Round Robin Scheduling

Round Robin is a preemptive CPU scheduling algorithm used in multitasking operating systems. It ensures that each process gets a fair share of the CPU time by allowing them to run in a cyclic queue for a fixed time slice called the time quantum. 

 

The Round Robin (RR) CPU scheduling algorithm is one of the most common algorithms used in operating systems to manage the execution of processes. It's designed to provide fair allocation of CPU time to all processes in the system. Here's how it works:

1. Queue: The ready queue contains all the processes that are ready to execute. In Round Robin, this queue is typically implemented as a first-in-first-out (FIFO) queue.

2. Time Quantum: A fixed time quantum (also known as a time slice or time interval) is defined. This time quantum represents the maximum amount of CPU time that a process can have before it's preempted.

3. Execution: The CPU scheduler selects the process at the front of the ready queue for execution. The selected process is allowed to run for the duration of one time quantum or until it voluntarily relinquishes the CPU, whichever comes first.

4. Preemption: At the end of the time quantum, if the process hasn't completed its execution, it's preempted and moved to the back of the ready queue. This allows other processes to get a chance to execute.

5. Next Process Selection: The CPU scheduler then selects the next process from the ready queue for execution. This process can be the same process that just completed its time quantum (if it's still ready), or it can be the next process in the queue.

6. Repeat: Steps 3-5 are repeated continuously, with processes being executed in a circular manner until they all complete their execution.

Round Robin scheduling ensures fairness by giving each process an equal share of the CPU time over a period. It's particularly useful in time-sharing systems where multiple users are interacting with the system simultaneously, as it prevents any single process from monopolizing the CPU for an extended period. However, the choice of time quantum is crucial, as a too-short quantum may lead to high overhead due to frequent context switches, while a too-long quantum may result in poor responsiveness.

 

Let’s dive into an example to understand how it works:

Consider the following three processes with their burst times:

Process P1: Burst time = 4

Process P2: Burst time = 3

Process P3: Burst time = 5

Assume a time quantum of 2 seconds.

Step 1: Execution begins with P1 (burst time 4). P2 and P3 are still in the waiting queue.

Step 2: At time = 2, P1 is added to the end of the queue, and P2 starts executing.

Step 3: At time = 4, P2 is preempted and added to the end of the queue. P3 starts executing.

Step 4: At time = 6, P3 is preempted and added to the end of the queue. P1 starts executing.

Step 5: At time = 8, P1 completes its burst time of 4. P2 starts execution.

Step 6: P2 has a burst time of 3. It has already executed for 2 seconds. At time = 9, P2 completes execution. Then, P3 starts execution until it completes.

 

 Average waiting time:

Waiting time for P1: 0 + 4 = 4

Waiting time for P2: 2 + 4 = 6

Waiting time for P3: 4 + 3 = 7

Average WT=5.666667 

Average Turn Around Time

Turn Around Time For P1=8

Turn Around Time For P2=9

Turn Around Time For P3=12

Average TAT=9.666667

In summary, Round Robin ensures fairness by allowing each process to run for a fixed time slice, preventing starvation and providing predictable response times. It’s widely used in traditional operating systems.

//**************************************************

// C Program Implementation Round Robin Scheduling

//**************************************************

#include <stdio.h>

// Structure to represent a process

struct Process {

    int id;

    int burst_time;

};

// Function to simulate Round Robin scheduling

void roundRobin(struct Process processes[], int n, int time_quantum) {

    int remaining_time[n];

    int wt[n],tat[n];

    for (int i = 0; i < n; ++i)

        remaining_time[i] = processes[i].burst_time;

    int current_time = 0;

    int completed = 0;

    while (completed < n) {

        int flag = 0;

        for (int i = 0; i < n; ++i) {

            if (remaining_time[i] > 0) {

                flag = 1;

                if (remaining_time[i] <= time_quantum) {

                    current_time += remaining_time[i];

                    remaining_time[i] = 0;

                    printf("Process P%d completed at time %d\n", processes[i].id, current_time);

                    tat[i]=current_time;

                    wt[i]=tat[i]-processes[i].burst_time;

                    printf("Process P%d WT=%d TAT=%d\n",processes[i].id,wt[i],tat[i]);

                    completed++;

                } else {

                    current_time += time_quantum;

                    remaining_time[i] -= time_quantum;

                }

            }

        }

        if (!flag)

            break;

    }

    float awt=0;

    float atat=0;

    for(int i=0;i<n;i++)

    {

        awt=awt+wt[i];

        atat=atat+tat[i];

    }

    printf("Average WT=%f Average TAT=%f",awt/n,atat/n);

    

}

int main() {

    int n = 3; // Number of processes

    struct Process processes[] = {{1, 4}, {2, 3}, {3, 5}};

    int time_quantum = 2;

    printf("Round Robin Scheduling:\n");

    roundRobin(processes, n, time_quantum);

    return 0;

}

Output:

Round Robin Scheduling:

Process P1 completed at time 8

Process P1 WT=4 TAT=8

Process P2 completed at time 9

Process P2 WT=6 TAT=9

Process P3 completed at time 12

Process P3 WT=7 TAT=12

Average WT=5.666667 Average TAT=9.666667

 

In the previous implementation we assumed that all process arrive at time 0.Lets consider a detailed implementation with linked list.Here the processes arrive in different time.

