UNIT-2

Real Time Scheduling: Common Approaches to Real Time Scheduling: Clock Driven
Approach, Weighted Round Robin Approach, Priority Driven Approach, Dynamic
Versus Static Systems, Optimality of Effective-Deadline-First (EDF) and Least-Slack-
Time-First (LST) Algorithms, Rate Monotonic Algorithm, Offline Versus Online
Scheduling, Scheduling Aperiodic and Sporadic jobs in Priority Driven and Clock
Driven Systems.
Real-Time Scheduling is designed to ensure that tasks are
executed in a predictable and timely manner, with minimal delay or
latency.
●The scheduler assigns priorities to tasks based on their importance
and urgency and allocates system resources accordingly to ensure
that critical tasks are executed on time.
Real-time scheduling can be classified into two categories: hard real-time
scheduling and soft real-time scheduling.
1. Hard real-time scheduling, requires that tasks are completed
within a specific deadline, and any delay can result in system
failure or loss of data.
2. Soft real-time scheduling, on the other hand, allows for some
degree of delay but still requires that tasks are completed within a
reasonable timeframe.
Common Approaches to Real-Time Scheduling: There are several common approaches to real-
time scheduling that are used to ensure that tasks are executed in a timely and efficient manner.
These approaches include
1. Clock-Driven Approach:
• The Clock Driven approach is a real-time scheduling technique that uses a fixed clock to divide
time into equal intervals and assigns tasks to each interval based on their deadline and priority.
• In this approach, tasks are assigned to a fixed time slot, and the scheduler ensures that each task
is executed within its allocated
time slot.
• If a task misses its deadline, it is either rescheduled or dropped, depending on the criticality of
the task.
• The Clock Driven approach is helpful in systems that have a high degree of predictability and
where tasks have fixed execution times.
• It is commonly used in embedded systems, where tasks are executed in a deterministic and
predictable manner.
• One advantage of the Clock Driven approach is that it provides a
simple and efficient way to allocate system resources, as tasks are
scheduled based on their priority and deadline.
• it can be less flexible than other scheduling techniques, as it assumes
that all tasks have fixed execution times and cannot adapt to changing
system conditions.
Here's an example of how clock-driven scheduling works:
Execution Time (C) Task A Task B Task C
0 ms
10 ms A
20 ms B
30 ms C
40 ms A
50 ms B
60 ms C
70 ms A
80 ms B
90 ms C
100 ms A
In this example,
● The system clock is divided into time slots of 10 ms, and three tasks, A, B, and C,
are scheduled to run.
● Task A is assigned the first time slot, task B is assigned the second time slot, and
task C is assigned the third time slot.
● At time 0 ms, no tasks are running.
● At time 10 ms, task A starts running and continues until the end of its time slot at
20 ms.
● At time 20 ms, task B starts running and continues until the end of its time slot at
30 ms.
● At time 30 ms, task C starts running and continues until the end of its time slot
at 40 ms.
● The process repeats until all tasks have completed their execution.
Advantages:
• The important advantage is conceptual simplicity.
• Time triggered System Based on clock driven scheduling is easy to validate, test
and certify.
• Clock driven scheduling paradigm are time triggered. in this systems interrupts to
external events are queued and polled periodically.
Disadvantages:
• The system based on clock driven approach is Brittle (soft).
• The release time of all jobs must be fixed.
• The pure clock driven approach is not suitable for many systems that contain
both hard and soft real time applications.
• In this system all combinations of periodic task that might execute at same time.
2. Weighted Round Robin (WRR):
● Weighted Round Robin (WRR) is a real-time scheduling algorithm that is an
extension of the Round Robin approach.
● In this approach, tasks are assigned a weight value that determines the amount of
CPU time they receive during each round.
● Tasks with higher weight values are allocated more CPU time than tasks with lower
weight values.
Here is an example diagram of the weighted round-robin approach:
Task A: weight 3, requires 15 ms to complete
Task B: weight 2, requires 10 ms to complete
Task C: weight 1, requires 5 ms to complete
Time quantum: 5 ms
Execution Time (C) Task A Task B Task C

