Chapter-4: Queuing Theory (Waiting Line Theory) in Operations Research

4.1 Introduction to Queuing Theory

Queuing Theory (also known as Waiting Line Theory) is the mathematical study of
waiting lines or queues. It is used to model situations where there is a need to manage the
flow of customers, information, or materials in a system, and where delays may occur due to
limited capacity or resources. Queuing theory helps to analyze the arrival of customers, the
service process, and the waiting time before customers can be served.

The goal of queuing theory is to provide insights that can help minimize waiting times,
optimize service processes, and improve the efficiency of systems in various industries, such
as telecommunications, healthcare, transportation, and manufacturing.

4.2 Objective of Queuing Theory

The primary objectives of queuing theory are:

1. Optimize Service: To design systems that minimize waiting time, balance service
speed with demand, and allocate resources efficiently.
2. Analyze System Behavior: To predict the behavior of waiting lines under various
conditions, such as changes in arrival rates or service rates.
3. Improve Customer Satisfaction: By minimizing the amount of time customers wait
for service; queuing theory aims to improve the customer experience.
4. Cost-Effective Solutions: To develop cost-effective solutions for managing resources
and facilities, balancing the cost of service provision with customer wait times.

4.3 Characteristics of Queuing Model

A queuing model typically consists of the following components:

1. Arrival Process: The process by which customers arrive at the queue. It can be
characterized by:
o Arrival rate (λ\lambdaλ): The average number of customers arriving per unit
time. Commonly modeled by Poisson distribution.
o Inter-arrival time: The time between two consecutive customer arrivals,
often modeled using exponential distribution.
2. Service Process: The process by which customers are served. It can be characterized
by:
o Service rate (μ\muμ): The average number of customers a server can handle
per unit of time.
o Service time: The time required to serve a customer, usually modeled using
an exponential distribution.
3. Queue Length (L): The number of customers waiting in the queue at a given time.
4. Number of Servers (c): The number of servers or service channels available in the
system.
5. System Capacity: The maximum number of customers that can be accommodated in
the system, including both those is served and those waiting.
6. Waiting Time (W): The amount of time a customer spends in the queue before being
served.
5.4 Assumptions of Queuing Model

The basic assumptions of queuing models typically include:

1. Arrival Process:
o Customer arrivals follow a Poisson process (i.e., they occur randomly over
time).
o The average arrival rate is λ\lambdaλ (customers per unit of time).
2. Service Process:
o The service times follow an exponential distribution (i.e., the time between
completions of two service operations is independent of each other).
o The average service rate is μ\muμ (customers served per unit of time).
3. Queue Discipline:
o The service discipline is typically First-Come, First-Served (FCFS), but
other disciplines can be considered.
4. Infinite Queue Capacity:
o The queue can hold an unlimited number of customers.
o There is no limit to the number of customers that can wait in line.
5. Exponential Service Time:
o The time between service completions follows an exponential distribution,
which is a typical assumption for simple queuing models.
6. No Balking or Reneging:
o Customers do not leave the queue before receiving service (no reneging) and
do not decide not to enter the system (no balking).

4.5 Classification of Queues

Queuing systems can be classified based on the following parameters:

1. Arrival Process (A):

o Poisson (M): The inter-arrival times follow an exponential distribution, often
referred to as Markovian (M).
o Deterministic (D): The inter-arrival times are constant.
o General (G): The inter-arrival times can follow any distribution.
2. Service Process (B):
o Markovian (M): The service time follows an exponential distribution
(memoryless).
o Deterministic (D): The service times are fixed and known.
o General (G): The service time can follow any distribution.
3. Number of Servers (c):
o Single Server (1): Only one server is available.
o Multiple Servers (c): Multiple servers are available to serve customers
simultaneously.
4. Queue Capacity (K):
o Infinite Capacity (∞): The queue can accommodate an unlimited number of
customers.
o Finite Capacity (K): There is a limit on the number of customers that can
wait in the queue.
 5. Queue Discipline:
o First-Come, First-Served (FCFS): Customers are served in the order they
arrive.
o Priority Queuing: Customers are served based on priority, regardless of their
arrival time.
o Shortest Job Next (SJN): The customer with the shortest expected service
time is served first.

4.6 Queue Discipline

Queue discipline defines the order in which customers are served. The most common types
of queue disciplines are:

1. First-Come, First-Served (FCFS): Customers are served in the order they arrive.
This is the most common and simplest discipline.
2. Shortest Job First (SJF): Customers with the shortest service times are served first.
This can reduce waiting time but is not always practical because it requires knowing
the exact service time in advance.
3. Priority Queueing: Customers with higher priority are served first, regardless of their
arrival time. This is used in emergency situations, such as in hospitals or customer
support systems.
4. Round Robin (RR): This is used in systems where each customer is given a fixed
time slot or quantum. After each time quantum, the service moves to the next
customer in line.

4.7 Single-Channel and Multiple-Channel Queuing Models

1. Single-Channel Queuing Model:

o In this model, there is only one server available to serve the customers. This
model is simple and is typically used in situations like a single cashier at a
grocery store.

Notation: M/M/1M/M/1M/M/1

o M: Poisson arrival process (Markovian arrival).

o M: Exponential service time (Markovian service).
o 1: Single server.
2. Multiple-Channel Queuing Model:
o In this model, multiple servers are available to serve customers
simultaneously. This is used in situations where multiple resources are needed
to handle demand, such as multiple cash registers at a shopping mall.

Notation: M/M/cM/M/cM/M/c

o M: Poisson arrival process.

o M: Exponential service time.
o c: Number of servers.

The service system with multiple servers helps reduce waiting time and improve customer
satisfaction.
4.8 Limitations of Queuing Theory

While queuing theory is a powerful tool for understanding and optimizing service systems, it
has several limitations:

1. Simplifying Assumptions:
o Queuing models typically assume Poisson arrivals and exponential service
times, which may not always reflect real-world scenarios.
o The assumption of infinite queue capacity is often not practical, as many
systems have finite resources.
2. Complexity in Real Systems:
o Queuing models may not account for all real-world complexities, such as
varying service speeds, non-arrival patterns, or customer behavior.
3. Static Nature:
o Most basic queuing models assume fixed arrival and service rates, but in
reality, these rates can change over time due to factors like seasonality or
unexpected events.
4. No Human Factors:
o Queuing theory often ignores human behavior aspects such as impatience,
which can impact customer satisfaction and decision-making.
5. Over-Simplification of Multi-Stage Systems:
o Real-world queuing systems often have multiple stages (e.g., in hospitals,
where a patient must pass through various departments). While multi-phase
queuing models exist, they can be highly complex and computationally
expensive.