Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2610.26653

Direct Sample-Average Learning for Queueing Control: Gradient-Free Top-$K$ Coarse-to-Fine Optimization

Mehrdad Nozohour Yazdi·Yiqiang Q. Zhao
View original source ↗Read source document ↗Export metadata ↓

Abstract

We study steady-state queueing control when decisions must be learned from a single persistent operating trajectory: only the implemented action produces data, and the queue is not reset when the action changes. Fixed-event-count cycles yield a direct ratio-of-totals sample-average approximation (SAA) of each action's steady-state cost rate. A Top-K coarse-to-fine search ranks these estimates, retains several promising refinement paths, and concentrates later operation on finer candidate sets, without requiring gradients, convexity, or differentiability. We decompose direct-SAA error into accumulated conditional transient bias and martingale fluctuation. For regenerative queues, a bias-function representation reduces transient effects within contiguous action blocks to boundary terms. Finite-sample stagewise bounds then link estimation accuracy to master-grid identification through guardian retention, and the required queueing and sampling conditions are verified for a persistent-workload M/M/1 implementation. At a common active set, Top-K has a weakly larger guardian margin than Top-1, and an exact-objective example shows Top-1 can discard the optimal refinement branch. Numerical experiments delineate scope rather than universal dominance: Top-K improves terminal identification in the reported LiQUAR and mixed staffing comparisons, while LiQUAR can have lower cumulative learning cost, and GOLiQ is markedly stronger on a smooth benchmark with reliable queueing sensitivities. Direct ranking is therefore best viewed as a complementary method when model-specific gradients are unavailable, inconvenient, or poorly matched to the decision space.

Explore related subjects

math.OC

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Mehrdad Nozohour Yazdi, Yiqiang Q. Zhao. 2026-10-19. Direct Sample-Average Learning for Queueing Control: Gradient-Free Top-$K$ Coarse-to-Fine Optimization. https://arxiv.org/abs/2610.26653

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Exploring chordal sparsity in semidefinite programming with sparse plus low-rank data matrices

Semidefinite programming (SDP) problems are challenging to solve because of their high dimensionality. However, solving sparse SDP problems with small tree-width are known to be relatively easier because: (1) they can be decomposed into smaller multi-block SDP problems through chordal conversion; (2) they have low-rank optimal solutions. In this paper, we study more general SDP problems whose coefficient matrices have sparse plus low-rank (SPLR) structure. We develop a unified framework to convert such problems into sparse SDP problems with bounded tree-width. Based on this, we derive rank bounds for SDP problems with SPLR structure, which are tight in the worst case.

math.OC↗
arXiv2026

Bellman Optimality of Average-Reward Robust Markov Decision Processes with a Constant Gain

Learning and optimal control under robust Markov decision processes (MDPs) have received increasing attention, yet most existing theory, algorithms, and applications focus on finite-horizon or discounted models. Long-run average-reward formulations, while natural in many operations research and management contexts, remain underexplored. This is primarily because the dynamic programming foundations are technically challenging and only partially understood, with several fundamental questions remaining open. This paper steps toward a general framework for average-reward robust MDPs by analyzing the constant-gain setting. We study the average-reward robust control problem with possible information asymmetries between the controller and an S-rectangular adversary. Our analysis centers on the constant-gain robust Bellman equation, examining both the existence of solutions and their relationship to the optimal average reward. Specifically, we identify when solutions to the robust Bellman equation characterize the optimal average reward and stationary policies, and we provide one-sided weak communication conditions ensuring solutions' existence. These findings expand the dynamic programming theory for average-reward robust MDPs and lay a foundation for robust dynamic decision making under long-run average criteria in operational environments.

math.OC↗
arXiv2026

Service-Induced Congestion in Memory-Constrained LLM Serving

In large language model (LLM) serving, each request accumulates persistent graphics processing unit (GPU) memory during service as its key-value cache grows with every generated token. Under high concurrency, aggregate memory usage therefore increases endogenously over time: the service process itself creates future capacity pressure. When memory capacity is exceeded, systems evict active requests, discarding cached state and restarting them later, which wastes computation and reduces throughput. We develop a discrete-time dynamical model of memory-constrained LLM inference that captures admission, memory growth, and eviction under continuous batching. In the saturated-input regime, the system admits both eviction-free fixed points and limit cycles with evictions. For homogeneous workloads, we show that the eviction-free equilibrium is unstable and that, except for a Lebesgue-measure-zero exact-capture set, the system converges to a unique worst-case limit cycle that is asymptotically stable outside this exceptional set, with throughput losses as large as 50%. For heterogeneous workloads, we prove a stability criterion in the two-class common-input setting and explain how the survival-polynomial mechanism generalizes to multiple classes and heterogeneous-input lengths. Under an input-dominated scaling regime, coprime decoding lengths stabilize the eviction-free equilibrium, while non-coprime lengths create synchronized modes that drive instability. These results characterize when workload heterogeneity desynchronizes completions and helps stabilize memory-constrained serving. More broadly, we identify service-induced congestion as a structural instability mechanism and derive scheduling design principles for sustaining high throughput.

math.OC↗