Time synchronization is vital for accurate data collection and processing in sensor networks. Sensors in these networks often operate under fluctuating conditions. However, an accurate timekeeping mechanism is critical even in varying network conditions. Consequently, a synchronization method is required in sensor networks to ensure reliable timekeeping for correlating data accurately across the network. In this research, we present a novel dynamic NTP (Network Time Protocol) algorithm that significantly enhances the precision and reliability of the generalized NTP protocol. It incorporates a dynamic mechanism to determine the Round-Trip Time (RTT), which allows accurate timekeeping even in varying network conditions. The proposed approach has been implemented on an FPGA and a comprehensive performance analysis has been made, comparing three distinct NTP methods: dynamic NTP (DNTP), static NTP (SNTP), and GPS-based NTP (GNTP). As a result, key performance metrics such as variance, standard deviation, mean, and median accuracy have been evaluated. Our findings demonstrate that DNTP is markedly superior in dynamic network scenarios, a common characteristic in sensor networks. This adaptability is important for sensors installed in time-critical networks, such as real-time industrial IoTs, where precise and reliable time synchronization is necessary.
Keywords: FPGA; GPS-based network time protocol; internet-based network time protocol; network time protocol; time synchronization.