IT2032

Queuing Systems and Pseudorandom Number Generation • Kendall’s notation is a system of notation according to which various
A Queuing Model characteristics of a queuing model are concisely identified. It has been the
• Queues are commonly found in most systems where there exist one or more standard in the literature of queuing models (O’Reilly Media, Inc., n.d.). A
shared resources. general queuing system is denoted by (A / B / c), where:
• Queuing theory is the mathematical study of the congestion and delays of A – the interarrival time distribution
waiting in line. This theory examines every component of waiting in line to B – the service time distribution
be served, including the arrival process, service process, number of servers, c – the number of servers in the system
number of system places, and the number of customers such as people, N – the system capacity
data packets, cars, and mobile devices (Kenton, 2020). K – the size of the calling population
• Queuing theory helps in designing a balanced system that serves customers Note: N and K are usually dropped if it holds and infinite value.
quickly and efficiently in a sustainable manner.
• Queuing system pertains to any system that involves customers requesting Components of a queuing model (Güneş, n.d.):
for a particular service having a finite-capacity resource (Sokolowski & Bank, o Calling population – It refers to the population of potential customers,
2009). which can be assumed to be finite or infinite.
• Queuing systems can also be referred to as a system of flow. Thus, it is an  Infinite population model: The arrival rate is not affected by the
important application of discrete event system simulation, where the number of customers being served and waiting.
dynamics of waiting line ques are being analyzed (Chaturvedi, 2010).  Finite population model: The arrival rate depends on the number
of customers being served and waiting.
Strengths of a good queuing system (JRNI, 2020): o System capacity – It pertains to the limit on the number of customers
o It can increase operational efficiency. that may wait inline or in the system.
o It can improve the system’s productivity. o Arrival process
o It can reduce walkaway customers.  For infinite population models: This is in terms of the
o It can help in understanding trends within the system. interarrival time of consecutive customers. It involves two (2)
o It can increase customer lifetime value. arrival types which are the random arrival and the scheduled
arrival. In this model, at least one (1) customer is assumed to
Weaknesses of a queuing system (Kuklin, 2016): always be present. Thus, the server is never idle.
o The waiting time due to long queues.  For finite population models: This model is characterized by
o The possibility of queue jumping and reneging. tagging if the customer is pending (customer is outside the
o The minimization of waiting crowds. queuing system) or not and by the runtime of a customer (the
departure of the customer from the queuing system until the
• Queuing models are usually constructed by scientists, engineers, and customer’s next arrival to the queue).
developers to analyze the performance of a dynamic system where waiting o Queue behavior – It refers to the actions of customers while in a queue,
can occur. waiting for the service to begin.
• The goal of queuing models is to minimize the average number of waiting Examples:
for customers in a queue or simply to optimize the system’s service rate. It  A customer leaves upon seeing that the line is too long.
can also be used to forecast the estimated number of transactions in a  A customer leaves after being in the line because it is moving
queuing system. The key elements of queuing model are: too slow.
o Customer – This pertains to anything that arrives at a facility and  A customer moves from one line to a shorter line.
requires service.
o Server – This refers to any resource that provides a requested service.
04 Handout 1 *Property of STI
 student.feedback@sti.edu Page 1 of 5
 IT2032
o Queue discipline – It pertains to the logical order of customers in a o The service time vary from {1, 2, 3 …, 6} minutes.
queue that determines which customer is chosen for service when a Service Time
Probability
Cumulative
server becomes free. (minutes) Probability
1 0.10 0.10
Examples:
2 0.20 0.30
 First-in-first-out (FIFO)
3 0.30 0.60
 Last-in-first-out (LIFO) 4 0.25 0.85
 Shortest processing time first (SPTF) 5 0.10 0.95
 Service according to priority (PR) 6 0.05 1.00
o Service time – This can either be constant or random. It is usually
characterized as a sequence of independent and identically distributed Drawbacks:
random variables, such as exponential and gamma distribution. o The sample size is too small to be able to draw accurate/reliable
o Service mechanism – This involves the number of service centers and conclusions
interconnected queues on the system. Generally, a service center o The initial conditions within the system are not considered.
consists of a specific number of servers working in parallel, and upon
getting to the head of the queue, the customer takes the first available After the simulation run for 100 customers, the following data were gathered:
server. o Simulation System
Service Service Service
Interarrival Arrival time
Queuing Simulation Customer
Time (min.) (clock)
Time Start Time End Time
• Queuing simulations are performed to analyze and estimate the (min.) (clock) (clock)
performance of a queuing system and test the efficiency and effectiveness 1 -- 0 4 0 4
of a queuing model. It sometimes requires assumptions and sets of 2 1 1 2 4 6
3 1 2 5 6 11
limitations.
4 6 8 4 11 15
5 3 11 1 15 16
Example of a queuing simulation: 6 7 18 5 18 23
A queuing system at a small grocery store (Güneş, n.d.). …
Assumptions: 100 5 415 2 416 418
o There is only one (1) checkout counter. Total 415 -- 317 -- --
o Consider 100 customer transactions for the system.
o The customers arrive at a random time (interarrival time) from {1, 2, 3, o Performance Measures
…, 8} minutes. Customer’s
Interarrival Time Cumulative Waiting Time in Idle Time of
Probability time in the
(minutes) Probability Queue (min.) Server (min.)
System (min)
1 0.125 0.125 0 4 0
2 0.125 0.250 3 5 0
3 0.125 0.375 4 9 0
4 0.125 0.500 3 7 0
5 0.125 0.625 4 5 0
6 0.125 0.750 0 5 2
7 0.125 0.875 …
8 0.125 1.000 1 3 0
174 491 101

