M/M/1 Queue: The Single-Server Queueing Model
Also known as: Single-Server Markovian Queue, Birth-Death Queue, Poisson Queue, M/M/1 Kuyruk Modeli
The M/M/1 queue is the foundational single-server queueing model in which customers arrive according to a Poisson process with rate λ, are served one at a time by a single server with exponentially distributed service times at rate μ, and wait in an infinite-capacity first-come-first-served queue. Formalized within the Kendall notation framework by David Kendall in 1953, building on A. K. Erlang's early twentieth-century telephone traffic work, it yields closed-form steady-state performance measures when the traffic intensity ρ = λ/μ is less than one.
Key highlights
- Yields exact closed-form steady-state performance measures without simulation.
- Mathematically tractable due to the memoryless (Markov) property of both arrival and service processes.
- Serves as the canonical baseline for comparing more complex queueing models.
- Requires only two parameters (λ and μ), making it easy to calibrate from empirical data.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Use the M/M/1 model when a system has a single server, Poisson arrivals, exponentially distributed service times, infinite queue capacity, and a first-come-first-served discipline. It suits preliminary capacity planning and bottleneck analysis where exact distributional assumptions can be justified or serve as useful approximations. Key limitation: real service-time distributions are rarely exponential; heavy-tailed or deterministic service requires alternative models (M/G/1, D/D/1). The infinite-capacity assumption also breaks down in physically bounded systems. For multiple servers, the M/M/c model is the natural extension.
Strengths & limitations
- Yields exact closed-form steady-state performance measures without simulation.
- Mathematically tractable due to the memoryless (Markov) property of both arrival and service processes.
- Serves as the canonical baseline for comparing more complex queueing models.
- Requires only two parameters (λ and μ), making it easy to calibrate from empirical data.
- The exponential service-time assumption is often unrealistic; real service times frequently have lower or higher variance than exponential.
- Assumes a single server; multi-server or multi-stage systems require M/M/c or network-of-queues models.
- Infinite queue capacity is physically impossible in many real systems.
- Steady-state results do not describe transient behavior or systems near saturation where ρ approaches 1.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
What happens to the queue if ρ equals or exceeds 1?
When ρ ≥ 1, the arrival rate meets or exceeds the service rate. The queue grows without bound and no steady-state distribution exists. The closed-form formulas for L, W, L_q, and W_q become infinite or undefined. In practice, finite buffer models (M/M/1/K) or system redesign with additional servers are required.
How is the M/M/1 model related to Little's Law?
Little's Law (L = λW) is a general result that holds for any stable queueing system regardless of distributional assumptions. The M/M/1 model provides specific closed-form expressions for L and W; verifying that these satisfy Little's Law is a standard consistency check. Similarly, L_q = λW_q holds for the queue subsystem.
Can the M/M/1 model handle non-Poisson arrivals or non-exponential service?
No — the model is defined by Poisson arrivals and exponential service times. For general service-time distributions with Poisson arrivals, the M/G/1 queue (and the Pollaczek-Khinchine formula) applies. For general inter-arrival distributions, the G/G/1 model is used, though it lacks fully closed-form results.
Sources
- 1.Kendall, D. G. (1953). Stochastic processes occurring in the theory of queues and their analysis by the method of the imbedded Markov chain. The Annals of Mathematical Statistics, 24(3), 338–354.
You have read it. What now?
Cite this page
ScholarGate. (2026, June 2). M/M/1 Queue. ScholarGate. https://scholargate.app/operations-research/mm1-queue