An RSSI-Based Fingerprint Localization using Multi-Signal Mean Optimization Filter in Indoor Environment Onboard a Passenger Ship

The development of the passenger ship industry has driven the prosperity of tourism and the national maritime economy, and has become an important bridge for cultural exchange. With the rapid advancement of the shipping industry, the safety of passengers has been a major concern. Random movement of passengers on board not only affects the normal operation of the vessel but also poses a threat to their own safety. Therefore, how to accurately track the position of passengers on board has become a focus of attention in the passenger ship industry. However, due to the unique environment inside the ship, the accuracy of traditional indoor localization methods has decreased. Therefore, this paper proposes a Multi-Signal Mean Optimization Filtering (MSMOF) approach. MSMOF first uses boxplot to eliminate outlier data, and then employs hyperparameter optimization algorithms to adjust target parameters, reducing the decrease in localization accuracy caused by the frequent fluctuations of Received Signal Strength Indicator (RSSI) data. Experiments show that the MSMOF method can reduce the noise impact of raw data by 80% and effectively improve the accuracy of indoor localization on passenger ships.