0 - 5 ms A

5 - 10 ms B

10 -15 ms 5

15 -20 ms A -

20 -25 ms B -

25 -30 ms A - -
• At time 0 ms, the scheduler starts with Task A since it has the highest weight.
• Task A is executed for the first time quantum of 5 ms, until time 5ms.
• Since Task A has not been completed, it is moved to the end of the queue, and the
scheduler switches to Task B, which is executed for the next time quantum of 5 ms,
until time 10 ms.
• Since Task B has not been completed, it is moved to the end of the queue, and the
scheduler switches to Task C, which is executed for the next time quantum of 5 ms,
until 15 ms.
• Since Task C has been completed, it is removed from the schedule and the scheduler
switches back to Task A, which is executed for the next time quantum of 5 ms, until
time 20 ms.
• Since Task A has not been completed, it is moved to the end of the queue, and the
scheduler switches back to Task B, which is executed for the next time quantum of 5
ms, until time 25 ms.
• Since Task B has been completed, it is removed from the schedule, and the scheduler
switches back to Task A, which is executed for the final time quantum of 5 ms, until
time 30 ms
Example-2
3. Priority-Driven Approach:
• The priority-driven approach is a real-time scheduling algorithm that
assigns a priority to each task, based on its importance or urgency.
• The scheduler then schedules tasks based on their priority, with
higher-priority tasks being executed before lower-priority tasks.
Here is an example of the priority-driven approach:
• Task A: priority 1, requires 20 ms to complete
Task B: priority 2, requires 15 ms to complete
Task C: priority 3, requires 10 ms to complete
 Task A Task B Task C

20 ms 15 ms 10 ms

0-20 20-35 35-45

• At time 0 ms, the scheduler starts with Task A since it has the highest priority.
• Task A is executed until it completes at 20 ms.
• Since Task A has been completed, the scheduler switches to Task B,
which is the next highest priority task.
• Task B is executed until it completes at 35 ms.
• Finally, the scheduler switches to Task C, which is the lowest-priority
task.
• Task C is executed until it completes at 45 ms.
 Dynamic Vs Static Systems:
Feature Dynamic System Static System
Definition A system that changes over time in A system that does not change over time
response to external events or inputs and has a fixed configuration
Task arrival Tasks arrive at runtime All tasks are known in advance
Scheduling High, due to the need to respond to Low, since the schedule is fixed
complexity runtime changes
Resource Resources are allocated dynamically Resources can be pre-allocated
allocation based on the current workload or based on known requirements
demand
Less predictable, due to runtime More predictable since all requirements
Predictability changes in task arrival and resource are known in advance
demands
Flexibility More flexible since the system can Less flexible since the system is fixed
adapt to changing demands
Performance May be lower due to runtime Higher, since the system can be
overhead and resource allocation delays optimized for known requirements
Online reservation systems, traffic Assembly lines, air traffic control systems,
Examples management systems, adaptive control power grid systems
systems
The Effective Deadline First (EDF) algorithm:
● The Effective Deadline First (EDF) algorithm is a real-time scheduling
algorithm that schedules tasks based on their deadline.
● The task with the earliest deadline is given the highest priority and is
scheduled first.
● This algorithm ensures that tasks with the shortest deadlines are
executed first, which can help to meet critical timing requirements in
real-time systems.
Here's an example of how the EDF algorithm
works:
Task A: deadline 5 ms, requires 10 ms to complete
Task B: deadline 3 ms, requires 5 ms to complete
Task C: deadline 7 ms, requires 15 ms to complete
Execution Time (C) Task A Task B Task C

