Efficient offloading in disaster-affected areas using unmanned aerial vehicle-assisted mobile edge computing: A gravitational search algorithm-based approach

The collection and processing of real-time data from a disaster-affected area is challenging. Unmanned aerial vehicles (UAVs) can efficiently gather the data and then transfer it to the edge servers (ESs) to timely initiate the rescue process. Consideration of energy and delay in a UAV-assisted edge network is very important, as both the UAVs and the smart mobile devices (SMDs) in the network have energy constraints and low processing capacity. Offloading is a promising technique to preserve the precious energy of the SMDs. In this research, gravitational search algorithm (GSA)-based offloading is presented for UAV-assisted mobile edge computing (MEC)-enabled disaster-affected areas. The problem is first mathematically formulated and shown to be computationally hard. Efficient encoding of agents (solution vectors) is given for the offloading problem. Fitness function is designed by considering the energy, delay, and load balancing of the ESs. The proposed GSA is executed by considering multiple disaster scenarios, and its performance is compared with other evolutionary algorithms (EAs) like the genetic algorithm (GA), particle swarm optimization (PSO), and fireworks algorithm (FWA). It has been observed that the GSA outperforms the other EAs in almost all the considered experiment scenarios. GSA claims a 30%-40% improvement for delay, 3%-5% for energy consumption, and more than 40% for load balancing. Statistical and convergence analyses are also conducted. The convergence of the GSA is found to be faster than that of the other EAs.