A Comprehensive Review on Sensor Fusion Techniques for Localization of a Dynamic Target in GPS-Denied Environments

Although Global Navigation Satellite Systems (GNSS) typically provide sufficient accuracy for outdoor positioning, positioning systems face significant challenges in complex application scenarios, particularly in GPS-denied environments (GDE). As application demands increase, single-sensor positioning methods are increasingly inadequate to meet the requirements for high accuracy and robustness, making multi-sensor fusion techniques a growing research focus. In recent years, while many studies have aimed to achieve indoor positioning stability and reliability through Simultaneous Localization and Mapping (SLAM), Ultra-Wideband (UWB), Odometry, and Wi-Fi fusion technologies, these efforts have often been conducted in isolation, lacking a systematic synthesis of these fusion methods. Existing reviews predominantly focus on specific aspects of sensor fusion, without providing a comprehensive perspective on the integration of multiple sensor fusion approaches. Notably, there is a particular gap in the comprehensive analysis of Dynamic Target Localization (DTL). DTL involves real-time tracking of moving targets such as pedestrians, robots, vehicles, and Unmanned Aerial Vehicles (UAVs), which not only requires maintaining high accuracy in complex and dynamic environments but also demands the capability to handle the challenges posed by rapidly moving targets. This paper addresses this research gap by systematically reviewing the existing sensor fusion techniques, providing a detailed analysis of the application of SLAM fusion, UWB fusion, odometry fusion, and Wi-Fi fusion in DTL, and thoroughly evaluating their effectiveness and performance. Additionally, this paper summarizes the application of these technologies across different types of moving targets, discusses future development trends and potential challenges, and offers valuable insights for advancing the widespread application of multi-sensor fusion techniques in DTL. Finally, this paper discusses the challenges facing future work in this field, emphasizes what sets this review apart from existing studies, highlights the systematic analysis of multi-sensor fusion methods, and includes specific research findings to more clearly demonstrate the paper’s contributions.