An optimal workflow scheduling method has been developed for IoT-fog-cloud systems to minimize time and energy consumption.
The research examines the increasing reliance on the Internet of Things (IoT) and the resultant workflows requiring efficient processing on computing platforms. Traditional cloud computing solutions often face challenges due to their geographical distance from network edge devices, resulting in longer delays and increased costs. To combat this, the study presents the combination of Aquila and Salp Swarm Algorithms (ASSA) aimed at optimizing workflow scheduling, focusing particularly on reducing energy consumption (EC) and MakeSpan Time (MST) within the constraints of priority, deadlines, and reliability.
One of the major findings of the research indicates the proposed method significantly improves upon existing scheduling techniques. The ASSA algorithm utilizes various advanced techniques including Dynamic Voltage Frequency Scaling (DVFS) and Virtual Machine (VM) merging, leading to lower energy utilization and faster task completion compared to previous methodologies. For example, the efficiency of these techniques allows the ASSA to collectively minimize energy consumption, providing not only financial relief to resource providers but also contributing to overall environmental sustainability.
This paper proposes RMST, DVFS, and VM merging techniques, as well as the combined MH algorithm of ASSA, in order to reduce EC and workflow execution time. The combined approach is particularly notable as it allows for simultaneous multitasking by optimizing the parameters across different VMs for smooth execution of IoT workflows.
Experimental results demonstrate the effectiveness of the proposed method when benchmarked against other techniques, yielding superior performance metrics. Among the various methods tested, the ASSA showed lower energy consumption and MakeSpan Time values compared to methods such as Genetic Algorithms and Particle Swarm Optimization, making it a promising solution for future scheduling challenges.
Interestingly, the traditional concept of merely increasing computational efficiency runs counter to the study's findings, which advocate for smart energy management through reduced CPU frequency during less demanding tasks. This balance between energy conservation and task scheduling is pivotal, illustrating to researchers and practitioners alike the clear necessity of innovative solutions amid growing digital demands.
Notably, such systems are classified as NP-hard problems, presenting significant computational challenges. Despite this, the ASSA methodology provides systematic solutions which focus on optimizing energy use dynamically as workloads fluctuate. By ensuring compatibility with the demanding criteria of reliability and deadlines, this new scheduling approach signifies a considerable advance for IoT-fog-cloud computing.
The method’s holistic approach, integrating multiple strategies, allows for broader applications, paving the way for potential scalability across various sectors. Leveraging the advantages of ASSA could enable significant reductions in operational costs, making IoT applications more viable for small businesses and large enterprises alike.
The research concludes by noting the steady rise of IoT applications and the pivotal role efficient scheduling will play moving forward. The authors recommend continued exploration of hybrid algorithms to bolster sustainability via reduced energy consumption without compromising performance.