0 - 5 ms B

5 - 15 ms A

15 - 30 ms C
• At time 0 ms, the scheduler starts with Task B since it has the earliest
deadline.
• Task B is executed until it completes at 5 ms.
• Since Task B has been completed, the scheduler switches to Task A,
which has the next earliest deadline.
• Task A is executed until it completes at 15 ms.
• Finally, the scheduler switches to Task C, which has the latest
deadline. Task C is executed until it completes at 30 ms.
Example-2
Least-Slack-Time-First (LST) Algorithms:
● Least-Slack-Time-First (LST) is a real-time scheduling algorithm that
is used to optimize the scheduling of tasks based on their deadlines
and processing times.
● LST chooses the task with the least slack time as the next task to be
executed.
● Slack time is defined as the difference between a task's deadline and
the time remaining until the task must be completed.
• For example, if a task has a deadline of 100ms and has already been
running for 80ms, then its slack time is 20ms (i.e., the difference
between 100ms and 80ms).
The LST algorithm works as follows:
1. Initialize the slack time for all tasks in the system.
2. Select the task with the smallest slack time as the next task to be
executed.
3. Execute the selected task for the remaining slack time.
4. Recalculate the slack time for all tasks that have not been
completed.
5. Repeat steps 2-4 until all tasks have been completed.
Remark: If a task's slack time is negative, it means that the task has
already exceeded its deadline, and it should be scheduled immediately
to avoid missing the deadline. In this case, LST would prioritize the task
with the smallest negative slack time.
Here's an example of how the LST algorithm works:
Task A: deadline 50 ms, requires 20 ms to complete
Task B: deadline 30 ms, requires 15 ms to complete
Task C: deadline 40 ms, requires 10 ms to complete
Execution Time (C) Task A Task B Task C

0 - 15 ms B

15 - 25 ms C

25 - 45 ms A
• At time 0 ms, all tasks are ready to be executed, so the scheduler
selects the task with the smallest slack time, which is Task B.
• Task B is executed until it completes at 15 ms.
• At 15 ms, the scheduler checks the slack times of the remaining tasks.
• Task A has a slack time of 35 ms, and Task C has a slack time of 25 ms.
• Since Task C has the smallest slack time, it is given the highest priority and
is executed next.
• Task C is executed until it completes at 25 ms.
• Finally, the scheduler switches to Task A, which has the largest slack time.
• Task A is executed until it completes at 45 ms.
 Effective Deadline First (EDF) `Vs Least-Slack-Time-First (LST)
Algorithms:
Feature EDF LST

Definition Prioritizes tasks based on their Prioritizes tasks based on the amount
relative deadlines. of time remaining until their deadlines.

Type of scheduling Preemptive Preemptive

Algorithm complexity Moderate Moderate

Real-time systems Suitable for hard real-time Suitable for Soft real-time systems
systems
Deadline Requirement Each task must have a deadline
specified.
Each task must have a deadline specifi
ed.
Tasks with the least slack time (time
Selection Criteria Tasks with the earliest deadlines remaining until the deadline) are
are given priority. given
priority.
Utilization is Better for high-utilization Better for low utilization systems.
systems.
Rate Monotonic Algorithm:
• The Rate Monotonic (RM) algorithm is a real-time scheduling
algorithm that assigns priorities to tasks based on their periods, with
shorter-period tasks having higher priority.
• This algorithm assumes that the tasks are periodic and that their
execution times are constant and known in advance.
Here's an example of how the RM algorithm works:
• Task A: period 50 ms, requires 20 ms to complete
• Task B: period 30 ms, requires 15 ms to complete
• Task C: period 40 ms, requires 10 ms to complete
Execution Time (C) Task A Task B Task C

0 - 15 ms B

15 - 25 ms C

25 - 45 ms A
• At time 0 ms, all tasks are ready to be executed, so the scheduler selects
the task with the highest priority, which is Task B.
● Task B is executed until it completes at 15 ms.
● At 15 ms, the scheduler checks which task is ready to be executed next.
● Since Task C has a period of 40 ms, it is prepared to be executed next.
● Task C is executed until it completes at 25 ms.
● At 25 ms, the scheduler checks which task is prepared to be executed
next.
● Since Task A has a period of 50 ms, it is prepared to be executed next.
● Task A is executed until it completes at 45 ms.
The scheduler then repeats the cycle, executing the tasks in the order of
priority.
Example-2
Offline Vs Online Scheduling:

Offline Scheduling Online Scheduling

Offline scheduling involves creating a schedule online scheduling involves scheduling tasks
for all tasks in advance, before the execution of dynamically during the execution of the
the system. system.
Offline Vs Online Scheduling:

The schedule is then followed during the The scheduler determines which task to execute
execution of the system. next based on the current state of the system,
the task's priority, and its timing requirements.

In offline scheduling,the scheduler has This approach is suitable when the task workload
complete information about the tasks and and timing requirements are not known in
advance,
can optimize the schedule accordingly.

this approach does not handle unexpected Online scheduling is more exible and adaptable to
events or changes in task requirements changes, but it may not be able to optimize the
that occur during the execution. schedule as effectively as offline scheduling
Difference between Sporadic and Aperiodic Real-time Tasks:
SPORADIC TASK Aperiodic Task
It has a hard deadline. It has a soft deadline or no deadline.

