Vehicular Delay-Tolerant Networks for Smart Grid Data Management Using Mobile Edge Computing

With the widespread popularity and usage of ICT around the world, there is increasing interest in replacing the traditional electric grid by the smart grid in the near future. Many smart devices exist in the smart grid environment. These devices may share their data with one another using the ICT-based infrastructure. The analysis of the data generated from various smart devices in the smart grid environment is one of the most challenging tasks to be performed as it varies with respect to parameters such as size, volume, velocity, and variety. The output of the data analysis needs to be transferred to the end users using various networks and smart appliances. But sometimes networks may become overloaded during such data transmissions to various smart devices. Consequently, significant delays may be incurred, which affect the overall performance of any implemented solution in this environment. We investigate the use of VDTNs as one of the solutions for data dissemination to various devices in the smart grid environment using mobile edge computing. VDTNs use the store-and-carry forward mechanism for message dissemination to various smart devices so that delays can be reduced during overloading and congestion situations in the core networks. As vehicles have high mobility, we propose mobile edge network support assisted by the cloud environment to manage the handoff and the processing of large data sets generated by various smart devices in the smart grid environment. In the proposed architecture, most of the computation for making decisions about charging and discharging is done by mobile devices such as vehicles located at the edge of the network ( also called mobile edge computing). The computing and communication aspects are explored to analyze the impact of mobile edge computing on performance metrics such as message transmission delay, response time, and throughput to the end users using vehicles as the mobile nodes. Our empirical results demonstrate an improved performance 10-15 percent increase in throughput, 20 percent decrease in response time, and 10 percent decrease in the delay incurred with our proposed solution compared to existing state-of-the-art solutions in the literature.