Definition
A Queuing System is a model of a dynamic system where “customers” (entities) request “service” from “servers” (finite-capacity resources), often resulting in waiting lines or queues. It is a fundamental application of Discrete Event Simulation (DES).
Why It Matters
Queuing is the math of “friction.” If you ignore it, your systems—factories, call centers, networks—will experience “unstable” growth, leading to infinite delays and total system collapse. It is the essential tool for finding the balance between the “cost of waiting” (unhappy customers) and the “cost of capacity” (idle servers).
Core Concepts
- Calling Population: The pool of potential customers.
- Arrival Pattern: Usually stochastic, often modeled with a Poisson Distribution.
- Service Pattern: Time taken to serve, often modeled with an Exponential Distribution.
- Queue Discipline: Logic for selecting next customer (FIFO, LIFO, Priority).
- Impatient Behavior: Balking, Reneging, Jockeying.
- System Stability: To prevent infinite growth, .
- How to read: “The lambda is less than mu.”
- Meaning / when to use: Arrival rate lambda must be strictly less than service rate mu for the queue to be stable (steady-state exists, no explosion). The fundamental stability condition in queuing theory (from Kendall’s notation and birth-death processes).