<p>Wireless sensor network (WSN) typically consists of densely distributed sensor nodes deployed to gather information about their surrounding environment. However, the utility of this data is highly dependent on the knowledge of the exact spatial coordinates from which it is collected. Consequently, accurate localization of sensor nodes is a fundamental requirement for many WSN applications, such as intrusion detection in defence systems. The core objective of node localization is to compute the spatial positions of target nodes with the aid of anchor nodes whose positions are known. This study presents a modified Fruit Fly Optimization Algorithm integrating Levy flight and a dynamic search radius (FOA-LV), to enhance node localization efficiency and overcome the inherent drawbacks of the traditional fruit fly optimization algorithm (FOA), such as early stagnation in suboptimal solutions. The proposed FOA-LV algorithm refines the balance between exploration and exploitation by incorporating Levy flight mechanism and an adaptive search radius strategy. To validate the efficiency of the FOA-LV method, detailed simulations are performed across multiple node deployment configurations involving anchor and target nodes. Moreover, a comprehensive comparative analysis of FOA-LV method was conducted against other existing optimization algorithms across multiple error metrics and localization efficiency. The performance of the proposed method was systematically analyzed across varying transmission ranges, anchor densities, population sizes, Levy flight parameters, search radius, and monitoring area sizes. Extensive simulations confirm that FOA-LV achieves faster convergence, lower localization error, and 100% localization efficiency with reduced computation time, outperforming conventional FOA and other state-of-the-art optimization techniques in large-scale WSN.</p>