Performance improvement of data offloading using Krill herd optimization algorithm

The new pattern of user traffic generation in recent years and the variety of traffic services including data, voice and video have led to a large load in the access of cellular networks. One of the promising solutions in the field of reducing this traffic load is data offloading, which is based on exploiting the unused bandwidth of wireless technologies overlapping with the cellular network. As a widespread technology, Wi-Fi networks have been proposed as a suitable solution for data offloading in cellular networks. Considering the effect of access points (APs) on the performance of Wi-Fi networks, deploying APs can affect the efficiency and cost of Wi-Fi-based data offloading. This issue is the main research part of the current paper. In this paper, an optimization problem is proposed to find the best location for the Wi-Fi APs, providing the maximum performance metric for offloading. Two optimization algorithms are proposed to solve the problem: the Krill Herd Algorithm (KHA) and the Greedy algorithm. The evaluation results indicate that the feature of global optima in the exploration phase of the KHA algorithm leads to finding a better location of the APs than the Greedy algorithm.