Using intelligence in resource allocation and task off-loading for the IoE-edge networks

With the increased usage of Internet of Everything (IoE) capable devices and new communication technologies such as Sixth Generation (6G), more and more services can be made available close to the edge of networks formulating the IoE-based edge networks. In edge networks, several devices communicate with each other for the purpose of sharing information and lending each other various computing resources. This has raised challenges of how efficiently and effectively computing resources can be shared among IoE devices so that users can achieve high quality of service and the network resources are optimally utilised. This paper addresses the issue of computing resource allocation among various devices in such a way that every device can get its task done while lending its unutilised resources to other tasks. We proposed a local scheduler-based architecture where the central scheduler has up-to-date information on the resources available within the network and then allocates them under a certain pre-defined scheduling policy. We introduced the intelligence in the system based on various characteristics of the devices such as each device’s battery level, storage capacity, and computing capability. The proposed algorithm is named the Intelligent Main Task Off-loading Algorithm (iMTOSA). Novel scheduling schemes using these intelligence-based characteristics for device identification, selection, scheduling, and task management within the IoE cluster at Layer 1 supersedes conventional scheduling policies. To evaluate the performance of the proposed iMTOSA algorithm, we used the Program Evaluation and Review Technique (PERT) and Central Limit Theorem (CLT) to calculate the Z-scores for the successful completion of each task. The proposed algorithm was evaluated through extensive simulations, showing that intelligent scheduling algorithms (iRR, iSC, iMR, iPF, iPB) achieved task success rates of 89% to 99.8% significantly outperforming non-intelligent algorithms, which ranged from 30% to 40%. The proposed algorithm enhances the 6G system’s overall performance compared to similar techniques regarding successful task allocation, achieving higher efficiency rates than non-intelligent algorithms. We compared the performance of the proposed algorithm with the existing similar scheme in the literature and it is shown that our proposed algorithm has better performance and stands out when compared under similar network settings. The proposed iMTOSA approach is suitable for IoE-Edge cluster-based industrial environments and business scenarios like 6G to scale its enormous volume of IoE-generated data.