04 Handout 1 *Property of STI

 student.feedback@sti.edu Page 2 of 5
 IT2032
The following are some statistics applied to the data: readily available in this application which can support the formulation of the
Average ∑ 𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑖𝑖𝑖𝑖 𝑞𝑞𝑞𝑞𝑞𝑞𝑞𝑞𝑞𝑞 174 interpretations through generating graphical representations based on the
waiting time:
𝑤𝑤 = = = 𝟏𝟏. 𝟕𝟕𝟕𝟕𝟕𝟕𝟕𝟕𝟕𝟕 gathered data, as seen in Figure 1.
𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑜𝑜𝑜𝑜 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 100
Probability that 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑜𝑜𝑜𝑜 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑤𝑤ℎ𝑜𝑜 𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤 46
a customer 𝑝𝑝(𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤) = = = 𝟎𝟎. 𝟒𝟒𝟒𝟒
has to wait: 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑜𝑜𝑜𝑜 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 100
Proportion of ∑ 𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑜𝑜𝑜𝑜 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 101
server idle 𝑝𝑝(𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠) = = = 𝟎𝟎. 𝟐𝟐𝟐𝟐
time: 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑟𝑟𝑟𝑟𝑟𝑟 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 418
∑ 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 317
𝑠𝑠 = = = 𝟑𝟑. 𝟏𝟏𝟏𝟏𝟏𝟏𝟏𝟏𝟏𝟏
𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑜𝑜𝑜𝑜 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 100
Average ∞
service time: 𝐸𝐸(𝑠𝑠) = � 𝑠𝑠 ⋅ 𝑝𝑝(𝑠𝑠) = 0.10(10) + 0.20(20) + ⋯ + 0.05(6)
𝑠𝑠=0
= 𝟑𝟑. 𝟐𝟐𝟐𝟐𝟐𝟐𝟐𝟐 Figure 1. Histogram for the average waiting time (50 trials/transactions only).
Source: Chapter 2: Simulation Examples by Prof. Dr. Mesut Güneş, n.d. slide 18
∑ 𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 415
Average time 𝜆𝜆 = = = 𝟒𝟒. 𝟏𝟏𝟏𝟏𝟏𝟏𝟏𝟏𝟏𝟏
between 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑜𝑜𝑜𝑜 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 − 1 100 − 1 The Process of Pseudorandom Number Generation
arrivals: 𝑎𝑎 + 𝑏𝑏 1−8 • Random number generator (RNG) is a mathematical concept, either
𝐸𝐸(𝜆𝜆) = = = 𝟒𝟒. 𝟓𝟓𝟓𝟓𝟓𝟓𝟓𝟓
2 2 computational or as a device, which is designed to generate a random set
Average
of numbers that should not display any distinguishable patterns in its
∑ 𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑖𝑖𝑖𝑖 𝑞𝑞𝑞𝑞𝑞𝑞𝑞𝑞𝑞𝑞 174 generation process. It is usually in the form of a function or blocks of code
waiting time of 𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤 = =
those who 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑜𝑜𝑜𝑜 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑤𝑤ℎ𝑜𝑜 𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤 54 used in software applications where an element of chance is required
waited: = 𝟑𝟑. 𝟐𝟐𝟐𝟐𝟐𝟐𝟐𝟐𝟐𝟐 (Techopedia, n.d.).
Average time a
• In RNG, the true randomness can be obtained with the inclusion of entropy
∑ 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶′𝑠𝑠 𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝑖𝑖𝑖𝑖 𝑡𝑡ℎ𝑒𝑒 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 491 from the natural environment through a device. Atmospheric noise, thermal
customer 𝑡𝑡 = = = 𝟒𝟒. 𝟗𝟗𝟗𝟗𝟗𝟗𝟗𝟗𝟗𝟗
spends in the 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑜𝑜𝑜𝑜 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 100 noise, and other quantum and electromagnetic phenomena are some
system: 𝑡𝑡 = 𝑤𝑤 + 𝑠𝑠 = 1.74 + 3.17 = 𝟒𝟒. 𝟗𝟗𝟗𝟗𝟗𝟗𝟗𝟗𝟗𝟗 examples of entropy from the natural environment.
• Pseudorandom number generator (PRNG) pertains to algorithms that use
Some interpretations: mathematical formulas to produce sequences of random numbers. PRNGs
o The average waiting time within the system is not high. generate a sequence of numbers approximating the properties of random
o The server is well loaded since the proportion of the server’s idle time is numbers, which are determined by a small group of initial values
less than one (1). (Techopedia, n.d.).
o Nearly half of the customers have to wait in the queue. • PRNG is also known as a deterministic random number generator. It is
o The average time a customer spends in the system is not long. fundamentally used in cryptographic mechanisms and key generation, as
they ensure message uniqueness.
Note that both the simulations and the Statistical computations can be processed
through the Microsoft Excel application. Specific formulas and functions are

