Project 1
Discrete Event Simulation of a Mass Vaccination Site
In this project, your group will implement some of the basic elements of an eventdriven
simulation of a system with two dependent queues. You will implement
queues, servers, event routines, and periodically print statistics. You will also
design and implement data structures and the routines to maintain them, and a
main routine to coordinate all parts of the simulation.The details of the system are
described below.
Background:
You are asked by the provincial government to design a mass vaccination site for
COVID-19 vaccinations. Vaccinations are organized at a large event venue that
has a maximum occupancy of B visitors (to accommodate physical distancing
guidelines). Vaccinations are administered from 9 AM to 5 PM daily. Vaccine
doses are given to visitors on a first-come, first-serve basis. Each visitor entering
the site needs to wait in line to get checked in (for ID verification). Check-in is
done by m1 check-in volunteers. After check-in, visitors wait in another line to
receive their vaccine dose by m2 nurses. After a vaccine dose is administered, each
visitor needs to wait for t minutes (to make sure they don’t have any obvious side
effects) before departing the vaccination site.
System to Simulate:
The system consists of two queues, one after the other. Once visitors enter the site,
they stay in either the check-in queue, the vaccination queue, or the waiting
area. The first “check-in” queue has m1 servers. The second “vaccination” queue
has m2 servers. The total number of visitors in the whole system should not
exceed system capacity B. This maximum capacity includes visitors in both queues,
visitors being served and visitors in the waiting area. Visitors arrive and enter the
check-in queue if the system is not full. The visitor at the head of each queue
receives service when any of the servers becomes available. Each visitor has to pass
both queues and the waiting area before departing the system. Simulation will
start at the beginning of the day (9 AM) end when the last visitor that entered
before or at 5 PM leaves the site.
Queues:
You will implement two M/M/m queues. You can assume that visitors come from
an infinite population, and that vaccine supply is sufficient to vaccinate all visitors
who enter the site. Assume that the queueing discipline for each queue is FCFS
(First-come, First-served). Arriving visitors leave immediately if the site is at full
capacity and do not return the same day.
Random Variate Generation:
Visitor interarrival time and both queues’ service time distributions are
exponential. Arrivals are generated based on the mean interarrival time 1/
λ minutes. Service time of the check-in queue is based on a service time
distribution with mean service time 1/µ1 minutes. Service time of the vaccination
queue is based on a service time distribution with mean service time 1/µ2 minutes.
Time in the waiting area is deterministic at t minutes. You will need to write a
separate function for each distribution to generate values for arrival times and
service times per server based on the exponential distribution. You should use C’s
uniformly-distributed random number generator as the basis for your random
variate generation (similar to HW3).
Event Routines:
You need to write a separate function for each of the event routines: Arrival, start
service of check-in queue, depart from check-in queue and arrive at vaccination
queue, start service at vaccination queue, depart from vaccination queue and enter
waiting area, depart waiting area and system.
End of Simulation:
Simulation starts at 9 AM. Simulation ends when the last visitor who arrived
before or at 5 PM leaves the system. No visitors are allowed to enter the system
after 5 PM.
Performance Metrics:
You need to keep track of the following performance metrics for the duration of
the simulation:
• Current number of visitors in system
• Response time for each departing visitor (time between when a visitor
enters and leaves the whole system)
• Waiting time in check-in queue for each visitor that starts service with
a check-in volunteer
• Waiting time in vaccination queue for each visitor that starts service
with a vaccination nurse
• Number of visitors who leave the system because it’s at full capacity
when they arrive
Report Generation:
Using the above statistics (and others you might want to track), you need to
generate a report with these performance metrics:
• Total number of departures, or the total number of vaccinated visitors
leaving the site
• Average number in system
• Average response time
• Average waiting time in check-in queue
• Average waiting time in vaccination queue
• Number of visitors who leave the system because it's at full capacity
when they arrive.
This report should be generated periodically after each hour of simulated time
(i.e., at 10 AM, 11 AM, 12 PM, ..., 5 PM), as well as at the end of simulation. You
need to compute these statistics during runtime.
Your simulation should correctly handle any correct choice of input
parameters λ, µ1, µ2, t, B, m1 and m2. All these parameters should be
arguments to call the program, in addition to the random seed S.
Submission instructions:
Submit a single file “proj1.c” containing the source code of your simulation
program (including all the data structure definitions and appropriate comments).
The first line in the file should be a comment that includes the command line you
use to compile the program. Your executable should run using the following
command line parameter format:
./proj1 λ µ1 µ2 t B m1 m2 S
The arrival and service rates as well as wait time t can be real numbers, but the
remaining parameters are all integers.
For example: ./proj1 10 0.5 0.25 15 500 30 50 821
Should simulate a system with arrival rate of 10 visitors per minute, service rate
per check-in server of 0.5 visitors per minute, service rate per vaccination server
of 0.25 visitors per minute, a wait time of 15 minutes, site capacity of 500, 30
check-in servers, 50 vaccination servers, and a random seed of 821.
Important Note: Your code should be readable. It should use easy to understand
variable and function names. It should include a lot of comments to enable others
to understand it. Please err on the side of more comments and more explanation.
The quality of your code and ease of understanding it will be part of your grade.
Choosing efficient data structures that speed up simulation time is a plus.