Multi-Resource Allocation with Percentile Delay Guarantees based on Stochastic Network Calculus

Resource allocation has become increasingly vital in the era of advanced industrial informatization, where technology evolves rapidly, and the efficient use of resources is paramount. This paper adapts the Moment Generating Function-Stochastic Network Calculus (MGF-SNC) method, traditionally employed for probabilistic delay analysis, to resource allocation, specifically targeting Service Level Agreements (SLAs) response time constraints. Unlike methods that rely on average or worst-case response times, our approach focuses on the tail of response time distributions. This shift, enabled by the robust mathematical models of MGF-SNC theory, facilitates handling diverse resource configurations and complex scenarios with excellent scalability.