04 Handout 1 *Property of STI

 student.feedback@sti.edu Page 3 of 5
 IT2032
Pseudorandom number generation in simulation (Chaturvedi, 2010): m – The modulus
• In a simulation, particularly with discrete-event systems simulation, a X0 – The initial value of the sequence (the seed)
pseudorandom number generator (PRNG) needs to generate random
variables with a variety of distributions depending on the system model, Assumptions:
which encompasses specific distribution parameters and an initial state, also o a and c are constants
called the seed. o a should be greater than zero (0) but less than m
• The generation is usually carried out by one (1) or more pseudorandom o c should be greater than or equal to zero (0) but less than m
number generator. These PRNGs generate random numbers in an event- o m should be greater than zero (0)
based manner each time an entity arrives for simulation.
• The pseudorandom numbers, as opposed to the true random numbers, Note: The selection of the values for a, c, m, and X0 drastically affects the
benefit the simulation when there is a need to rerun the simulation with statistical properties and the period length.
exactly the same set of behavior.
• The seed resets to its initial value each time a simulation starts making the Use the LCG algorithm and the following values to generate pseudorandom
random behavior repeatable. numbers:
a = 13; c = 0; m = 64; and X0 = 2
Characteristics of pseudorandom number generator (GeeksforGeeks, 𝑋𝑋𝑖𝑖+1 = (𝑎𝑎𝑎𝑎𝑖𝑖 + 𝑐𝑐) 𝑚𝑚𝑚𝑚𝑚𝑚 𝑚𝑚, 𝑖𝑖 ≥ 0
2019):
• Efficient – A pseudorandom number generator (PRNG) can produce 𝑋𝑋0+1 = 𝑋𝑋1 = (13 × 𝑋𝑋0 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64, 𝑖𝑖 = 0
multiple pseudorandom numbers in a short period of time, which makes it 𝑋𝑋1 = (13 × 2 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64
advantageous for applications that require many pseudorandom numbers 𝑋𝑋1 = 26 𝑚𝑚𝑚𝑚𝑚𝑚 64
for specific processes to work. 𝑿𝑿𝟏𝟏 = 𝟐𝟐𝟐𝟐
• Deterministic – In a PRNG, a given sequence of numbers can be 𝑋𝑋1+1 = 𝑋𝑋2 = (13 × 𝑋𝑋1 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64, 𝑖𝑖 = 1
reproduced at a later stage, and become an advantage, if the starting point 𝑋𝑋2 = (13 × 26 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64
of the sequence is known. 𝑋𝑋2 = 338 𝑚𝑚𝑚𝑚𝑚𝑚 64
• Periodic – PRNGs are periodic. This means that the sequence will 𝑿𝑿𝟐𝟐 = 𝟏𝟏𝟏𝟏
eventually repeat itself at some point in time (at a specific period). Despite 𝑋𝑋2+1 = 𝑋𝑋3 = (13 × 𝑋𝑋2 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64, 𝑖𝑖 = 2
being periodic, most PRNG algorithms cover a period that is so long that it 𝑋𝑋3 = (13 × 18 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64
can be ignored for most practical purposes. Proper choice of a, c, m, and X0 𝑋𝑋3 = 234 𝑚𝑚𝑚𝑚𝑚𝑚 64
can result in attaining the maximum density of a period and avoid cycling. 𝑿𝑿𝟑𝟑 = 𝟒𝟒𝟒𝟒
