Presentation Information
[10p-E219-8]Reliability Analysis and Cost Optimization of HAPS-Assisted Cloud-FogEnabled ITS Using Semi-Markov Model with RAG Integrationand Quantum-Inspired QAOA Optimization
〇(D)Ishu Jain Jain1, Anupam Gautam1 (1.Netaji Subhas University of Technology)
Keywords:
High Altitude Platform Station,Low Earth Orbit,Quantum Approximate Optimization Algorithm
Modern Intelligent Transportation Systems (ITS) are undergoing a fundamental transformation driven by the convergence of Internet of Vehicles (IoV), High-Altitude Platform Stations (HAPS), Low Earth Orbit (LEO) satellites, and fog computing paradigms. This integration enables pervasive connectivity and ultra-low-latency processing at the network edge, positioning vehicular fog servers as the critical computational nexus for real-time traffic management, collision avoidance, and autonomous vehicle coordination. Ensuring the reliable and continuous operation of these servers is paramount. A single point of failure in the fog layer can cascade into network-wide disruptions affecting passenger safety and traffic efficiency. Despite this operational criticality, rigorous analytical frameworks for evaluating vehicular fog server dependability remain scarce. Semi-Markov processes offer a mathematically tractable yet realistic modeling approach, accommodating non-exponential sojourn-time distributions that better reflect the phase-type haracteristics of real hardware failure and repair cycles.
This paper presents a unified semi-Markov reliability and cost optimization framework for a HAPS-assisted cloud-fog Intelligent Transportation System (ITS). Steady-state availability is derived under deterministic and hypoexponential sojourn-time distributions. A Retrieval-Augmented Generation (RAG) pipeline is integrated as a dependable AI subsystem across the fog and cloud layers, modeled within the same semi-Markov framework. The Cost-Effectiveness Ratio (CER) is minimized using Differential Search Algorithm (DSA), Particle Swarm Optimization (PSO), and the quantum-inspired Quantum Approximate Optimization Algorithm (QAOA). QAOA converges within 78–83 iterations versus over 100 for DSA and PSO, achieving up to 15.3% lower CER in high-cost scenarios. RAG integration improves availability by up to 35.4% over the analytical baseline.
This paper presents a unified semi-Markov reliability and cost optimization framework for a HAPS-assisted cloud-fog Intelligent Transportation System (ITS). Steady-state availability is derived under deterministic and hypoexponential sojourn-time distributions. A Retrieval-Augmented Generation (RAG) pipeline is integrated as a dependable AI subsystem across the fog and cloud layers, modeled within the same semi-Markov framework. The Cost-Effectiveness Ratio (CER) is minimized using Differential Search Algorithm (DSA), Particle Swarm Optimization (PSO), and the quantum-inspired Quantum Approximate Optimization Algorithm (QAOA). QAOA converges within 78–83 iterations versus over 100 for DSA and PSO, achieving up to 15.3% lower CER in high-cost scenarios. RAG integration improves availability by up to 35.4% over the analytical baseline.