VTR

VTR quantifies timely reliability of virtual machines and end-to-end timely reliability in edge computing networks under VM migration and recovery to assess compliance with end-to-end delay constraints for delay-sensitive tasks.

Key Features:

• Timely Reliability Analysis: Quantifies timely reliability of virtual machines (VTR) and end-to-end timely reliability (ETR) within edge servers to evaluate the impact of VM migration on service downtime and delays.
• State Modeling with CTMC: Employs a continuous-time Markov chain (CTMC) to model three VM states—operational, migrating, and recovering—for analysis of failure, migration, and recovery dynamics.
• Analytical Methods: Uses matrix-geometric and first passage time methods to derive timely reliability metrics and related analytical insights.
• Simulation Verification: Validates analytical results through simulations implemented in OMNeT++.

Scientific Applications:

• Network Robustness Enhancement: Analyzes the impact of VM migration on network performance to inform design of more resilient edge computing networks for delay-sensitive applications.
• Optimization of Resource Allocation: Provides insights into trade-offs between task VMs (TVMs) and backup VMs (BVMs) to guide optimal allocation when the total number of VMs is constrained.

Methodology:

Defines a CTMC with VM states operational, migrating, and recovering; applies matrix-geometric and first passage time methods to compute VTR and ETR; and validates analytical outcomes using OMNeT++ simulations.