𝑋𝑋3+1 = 𝑋𝑋4 = (13 × 𝑋𝑋3 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64, 𝑖𝑖 = 3
Example of a pseudorandom number generator (PRNG) algorithm (Güneş, n.d.): 𝑋𝑋4 = (13 × 42 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64
Linear Congruential Generator (LCG) – It is one of the oldest and well-known 𝑋𝑋4 = 546 𝑚𝑚𝑚𝑚𝑚𝑚 64
PRNG algorithms used for generating sequences of random-like numbers in a 𝑿𝑿𝟒𝟒 = 𝟑𝟑𝟑𝟑
specific range, which is easy to understand and implement. To produce a 𝑋𝑋4+1 = 𝑋𝑋5 = (13 × 𝑋𝑋4 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64, 𝑖𝑖 = 4
sequence of pseudorandom numbers, this recursive relationship is defined: 𝑋𝑋5 = (13 × 34 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64
𝑿𝑿𝒊𝒊+𝟏𝟏 = (𝒂𝒂𝒂𝒂𝒊𝒊 + 𝒄𝒄) 𝒎𝒎𝒎𝒎𝒎𝒎 𝒎𝒎, 𝒊𝒊 ≥ 𝟎𝟎 𝑋𝑋5 = 442 𝑚𝑚𝑚𝑚𝑚𝑚 64
Where: 𝑿𝑿𝟓𝟓 = 𝟓𝟓𝟓𝟓
X – The sequence of pseudorandom numbers 𝑋𝑋5+1 = 𝑋𝑋6 = (13 × 𝑋𝑋5 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64, 𝑖𝑖 = 5
a – The multiplier 𝑋𝑋6 = (13 × 58 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64
c – The increment 𝑋𝑋6 = 754 𝑚𝑚𝑚𝑚𝑚𝑚 64
04 Handout 1 *Property of STI
 student.feedback@sti.edu Page 4 of 5
 IT2032
𝑿𝑿𝟔𝟔 = 𝟓𝟓𝟓𝟓 Kenton, W. (2020, November 1). What is queuing theory? In Investopedia.com. Retrieved on March 5, 2021,
𝑋𝑋6+1 = 𝑋𝑋7 = (13 × 𝑋𝑋6 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64, 𝑖𝑖 = 6 from https://www.investopedia.com/terms/q/queuing-theory.asp
Kuklin, P. (2016, April 20). How to improve your waiting lines: 10 common queue problems and solutions
𝑋𝑋7 = (13 × 50 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64 your customer will love. Retrieved on March 11, 2021, from https://www.lavi.com/en/resources-
𝑋𝑋7 = 650 𝑚𝑚𝑚𝑚𝑚𝑚 64 detail/10-ideas-to-solve-queuing-problems
𝑿𝑿𝟕𝟕 = 𝟏𝟏𝟏𝟏 O’Reilly Media, Inc. (2006, June). Quantitative techniques: Theory and problems. Retrieved on March 12,
𝑋𝑋7+1 = 𝑋𝑋8 = (13 × 𝑋𝑋7 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64, 𝑖𝑖 = 7 2021, from https://www.oreilly.com/library/view/quantitative-techniques-
𝑋𝑋8 = (13 × 10 + 0) 𝑚𝑚𝑚𝑚𝑚𝑚 64 theory/9789332512085/xhtml/ch9sec10.xhtml
Sokolowski, J. & Bank, C. (2009). Principles of modeling and simulation a multidisciplinary approach. John
𝑋𝑋8 = 130 𝑚𝑚𝑚𝑚𝑚𝑚 64 Wiley & Sons, Inc.
𝑿𝑿𝟖𝟖 = 𝟐𝟐 Techopedia. (n.d.). Pseudorandom number generator (PRNG). Retrieved on March 12, 2021 from
https://www.techopedia.com/definition/25842/pseudo-random-number-generator-prng
Some interpretations: Techopedia. (n.d.). Random number generator (RNG). Retrieved on March 12, 2021 from
o The seed (X0) greatly influences the sequence of the pseudorandom https://www.techopedia.com/definition/9091/random-number-generator-rng
Zeigler, B., Muzy, A. & Kofman, E. (2019). Theory of modeling and simulation discrete event and iterative
numbers. system computational foundations (3rd ed.). Academic Press – Elsevier Inc.
o The period encompasses a short cycle since, at i = 7, the value of X is
already equal to the initial seed.

