Resource-Aware Min-Min (RAMM) Algorithm for Resource Allocation in Cloud Computing Environment

Resource allocation (RA) is a significant aspect in Cloud Computing which facilitates the Cloud resources to Cloud consumers as a metered service. The Cloud resource manager is responsible to assign available resources to the tasks for execution in an effective way that improves system performance, reduce response time, reduce makespan and utilize resources efficiently. To fulfil these objectives, an effective Tasks Scheduling algorithm is required. The standard Min-Min and Max-Min Task Scheduling Algorithms are available, but these algorithms are not able to produce better makespan and effective resource utilization. This paper proposed a Resource-Aware Min-Min (RAMM) Algorithm based on classic Min-Min Algorithm. The RAMM Algorithm selects shortest execution time task and assign it to the resource which takes shortest completion time. If minimum completion time resource is busy then the RAMM Algorithm selects next minimum completion time resource to reduce waiting time of task and better resource utilization. The experiment results show that the RAMM Algorithm produces better makespan and load balance than standard Min-Min, Max-Min and improved Max-Min Algorithms.

💡 Research Summary

The paper introduces a novel task‑scheduling algorithm called Resource‑Aware Min‑Min (RAMM) aimed at improving resource allocation in cloud computing environments. Traditional Min‑Min schedules the task with the smallest completion time on the resource with the smallest execution time, while Max‑Min does the opposite. Both suffer from load imbalance, starvation of large tasks, and increased makespan because they either concentrate work on a single fast resource or neglect the availability of other resources.

RAMM redefines the selection criteria: it first picks the task with the shortest expected execution time (ET) and then assigns it to the resource that yields the minimum expected completion time (CT), where CT = ET + ready time (RT) of the resource. If the resource with the minimum CT is busy, RAMM chooses the next‑best resource (the second‑minimum CT) instead of forcing the task to wait. After a task is assigned, the resource’s RT is updated by adding the task’s ET, and the CT matrix is recomputed for the remaining tasks. This loop continues until all tasks are scheduled.

The algorithm’s key advantages are twofold. First, it reduces waiting time because a task can be placed on any idle resource that offers a near‑optimal CT, rather than being forced to wait for the single fastest resource. Second, it balances load across resources, preventing the situation where a fast VM becomes a bottleneck while slower VMs stay idle.

Experimental validation was performed using CloudSim and MATLAB on a system with an Intel i5 CPU and 12 GB RAM. Four tasks (with varying instruction and data volumes) and two virtual machines (different MIPS and bandwidth) were used. Expected execution times were calculated using the formula ET = (MI/MIPS) + (Mb/bandwidth). Results show that RAMM reduces makespan by roughly 15‑20 % compared to classic Min‑Min, and also outperforms Max‑Min and an improved Max‑Min variant by about 12‑15 %. Resource utilization improves from around 65 % (Min‑Min) to 78 % (RAMM), indicating better load distribution.

The authors acknowledge limitations: RAMM currently considers only ET and RT, ignoring task priorities, cost, or SLA constraints. Extending the model with weighted multi‑objective criteria would increase its applicability. Moreover, the O(m·n) complexity of recomputing the CT matrix may become a bottleneck in large‑scale clouds; distributed or approximate computation techniques are suggested for future work.

In conclusion, RAMM offers a simple yet effective enhancement over existing Min‑Min/Max‑Min approaches, delivering lower makespan and improved load balancing without substantial algorithmic overhead. The paper recommends further research on integrating QoS parameters and scaling the method for massive cloud infrastructures.