Delay Optimization In Smart Health Systems By Employing Dynamic Scheduling Approach With Gaussian Mixture Model

Abstract

With the increase in the number of Internet of Things (IoT) smart devices drastically, Low Power Wide Area Network (LPWAN) technologies have become an overwhelming choice worldwide. The researchers have used a variety of LPWAN technologies to solve difficulties like higher collision rates, retransmissions, delays, and energy usage. In contrast, the most appealing and appropriate technology in terms of energy efficiency, cheap cost, and delay optimization is Sigfox. The primary problem with Sigfox is the high percentage of packet drops caused by collisions. The Pure Aloha MAC technique, which Sigfox uses to transmit frames or data readings, is the main cause of this packet drop rate and ultimately retransmissions. Many retransmissions result from communication between Sigfox smart devices and Pure Aloha. The delay in Sigfox network has increased even further, with the increase in the number of retransmissions. This work uses the Gaussian Mixture Model (GMM), an unsupervised probabilistic technique, in conjunction with a Dynamic Scheduling Approach (DSA) to optimize the delay in Sigfox network. Retransmissions are decreased when DSA is used in conjunction with GMM, optimizing the latency in Sigfox. According to the findings, our method reduces Frame Collision Rate (FCR) by 15% when compared to traditional Sigfox. Furthermore, a 39% increase in Frame Success Ratio (FSR) is seen when comparing the traditional Sigfox. Moreover, 79% of the delay is optimized. This study may be useful in situations when patients' vital information must be transmitted to gateways with the least amount of delay and with optimal retransmissions.