Predicting Burst Time | SJF Scheduling

SJF Scheduling-

 

Before you go through this article, make sure that you have gone through the previous article on SJF Scheduling.

 

In SJF Scheduling,

• Out of all the available processes, CPU is assigned to the process having smallest burst time.
• The main drawback of SJF Scheduling is that it can not be implemented practically.
• This is because burst time of the processes can not be known in advance.

 

In this article, we will discuss techniques to predict burst time.

 

Techniques to Predict Burst Time-

 

There are several techniques which try to predict the burst time for the processes so that the algorithm can be implemented.

These techniques are-

 

 

Static Techniques-

 

There are two static techniques-

1. Based on process size
2. Based on process type

 

1. Based on Process Size-

 

• This technique predicts the burst time for a process based on its size.
• Burst time of the already executed process of similar size is taken as the burst time for the process to be executed.

 

Example-

 

• Consider a process of size 200 KB took 20 units of time to complete its execution.
• Then, burst time for any future process having size around 200 KB can be taken as 20 units.

 

NOTE The predicted burst time may not always be right. This is because the burst time of a process also depends on what kind of a process it is.

 

2. Based on Process Type-

 

• This technique predicts the burst time for a process based on its type.
• The following figure shows the burst time assumed for several kinds of processes.

 

 

Dynamic Techniques-

 

There are two dynamic techniques-

1. Based on simple averaging
2. Based on exponential averaging

 

1. Based on Simple Averaging-

 

• Burst time for the process to be executed is taken as the average of all the processes that are executed till now.
• Given n processes P₁, P₂, … , P_n and burst time of each process P_i as t_i, then predicted burst time for process P_n+1 is given as-

 

 

2. Based on Exponential Averaging-

 

• Given n processes P₁, P₂, … , P_n and burst time of each process P_i as t_i. Then, predicted burst time for process P_n+1 is given as-

 

 

where-

• α is called smoothening factor (0<= α <=1)
• t_n = actual burst time of process P_n
• T_n = Predicted burst time for process P_n

 

PRACTICE PROBLEM BASED ON PREDICTING BURST TIME-

 

Problem-

 

Calculate the predicted burst time using exponential averaging for the fifth process if the predicted burst time for the first process is 10 units and actual burst time of the first four processes is 4, 8, 6 and 7 units respectively. Given α = 0.5.

 

Solution-

 

Given-

• Predicted burst time for 1^st process = 10 units
• Actual burst time of the first four processes = 4, 8, 6, 7
• α = 0.5

 

Predicted Burst Time for 2nd Process-

 

Predicted burst time for 2^nd process

= α x Actual burst time of 1^st process + (1-α) x Predicted burst time for 1^st process

= 0.5 x 4 + 0.5 x 10

= 2 + 5

= 7 units

 

Predicted Burst Time for 3rd Process-

 

Predicted burst time for 3^rd process

= α x Actual burst time of 2^nd process + (1-α) x Predicted burst time for 2^nd process

= 0.5 x 8 + 0.5 x 7

= 4 + 3.5

= 7.5 units

 

Predicted Burst Time for 4th Process-

 

Predicted burst time for 4^th process

= α x Actual burst time of 3^rd process + (1-α) x Predicted burst time for 3^rd process

= 0.5 x 6 + 0.5 x 7.5

= 3 + 3.75

= 6.75 units

 

Predicted Burst Time for 5th Process-

 

Predicted burst time for 5^th process

= α x Actual burst time of 4^th process + (1-α) x Predicted burst time for 4^th process

= 0.5 x 7 + 0.5 x 6.75

= 3.5 + 3.375

= 6.875 units

 

To gain better understanding about predicting burst time,

Watch this Video Lecture

 

Next Article- LJF Scheduling | LRTF Scheduling

 

Get more notes and other study material of Operating System.

Watch video lectures by visiting our YouTube channel LearnVidFun.