It is a highly critical task. It is a low or moderate critical task.

The minimum separation between two the minimum separation between two
consecutive instances can not consecutive instances can be zero.
be zero.
It includes hard real-time tasks. It includes soft real-time tasks.
The deadline for all instances of Meeting the deadline of all
sporadic tasks can be met easily. instances
of aperiodic tasks is difficult.
It goes through an acceptance No testing is performed on the
test. aperiodic task.
It is executed only when sufficient Its execution does not depend on
slack time is available. available slack time.

It gets rejected by the scheduler It never gets rejected by the scheduler.

when there is less slack time.

It includes commands given by the system. It includes interactive commands given by the
user.

Example: Security alert program in the system. Example: Logging task in the system.
Scheduling Aperiodic and Sporadic jobs in Priority Driven:
● In priority-driven scheduling, tasks are assigned priorities based on their importance and urgency.
● The scheduler then selects the task with the highest priority to execute next.
● Aperiodic and sporadic jobs are two types of tasks that can be scheduled in a priority-driven system.
Aperiodic tasks are tasks that do not have a regular or periodic execution pattern, and they can
arrive at any time.
● These tasks have specific deadlines or response times, and they must be executed within those
deadlines.
● Examples of aperiodic tasks include user inputs, interrupts, and system events.
Sporadic tasks are a type of aperiodic task that has a minimum inter-arrival time between
successive task arrivals.
● Sporadic tasks have a deadline or response time, just like aperiodic tasks, but they also have a
minimum time between successive arrivals.
● Examples of sporadic tasks include periodic system health checks, periodic communication tasks, or
periodic maintenance tasks.
Priority-driven scheduling can handle both aperiodic and sporadic tasks by
assigning priorities to them based on their timing requirements and
importance. A high-priority task can preempt a low-priority task, allowing the
system to ensure that aperiodic and sporadic tasks meet their timing
requirements.
For example, let's consider a system that has three tasks:
1. Aperiodic task A, sporadic task B, and periodic task C.
2. Task A has a deadline of 50 ms, task B has a minimum inter-arrival time of
100 ms and a deadline of 150 ms, and task C has a period of 200 ms and a
deadline of 190 ms.
3. The system assigns task A the highest priority, followed by task B and task C.
4. If task A arrives at time 0, the scheduler will execute it immediately since it
has the highest priority.
5. If task B arrives at a time of 70 ms, the scheduler will preempt the execution
of task A and execute task B since it has a higher priority than task A.
6. Once task B is complete, the scheduler will return to executing task
A.
7. If task C arrives at a time of 200 ms, the scheduler will execute it since it is
the highest-priority task at that time.
8. If task C is not complete before its deadline, it will be considered a missed
deadline.
In this way, priority-driven scheduling can handle aperiodic and sporadic tasks
by assigning priorities based on their timing requirements and ensuring they
meet their deadlines
Scheduling Aperiodic and Sporadic jobs in Clock-Driven Systems:
● Clock-driven scheduling systems are used in real-time systems to
schedule periodic tasks that execute at regular intervals.
● However, these systems can also support aperiodic and sporadic tasks
with the help of additional scheduling algorithms.
● One way to schedule aperiodic and sporadic tasks in clock-driven
systems is to use priority-based scheduling algorithms.
● In these algorithms, tasks are assigned priorities based on their
importance, and the scheduler selects the task with the highest priority
for execution.
For example, consider a clock-driven system that has two
periodic tasks that execute every 50ms and 100ms
respectively.
1. The system also has two aperiodic tasks and one sporadic task.
2. Task A is an aperiodic task that requires immediate attention, task
B is an aperiodic task that can wait for up to 30ms, and task C is a
sporadic task that must execute within 20ms of its arrival.
3. Task A can be assigned the highest priority, followed by task C and task B.
4. The periodic tasks can be assigned lower priorities than the aperiodic and sporadic
tasks.
5. During execution, the scheduler checks for new arrivals of aperiodic and sporadic tasks
and assigns them priorities based on their requirements.
6. The scheduler then selects the task with the highest priority for execution.