Brief Announcement: Generative Markov Model for Distributed Computing Systems

2026-06-01

AI Summary

This work addresses the lack of a unified formal model for efficiently harnessing the heterogeneous, dynamic, and complex distributed resources in computing continua. It proposes a generative Markov model grounded in structured system states, which factorizes high-dimensional state variables via sparse dependencies. By integrating generative modeling with reinforcement learning—a novel combination in this context—the approach enables scalable modeling, inference, and adaptive decision-making for large-scale distributed systems. Experimental evaluation in collaborative AI inference scenarios demonstrates that the method facilitates efficient decentralized scheduling, significantly reducing both task latency and server resource consumption, and outperforms conventional centralized strategies.

Abstract

Emerging distributed computing paradigms, such as the computing continuum, are inherently heterogeneous, stochastic, and complex. Efficiently and effectively utilizing all available resources across the continuum demands a unified formal model of the system. To address this gap, we propose a general framework for modeling distributed computing systems as a generative Markov model, factorized over a structured system state. In our model, the state decomposes into high-dimensional variables, each further factorized over its elements, reflecting the sparse dependency structure inherent to distributed systems. This yields a tractable model enabling simulation, inference, and policy learning over otherwise intractable system states, bridging distributed computing with Markov chain theory and reinforcement learning (RL). We demonstrate our framework through a case study of collaborative AI inference, in which a dedicated server combines resources with those volunteered by service users. Our results show that centralized scheduling becomes a bottleneck at scale, while distributing computation across user devices reduces both latency and server resource consumption. These findings highlight the value of adaptive decision-making in distributed computing systems and demonstrate the framework's utility for modeling, simulation, and optimization.

Problem

• distributed computing
• computing continuum
• system modeling
• resource utilization
• heterogeneous systems

Innovation

• Generative Markov Model
• Distributed Computing
• Structured State Factorization
• Reinforcement Learning
• Computing Continuum