There are six processes named as P1, P2, P3, P4, P5 and P6. Their arrival time and burst time are given below in the table. The time quantum of the system is 4 units.
Process ID    Arrival Time        Burst Time
1                             0                         5
2                             1                         6
3                             2                         3
4                             3                         1
5                             4                         5
6                             6                         4

GANTT CHART

Note: In the following implementation we will assume that the processes are in the order of arrival time.The print queue function can be used to print the ready queue.

// C Program Implementation 

#include <stdio.h>
#include <stdio.h>
#include <stdlib.h>

// Structure to represent a process
typedef struct {
    int name;      // Process name (1,2,3 ...)
    int arrival;    // Arrival time
    int burst;      // Burst time
    int remaining;  // Remaining burst time
    int status;  // in queue or not
    int completion;//completion time
    int waiting; // waiting time
    int turnaround; // turn around time
} Process;

// Structure to represent a node in the waiting queue
typedef struct Node {
    Process process;
    struct Node* next;
} Node;
Node *front=NULL;
Node *rear=NULL;

// Function to enqueue a process into the waiting queue
void enqueue(Process process) {
    Node* newNode = (Node*)malloc(sizeof(Node));
 
    newNode->process = process;
    newNode->next = NULL;

    if (front == NULL) {
        front=rear = newNode;
    } else {
        rear->next = newNode;
        rear = newNode;
    }
}

// Function to dequeue a process from the waiting queue
Process dequeue() {
   
    Node* temp = front;
    Process process = temp->process;
    front = front->next;
    free(temp);
    return process;
}
void printqueue()
{
    Node *temp=front;
    while(temp!=NULL)
    {
        printf("%d-->",temp->process.name);
        temp=temp->next;
        
    }
    printf("NULL\n");
}
int main() {
    // Define the processes
    Process processes[] = {
        {1, 0, 5, 5,0,-1,-1,-1},  
        {2, 1, 6, 6,0,-1,-1,-1},  
        {3, 2, 3, 3,0,-1,-1,-1},
        {4, 3, 1, 1,0,-1,-1,-1},  
        {5, 4, 5, 5,0,-1,-1,-1},  
        {6, 6, 4, 4,0,-1,-1,-1}
        
    };
    int n = sizeof(processes) / sizeof(Process); // Number of processes

    // Time quantum
    int quantum = 4;

    // Initialize variables
    int currenttime = 0;
    int completed = 0;
   
    printf("Time  |  Process\n");
    printf("------+-------------------\n");

    // Loop until all processes are completed
     enqueue(processes[0]);
     processes[0].status=1;
    while (completed < n ) {
        // Enqueue processes that arrive at the current time
        
       //  printqueue();
        // Execute processes in the waiting queue
        if (front != NULL) {
            Process currentProcess = dequeue();
            int execution_time = (currentProcess.remaining < quantum) ? currentProcess.remaining : quantum;
            currentProcess.remaining -= execution_time;
            currenttime += execution_time;
            for (int i = 0; i < n; i++) {
            if (processes[i].arrival <= currenttime && processes[i].status!=1) {
                enqueue(processes[i]);
                processes[i].status=1;
            }
        }
            printf("%-5d |  P%d\n", currenttime, currentProcess.name);
            if (currentProcess.remaining <= 0) {
                completed++;
                for(int i=0;i<n;i++) if (processes[i].name==currentProcess.name)
                           {processes[i].completion=currenttime;break;}
            } else {
                currentProcess.status=1;
                enqueue(currentProcess);
            }
        } else {
            currenttime++; // Move to the next time unit if no process is ready to execute
        }
    }

    float avwt=0.0;
    float avtat=0.0;
    printf("Process ArrivalTime BurstTime  CompletionTime  TurnAroundTime WaitingTime\n");
    for(int i=0;i<n;i++)
    {
    processes[i].turnaround=processes[i].completion-processes[i].arrival;
    processes[i].waiting=processes[i].turnaround-processes[i].burst;
    printf("P%d%10d%10d%20d%10d%20d\n",processes[i].name,processes[i].arrival,processes[i].burst,processes[i].completion,processes[i].turnaround,processes[i].waiting);
    avwt=avwt+processes[i].waiting;
    avtat=avtat+processes[i].turnaround;
    }
    printf("Average Waiting Time=%f\n",avwt/n);
    printf("Average Turn Around Time=%f\n",avtat/n);
      return 0;  
}

Output:

Time  |  Process

------+-------------------

4     |  P1

8     |  P2

11    |  P3

12    |  P4

16    |  P5

17    |  P1

21    |  P6

23    |  P2

24    |  P5

Process ArrivalTime BurstTime  CompletionTime  TurnAroundTime WaitingTime

P1                 0             5                      17                            17                  12

P2                 1             6                      23                            22                  16

P3                 2             3                      11                             9                   6

P4                 3             1                      12                             9                   8

P5                 4             5                      24                            20                  15

P6                 6             4                      21                            15                  11

Average Waiting Time=11.333333

Average Turn Around Time=15.333333