Development of a Double-Resampling-Based Least-Squares Particle Filter for Accurate Position Estimation of a GPS Receiver in Visakhapatnam Region of the Indian Subcontinent

Urban sensing plays a significant role in improving resource management, citizen engagement, environmental monitoring, urban planning, safety, and social equality. Global positioning system (GPS) is a crucial part in urban sensing as it provides accurate location tracking, real-time data collection, location-based services, mobility and transportation solutions, intelligent urban planning, and disaster management. However, there are various challenges associated with accurately estimating positions in urban environments due to various factors such as signal obstruction, urban canyons, multipath interference, noise and signal degradation, and differential GPS limitations. In the context of GPS-based urban sensing applications, the use of a navigation algorithm plays a critical role in extracting reliable data from corrupt sources, which can significantly impact inference performance for various signal processing applications. However, modeling all error sources that affect data quality can significantly increase system complexity, leading to challenges in terms of hardware and computation. To address this challenge, this article proposes a novel particle filter-based algorithm, called the double-resampling-based least-squares particle filter (DR-LPF), designed specifically for estimating the position of a GPS receiver without the need to model all error sources. By integrating current measurements (CMs) into the particle before resampling through the least-squares (LS) method, the DR-LPF allows the double-resampled particles to move toward high-likelihood regions, leading to improved estimation accuracy, reduced computation time, and reduced computational load. The application of the proposed DR-LPF algorithm finds wide applications in urban sensing environments where data quality can be affected by multiple error sources. By reducing the computational load and improving the estimation accuracy, the proposed DR-LPF algorithm can provide valuable insights into the movement and behavior of individuals and objects within an urban environment, enabling a wide range of smart city applications, such as traffic monitoring, environmental sensing, and crowd management.