Enhanced internet of things load balancing algorithm with awareness in resources availability

Abstract

Internet of Things (IoT) backed by blockchain preserves data from being tampered on peer-to-peer network communication between the IoT devices. Unfortunately, blockchain is an intensive computing process because of the consensus algorithm. Hence IoT device is not suitable for blockchain because IoT device is mostly a low computing device and made up with limited resources. This study propose the Awareness in Resources Availability (AIRA) algorithm that apply with weight least connection (WLC) algorithm at IoT blockchain platform. The purpose of the algorithm is to reduce the usage of resources on the IoT nodes; the parameters that will be compared are CPU percentage, RAM usage, and network throughput. The research methodology using three types of experiment in order to evaluate the AIRA algorithm performance and the experiment performs with a different type of load balancing. The load balancing algorithms used to compare with the AIRA algorithm are weight round-robin (WRR) and weight least connection (WLC). It was discovered that with a limited amount of resources at the low computing of IoT nodes, in most cases, the AIRA algorithm performs better compare to WRR and WLC in terms of reducing the CPU utilization, RAM usage, and network throughput. In conclusion, the AIRA algorithm is the best method to reduce the resources consumption of node at IoT-blockchain platform base on result, the average reduction of CPU utilization is 56 percent, the average reduction of RAM usage is 16 percent, the average reduction of network throughput for receive is 72 percent and finally the average network throughput for send is 81 percent.