate delay-sensitive and innovative applications. Since vehicular environments are inherently dynamic, it is challenging to effectively utilize available resources. Often a centralized resource distribution model is adopted for effective resource utilization, but this comes with significant overhead. Distributed task offloading solutions are presented to address the issue; however, due to highly dynamic nature of the vehicular network the dis tributed solution cannot capture the complete state space of the vehicular system and usually end up in the uneven workload distribution. For that purpose, we proposed a distributed task offloading solution that adopts the notions of non-cooperative game theory and deep reinforcement learning to utilize the resources at the vehicular edge. This work is divided into two parts. In the First phase, this thesis proposes a distributed non-cooperative task offloading framework for efficient resource utilization. Here, vehicles with heterogeneous task requirements interact with one another without directly influencing the actions of the neighboring vehicles. That is, the framework allows vehicles to communicate with neighbors to gather contextual information to revisit their offloading decisions. The shared information includes resource type, task residence time, system cost, and offloading inference. The effectiveness of the framework is evaluated against baseline schemes in terms of service delay, transmission delay, system cost, delivery rate, and system efficiency. The results demonstrate 50% reduced task residence time, 83 Mb/s throughput which in turn contributes to an improved system utilization across the resource-sharing network. In the second phase this thesis introduces a deep reinforcement learning solution that efficiently learns task-offloading decisions at multiple tiers i.e., locally, and globally. The proposed work results in fast convergence due to its collaborative learning model among vehicles and fog servers. The local model runs at the vehicular nodes and the global model runs at the fog servers. To reduce network overhead, the models are learned locally; thus, limited information is shared across the network this reduces the communication overhead and improves the privacy of the agents. The proposed system is compared with the greedy and stochastic approaches in terms of residence times, cost, delivery rate, and utilization ratio. It has been observed that the proposed approach has significantly reduced the task residence time, end-to-end delay, and overall system cost.