Optimization of a stochastic model having erratic server with immediate or delayed repair.

Area of Science:

• Operations Research
• Computer Science
• Applied Mathematics

Background:

• Queuing systems are integral to computing and networking.
• System reliability is challenged by internal faults like power failures and virus attacks.
• Optimizing system performance requires balancing operational costs and service quality.

Purpose of the Study:

• To propose a novel queuing model with an erratic server, delayed repair, and balking behavior.
• To analyze system performance under active and passive breakdown conditions.
• To provide a framework for enhancing service standards in computing and networking systems.

Main Methods:

• Development of a queuing model incorporating server breakdowns and balking.
• Application of the supplementary variable technique for performance indicator derivation.
• Comparison of analytical results with Adaptive Neuro-Fuzzy Inference System (ANFIS) results.
• Utilization of Particle Swarm Optimization (PSO) and Multi-Objective Genetic Algorithm (MOGA) for minimization problems.
• Formulation of minimization problems as convex programming for global optimality.

Main Results:

• The queuing model effectively captures system behavior under various fault conditions.
• ANFIS provides a robust comparison with analytical findings.
• PSO and MOGA successfully identify optimal solutions for single and bi-objective minimization.
• The study demonstrates a method to achieve a balance between operational costs and service quality.

Conclusions:

• The proposed queuing model offers valuable insights for industrial software developers and system managers.
• The integration of soft computing and optimization techniques enhances the analysis of complex systems.
• The approach facilitates informed decision-making for system evolution and resource management.