Virtual Machine Resource Utilization in a Data Center

Technology Overview: A resource‑management method that predicts application workloads, computes average service rates per VM pool, then forms collaborative sets of VMs drawn across pools to meet service‑level requirements—distributing incoming requests among selected VMs to raise utilization while controlling latency and cost.

Industry Pain Point: Standard VM schedulers over‑ or under‑provision resources, ignoring cross‑pool collaboration—leading to SLA violations, idle capacity, and higher operational expense.

NJIT Solution: Use predictive workload modeling and collaborative VM selection to dynamically right‑size and balance load across pools, improving resource utilization without breaching SLAs.

Key Features & Advantages

• Computes average service rates per pool and predicts workload to pre‑empt hot spots.
• Builds collaborative VM sets and routes requests to satisfy SLA targets.
• Aligns with green computing goals via higher utilization of active infrastructure.

Development Stage: TRL 3–4 – Algorithm/prototype.

Target Markets

• Public cloud providers & private cloud operators
• Enterprise virtualization platforms
• Telco cloud and edge data centers