Abstract:Motivated by the wide range of modern applications of the Erlang-B blocking model beyond communication networks and call centers to sizing and pricing in design production systems, messaging systems, and app-based parking systems, we study admission control for such a system but with unknown arrival and service rates. In our model, at every job arrival, a dispatcher decides to assign the job to an available server or block it. Every served job yields a fixed reward for the dispatcher, but it also results in a cost per unit time of service. Our goal is to design a dispatching policy that maximizes the long-term average reward for the dispatcher based on observing only the arrival times and the state of the system at each arrival that reflects a realistic sampling of such systems. Critically, the dispatcher observes neither the service times nor departure times so that standard reinforcement learning-based approaches that use reward signals do not apply. Hence, we develop our learning-based dispatch scheme as a parametric learning problem a'la self-tuning adaptive control. In our problem, certainty equivalent control switches between an always admit if room policy (explore infinitely often) and a never admit policy (immediately terminate learning), which is distinct from the adaptive control literature. Hence, our learning scheme judiciously uses the always admit if room policy so that learning doesn't stall. We prove that for all service rates, the proposed policy asymptotically learns to take the optimal action and present finite-time regret guarantees. The extreme contrast in the certainty equivalent optimal control policies leads to difficulties in learning that show up in our regret bounds for different parameter regimes: constant regret in one regime versus regret growing logarithmically in the other.

Admission Control of Parallel Queues with Fork Types of Jobs

Online Scheduling on Parallel Machines: A Survey

Online Hierarchical Scheduling: an Approach Using Mathematical Programming

On-line Scheduling of Parallel Jobs on Two Machines

Online scheduling with hard deadlines on parallel machines

Generalized Parallel-Server Fork-Join Queues with Dynamic Task Scheduling.

Optimal On-Line Algorithms for Scheduling on Parallel Batch Processing Machines

Admission Control for Double-ended Queues

On the Control of Fork-Join Networks

Scheduling in Parallel Queues with Randomly Varying Connectivity and Switchover Delay

Scheduling And Control For A Failure Prone System Of Open Shop With Job Overlaps

Modelling and Analysis of Fork-Join Discrete Event Dynamical System

Parallel Scheduling on Three Machines under a Grade of Service Provision

Online scheduling on parallel-batching machines with periodic availability constraints and job delivery

Optimal Admission Control for Multiclass Queues with Time-Varying Arrival Rates via State Abstraction

A Paradox for Admission Control of Multiclass Queueing Network with Differentiated Service

Non-Asymptotic Delay Bounds for Multi-Server Systems with Synchronization Constraints

Steady-State Analysis and Online Learning for Queues with Hawkes Arrivals

Optimal Scheduling Control in Fluid Models of General $n\times n$ Input-Queued Switches

Coflow Scheduling in Input-Queued Switches: Optimal Delay Scaling and Algorithms

Learning a Discrete Set of Optimal Allocation Rules in a Queueing System with Unknown Service Rate