RACE-IoT: Shannon–Dragonfly Optimized Clustered Load Balancing in IoT Edge Computing

With the proliferation of ubiquitous Internet of things (IoT) sensors and smart devices in a wide array of industries, involving transportation, agriculture, and healthcare, a huge amount of data is being generated, which requires ever-increasing calculations and services both at the cloud and at the edge of the network. Because cloud servers are specifically located a significant distance from IoT devices, time-sensitive Internet of things applications must be extended to the cloud architecture for the timely delivery of important services. Although IoT services can be allocated to the right edge nodes, ensuring low latency and efficient resource use is still challenging. To overcome these challenges, a novel resource allocation using clustering enabled optimization in IoT (RACE-IoT) has been proposed in this paper. This work introduces a novel Shannon–dragonfly (Sha-Dragon) optimization algorithm, which combines the dragonfly algorithm with the Shannon entropy for resource scheduling. The proposed RACE-IoT balances the load by reallocating the resources to maintain the same load ratio over all servers. The developed technique has been compared with the previous strategies involving PBSM, HRL-Edge-Cloud, and ERAM-EE in terms of specific parameters such as energy consumption, throughput, response time, end-to-end delay, resource utilization, makespan, and success rate. The outcomes demonstrate the efficiency of the developed RACE-IoT technique. The proposed technique achieves higher success rates of 35%, 17%, and 18% than existing techniques such as PBSM, HRL-Edge-Cloud, and ERAM-EE, respectively.