Optimal deployment of cloudlets based on cost and latency in Internet of Things networks

With the development of the Internet of Things (IoT), a large amount of data is generated on the network edge. Given the limited computing power of mobile devices (MDs) and access to computing resources from remote clouds, which leads to high latency to MDs, edge computing provides a way to reduce service latency by building a miniature cloud (Cloudlet). MDs transfer tasks they generate to nearby cloudlets for lower latency. Although a lot of research has been done in the field of edge computing, little attention has been paid to how to deploy cloudlets in the network. In this paper, we study the cloudlet deployment on a large number of wireless access points (APs) in an IoT network to optimize both deployment cost and network latency. When the cloudlets has been deployed in the network, we propose a fault-tolerant cloudlet deployment scheme. When the original cloudlets in the network fail, the software-defined network technology is used to start the fault-tolerant cloudlets in time to ensure the stability of the network latency. To address the above problems, we propose a binary-based differential evolution cuckoo search (BDECS) algorithm, which selects the permanent cloudlet deployment location among a large number of APs on the network. Extensive simulations reveal that the proposed algorithm has better performance in minimizing cost and latency compared with other deploymegt algorithms. Moreover, the convergence speed of the BDECS algorithm is also superior to other algorithms.