Energy efficient processing in opportunistic vehicular edge clouds

The Cisco Visual Networking Index of 2019 reports that more than six billion Machine-to-Machine (M2M) connections were added in 2017 and the number of connections is expected to grow by more than 50% by 2022. The proliferation of connected devices is accompanied by rapid growth in the generated traffic between the edge layer and data centres, and therefore is expected to lead to significant increase in the power consumption of the network infrastructure. This calls for new architectural designs capable of reducing the traffic congestion and power consumption in the network. At the same time, vehicles are going through a huge revolution in term of their on-board units and processing capabilities producing a new promising framework concept. This concept is a consequence of the integration of vehicles and cloud computing, referred to as Vehicular Edge Cloud (VEC).
This thesis investigates distributed processing in VECs, where a group of vehicles in a car park, at a charging station or at a road traffic intersection, cluster and form a temporary vehicular cloud by combining their computational resources in the cluster. We investigated the problem of energy efficient processing task allocation in VEC by developing a Mixed Integer Linear Programming (MILP) model to minimise power consumption by optimising the allocation of different processing tasks to the available network resources, cloud resources, fog resources and vehicular processing nodes resources. Three dimensions of processing allocation were investigated. The first dimension compared centralised processing (in the central cloud) to distributed processing (in the multi-layer fog nodes). The second dimension introduced opportunistic processing in the vehicular nodes with low and high vehicular node density. The third dimension considered non-splittable tasks (single allocation) versus splittable tasks (distributed allocation), representing real-time versus non real-time applications respectively. The results revealed that a power savings up to 70% can be achieved by allocating processing to the vehicles. However, many factors have an impact on the power saving percentage such the vehicle capacities, vehicles density, workload size, and the number of generated tasks. It was observed that the power saving is improved by exploiting the flexibility of task splitting among the available vehicles. In addition to the processing allocation problem, this thesis investigated the software matching problem in VEC. The vehicles involved may not be equipped with the full set of software needed to process the tasks requested. Moreover, as vehicles in VEC represent processing at the edge layer, we studied the impact of edge processing on the propagation and queuing delay in a joint optimisation modelling intended to minimise both power consumption and delay. Our investigation showed a significant impact on the processing allocation decision and therefore, the power consumption, attributed to the location of the processing node and the service rate of the network controller.