In recent years, edge computing (EC) has been widely studied as a new computing paradigm which extends cloud computing. It paves the way to further reduce the network latency between IoT/mobile devices (referred to as edge users hereafter) and service providers by pushing services and corresponding data from clouds to nearby edge servers located nearby edge users. The edge user allocation (EUA) problem is a new issue in EC environment. It aims at optimizing strategies to allocate edge users to those edge servers while fulfilling specific constraints, e.g., budget constraint, coverage constraint, etc. As the EUA problem is $\mathcal{NP}$-hard, effectively and efficiently solving it is still intractable. In this paper, we take allocating maximum edge users and employing minimum edge servers as objectives, then take both the proximity constraint and capacity constraint into account, and propose EUA-FOA, an Fruit fly Optimization Algorithm (FOA)-based approach, to solve the EUA problem. To extensively evaluate EUA-FOA's performance, we employ a widely used real-world dataset to conduct two sets of experiments, including small-scale EUA scenarios and large-scale EUA scenarios. We compare EUA-FOA against four representative approaches and the experimental results demonstrate that EUA-FOA is highly effective as it outperforms the state-of-the-art approaches significantly.

近年来，边缘计算（EC）作为一种扩展云计算的新计算范式得到了广泛的研究。它通过将服务和相应数据从云端推送到位于边缘用户附近的边缘服务器，为进一步减少物联网/移动设备（以下称为边缘用户）和服务提供商之间的网络延迟铺平了道路。边缘用户分配（EUA）问题是 EC 环境中的一个新问题。它旨在优化策略以将边缘用户分配给这些边缘服务器，同时满足特定约束，例如预算约束、覆盖约束等。 由于 EUA 问题是$\mathcal{NP}$- 困难，有效和高效地解决它仍然是棘手的。在本文中，我们以分配最大边缘用户和使用最少边缘服务器为目标，然后考虑邻近约束和容量约束，并提出基于果蝇优化算法（FOA）的方法 EUA-FOA，以解决EUA 问题。为了广泛评估 EUA-FOA 的性能，我们采用广泛使用的真实世界数据集进行两组实验，包括小规模 EUA 场景和大规模 EUA 场景。我们将 EUA-FOA 与四种代表性方法进行比较，实验结果表明 EUA-FOA 非常有效，因为它显着优于最先进的方法。