A New Optimization Approach Based on Firefly and Cuckoo Search Algorithm for Dynamic Multi-Objective Optimization Problem in Workflow Scheduling

Workflow scheduling is a well-explored topic for research in cloud computing, that focuses on maximizing the usage of cloud resources for different workflow activities. In this research, the topic of dynamic workflow scheduling is considered by addressing it as a Dynamic Multi-Objective Optimization Problem (DMOP). The dynamism in this challenge originates from two key sources: resource failures and the possible fluctuation in the number of goals. Resource failures may arise from software or hardware problems, producing the first form of dynamism. Furthermore, real-world circumstances in cloud computing might lead to changes in the number of goals throughout the execution of a process, providing the second source of dynamism. To tackle this issue, this research introduces a novel hybrid optimization algorithm that employs both the Firefly algorithm and the cuckoo search algorithm. The Firefly algorithm, inspired by the flashing activity of fireflies, aids in fast exploring the search space and fine-tuning solutions, whereas Cuckoo search maximizes the exploitation of the finest solutions found. This hybrid approach takes advantage of Firefly's global search capabilities and Cuckoo's local search intensification, resulting in a robust and adaptive scheduling strategy that can effectively allocate computational tasks to cloud resources, optimizing memory utilization, cost, energy consumption and makespan in dynamic cloud environments. the proposed approach FFBSO has showed remarkable results with a lowering the make span of only 10s, illustrating its extraordinary efficiency in managing tasks. Furthermore, when submitted to 400 jobs using the proposed technique, the model obtained an excellent energy consumption of 11kwh. In addition to its memory usage ability, the suggested model also succeeded in memory usage, obtaining a particularly low value of 87%.