TMaLB: A Tolerable Many-Objective Load Balancing Technique for IoT Workflows Allocation

Abstract

Recently, the load imbalance of network resources in the Internet of Things (IoT) due to the diverse and heterogeneous workflow of data, has attracted the attention of many researchers. To make appropriate load-balancing decisions, the influential Quality of Service (QoS) parameters should be identified. Most load-balancing techniques have focused on optimizing the multi-objective QoS while, the QoS optimization from the user, service provider and infrastructure provider directions simultaneously is an important many-objective problem. Extracting the QoS parameters from various directions, this paper proposes a tolerable many-objective load-balancing technique to efficiently allocate computing resources to workflows. Cost, response time, energy consumption, and CPU utilization have been investigated as the user, service provider, and infrastructure’s QoS parameters solved by Grey Wolf Optimization (GWO) algorithm. Also, a load-tolerance concept has been introduced to improve the QoS-based load balancing and task acceptance rate together. The experiment results show that TMaLB has about 10% improvement compared to MaPSO and NSGA-III algorithms.