Optimizing Task Scheduling and Resource Allocation Using Multi-Criteria Framework in Fog-Assisted IoT Networks with Preemption

The Internet of Things (IoT) can be developed further using fog and cloud computing environments. Task scheduling is highly effective for carrying out user requests in these settings, and the IoT-fog-cloud system's productivity is increased when IoT task requests are scheduled well. To address challenges such as latency, bandwidth overhead, and resource management, this paper proposes the Optimized Scheduling and Cluster-based Resource Allocation (OSCRA) model. The OSCRA model introduces a multi-criteria task scheduling with preemption using: (i) expectation-maximization (EM) clustering to group jobs by priority and deadline, (ii) a heap-based optimizer to schedule jobs based on SLA and QoS restrictions, and (iii) distributed resource management to assign resources effectively. Experimental results using iFogSim demonstrate that combining MSCFS and OSCRA enhances server utilization, reduces latency, improves throughput, and shortens response time, outperforming existing models.