Densification has been acknowledged as a main technological pillar for the Fifth Generation (5G) of networking, to address the foreseen huge number of connected devices owing to the wide adoption of Internet of Things (IoT), along with the unprecedented high amounts of data and requested data rates. Small cells (SCs) are at the center of such approach; however, the problem of small cell allocation (SCA) in heterogeneous networks (HetNets) yields a highly nonlinear, and integer programming formulation. In this letter, we address such challenging problem through the proposal of an innovative Integer-based Non-Dominated Sorting Genetic Algorithm (I-NSGA) to deal with the integer-based multi-objective optimization modeling in SCA. The proposed I-NSGA offers lower computation costs compared to conventional integer programming, while maintaining high optimality searching for the Pareto Front. Energy consumption and the total achieved data rates are optimized in SC deployment. The simulation results, not only prove the applicability of the I-NSGA in integer-based multi-objective optimization, but also confirm its superiority compared to existing solutions.

中文翻译：

---

用于小小区分配优化的基于整数的多目标算法

密集化已被公认为第五代 (5G) 网络的主要技术支柱，以解决由于物联网 (IoT) 的广泛采用以及前所未有的大量数据和请求的数据速率。小基站 (SC) 是这种方法的核心；然而，异构网络 (HetNets) 中的小小区分配 (SCA) 问题会产生高度非线性的整数规划公式。在这封信中，我们通过提出一种创新的基于整数的非支配排序遗传算法 (I-NSGA) 来处理 SCA 中基于整数的多目标优化建模，从而解决了这一具有挑战性的问题。与传统的整数规划相比，提议的 I-NSGA 提供了更低的计算成本，同时保持搜索帕累托前沿的高最优性。在 SC 部署中优化了能耗和实现的总数据速率。仿真结果不仅证明了 I-NSGA 在基于整数的多目标优化中的适用性，而且证实了其与现有解决方案相比的优越性。