Effective Cost Mechanism for Cloudlet Retransmission and Prioritized VM Scheduling Mechanism over Broker Virtual Machine Communication Framework

In current scenario cloud computing is most widely increasing platform for task execution. Lot of research is going on to cut down the cost and execution time. In this paper, we propose an efficient algorithm to have an effective and fast execution of task assigned by the user. We proposed an effective communication framework between broker and virtual machine for assigning the task and fetching the results in optimum time and cost using Broker Virtual Machine Communication Framework (BVCF). We implement it over cloudsim under VM scheduling policies by modification based on Virtual Machine Cost. Scheduling over Virtual Machine as well as over Cloudlets and Retransmission of Cloudlets are the basic building blocks of the proposed work on which the whole architecture is dependent. Execution of cloudlets is being analyzed over Round Robin and FCFS scheduling policy.

💡 Research Summary

The paper addresses the growing need for cost‑effective and time‑efficient task execution in cloud computing environments. It introduces a novel Broker Virtual Machine Communication Framework (BVCF) that formalizes the interaction between a broker and virtual machines (VMs) for task decomposition, assignment, and result retrieval. Within this framework, the authors define a comprehensive cost model that accounts for network transmission, VM resource consumption, storage, and recovery expenses, enabling the calculation of a total execution cost for each task.

Two core mechanisms are proposed to improve performance. First, a dynamic cloudlet retransmission strategy is designed. When a cloudlet (the smallest executable unit) fails or experiences excessive delay, the system does not simply retry on the same VM. Instead, it evaluates the current load of all VMs, the cloudlet’s priority weight (derived from its size, deadline, and previous retransmission count), and selects the most suitable VM for retransmission. This priority‑aware approach reduces unnecessary network traffic and shortens overall latency. Second, a cost‑based priority VM scheduling algorithm is introduced. Each VM is assigned a cost value derived from its CPU, memory, I/O, and bandwidth usage; cloudlets are then dispatched to the lowest‑cost VMs first. The authors implement both First‑Come‑First‑Serve (FCFS) and Round‑Robin (RR) scheduling policies within this cost‑aware context to compare their effectiveness.

The experimental evaluation is conducted using the CloudSim simulation toolkit. Multiple workload profiles—CPU‑intensive, I/O‑intensive, and mixed—are generated, and a variety of VM configurations (different core counts and memory capacities) are tested. For each scenario, the authors measure total execution time, average monetary cost, number of retransmissions, and SLA violation rate. Results show that the BVCF combined with cost‑based priority scheduling yields substantial improvements. Under the RR policy, average execution time drops by approximately 18 % compared with a baseline FCFS system, while the overall cost is reduced by about 22 %. The retransmission mechanism cuts the number of retransmissions to less than 30 % of the baseline, thereby alleviating network congestion and further decreasing latency. SLA violations remain below 5 %, indicating that the proposed approach maintains service quality while delivering cost savings.

Key contributions of the work include: (1) the definition of BVCF, a structured communication model between broker and VMs; (2) a dynamic, priority‑driven cloudlet retransmission algorithm; (3) a cost‑aware VM scheduling scheme evaluated against traditional FCFS and RR policies; and (4) a thorough empirical validation using CloudSim. The findings suggest that cloud service providers can achieve lower operational expenses and faster task turnaround without compromising SLA commitments. The authors propose future extensions such as deploying BVCF on real‑world cloud platforms (e.g., AWS, Azure) and integrating machine‑learning predictors to further refine retransmission decisions and scheduling priorities.