Planning of location area (LA) in cellular networks plays a vital role in utilizing the resources efficiently and economically. Location update and paging costs directly affect the cost to the service provider. When the size of the Mobile Switching Centre (MSC) is large, the paging cost becomes maximum. Though, no location update is required for this case. On the contrary, paging cost shoots high if each cell forms individual LA, but in this case, location update cost shoots to the maximum value. Therefore, deducing the network to the optimal amount of LA's is an intractable combinatorial optimization problem. In this paper, we intend to optimize the partitions of the MSC service area into an optimal number of LA, to curtail the total location management cost. We propose a novel scheme bearing in mind cell attributes of the network to deduce an optimum number of location areas, leading to minimum cost per call. It is revealed by the results of this work, attained after simulating the scenario on MATLAB 2018a that Cell Attributes Based Algorithm is a very promising scheme for cellular networks, and it outperforms the existing algorithms in a practical scenario.

中文翻译：

---

基于人工智能的蜂窝网络位置区规划新方案

蜂窝网络中位置区 (LA) 的规划在有效和经济地利用资源方面起着至关重要的作用。位置更新和寻呼成本直接影响服务提供商的成本。当移动交换中心(MSC)规模较大时，寻呼成本最大。但是，这种情况不需要位置更新。相反，如果每个小区形成单独的 LA，则寻呼成本会很高，但在这种情况下，位置更新成本会达到最大值。因此，将网络推导出最优数量的 LA 是一个棘手的组合优化问题。在本文中，我们打算将 MSC 服务区的分区优化为最优数量的 LA，以减少总的位置管理成本。我们提出了一种新颖的方案，考虑到网络的小区属性，以推导出最佳数量的位置区域，从而使每次呼叫的成本最低。这项工作的结果表明，在 MATLAB 2018a 上模拟场景后获得的基于单元属性的算法是一种非常有前途的蜂窝网络方案，并且在实际场景中优于现有算法。