Other pseudorandom number generator (PRNG) algorithms:

• Lagged Fibonacci Generator
• Linear Feedback Shift Register
• Blum Blum Shub
• Inverse Congruential Generator
• Advanced Randomized Systems

_______________________________________________________________
References:
Chaturvedi, D. (2010). Modeling and simulation of systems using MATLAB and Simulink. CRC Press – Taylor
& Francis Group, LLC
Dubois, G. (2018). Modeling and simulation challenges and best practices for industry. CRC Press – Taylor &
Francis Group, LLC
GeeksforGeeks. (2019, September 6). Pseudorandom number generator. Retrieved on March 15, 2021,
from https://www.geeksforgeeks.org/pseudo-random-number-generator-prng/
Güneş, M. (n.d.). Chapter 2: Simulation Examples. Retrieved on March 11, 2021, from https://www.mi.fu-
berlin.de/inf/groups/ag-
tech/teaching/2012_SS/L_19540_Modeling_and_Performance_Analysis_with_Simulation/02.pdf
Güneş, M. (n.d.). Chapter 6: Random-Number Generation. Retrieved on March 16, 2021, from
https://www.mi.fu-berlin.de/inf/groups/ag-
tech/teaching/2012_SS/L_19540_Modeling_and_Performance_Analysis_with_Simulation/06.pdf
Güneş, M. (n.d.). Chapter 8: Queueing Models. Retrieved on March 10, 2021, from https://www.mi.fu-
berlin.de/inf/groups/ag-tech/intern/19540-V-Simulation/08_Queueing_Models.pdf
JRNI. (2020, January 21). What are the advantages of a queuing system? Retrieved on March 11, 2021, from
https://www.jrni.com/blog/advantages-of-queuing-system

04 Handout 1 *Property of STI

 student.feedback@sti.edu Page 5 of 5