مشخصات پژوهش

The emergence of industry 4.0 and 5.0 has brought about profound transformations in both daily existence and industry. In many industries, the Internet of Things (IoT) is crucial for supplying information and carrying out the necessary tasks. One significant development for IoT activation is Low-Power Wide Area Networks (LPWANs), particularly LoRaWANs. Ac- curate tracking of machines, equipment, and objects to improve production and efficiency in industries is a requirement for the success of these revolutions. Received Signal Strength Indication (RSSI) fingerprint map is one of the advanced localization techniques. The measured RSSI is greatly affected by environmental changes, such as object displacement and weather variations. In outdoor settings, these alterations and displacements are more noticeable. To obtain improved localization accuracy, the fingerprint map needs to be updated frequently due to variations in RSSI caused by changes in the environment. For LoRa fingerprint-based localization, this poses a serious challenge. Environment related images, such as images from surveillance cameras, show environmental changes such as the movement of objects. Therefore, environmental images can be a useful tool for detecting environmental changes and updating fingerprint maps. This research helps to improve the accuracy and reliability of LoRaWAN fingerprint localization systems using image processing techniques to learn and predict the effect of environment changes on the RSSI and fingerprint map. To implement the proposed method, a real environment is used in a car parking environment, a place where the movement of vehicles is evident based on the measured RSSI. The results show that this method can greatly improve localization, which results in localization output that is significantly more accurate.