SRTF Implementation
Duration: 7 min
This video lesson is available to enrolled students.
AI Summary
An AI-generated summary of this video lecture.
This educational video lecture focuses on the Shortest Job First (SJF) CPU scheduling algorithm and addresses its primary limitation: the inability to know the exact CPU burst time of a process beforehand. To overcome this, the instructor introduces the concept of predicting the next CPU burst using the exponential averaging technique. The lecture details the mathematical formula for this prediction, defining the variables involved, including the smoothing factor alpha. The instructor provides a step-by-step calculation example on a whiteboard and demonstrates the application of the formula using a process table. Finally, the video concludes by discussing the theoretical nature of this scheme, noting its limitations when burst times vary significantly or remain constant.
Chapters
0:00 – 2:00 00:00-02:00
The lecture begins with the instructor standing in front of a slide that states, 'As SJF is not implementable, we can use the one technique where we try to predict the CPU burst of the next coming process.' He explains that since the Shortest Job First algorithm requires knowledge of the burst time which is not available, a prediction method is necessary. The slide text further clarifies, 'The method is used as exponential averaging technique, where we consider the previous value and previous prediction.' The instructor emphasizes that this technique relies on historical data to estimate future behavior, setting the stage for the mathematical derivation to follow.
2:00 – 5:00 02:00-05:00
The instructor transitions to the core formula displayed on the screen: 'Tau(n+1) = alpa tn + (1- alpa) tau n'. He explains that Tau(n+1) represents the predicted burst time for the next process, tn is the actual burst time of the current process, and tau(n) is the previous predicted burst time. He writes on the whiteboard to illustrate the concept, listing 'M1' with values '250' and '300'. He calculates a resulting value of '275' to show how the average works. He then discusses the parameter 'alpha' (alpha), writing 'alpha = 0', 'alpha = 1', and 'alpha = 0.5' to demonstrate how different weights affect the prediction, explaining that alpha determines the influence of the most recent history versus older history.
5:00 – 6:40 05:00-06:40
A table appears on the screen with columns labeled 'Process', 't', and 'tau'. The rows list processes P1, P2, P3, and P4. For P1, the actual time 't' is 10 and the initial 'tau' is 20. The instructor proceeds to calculate the 'tau' values for subsequent processes based on the formula. He writes '15' in the tau column for P2 and '13.5' for P3, showing the iterative nature of the prediction. The video concludes with a slide stating, 'This idea is also more of theoretical importance as most of the time the burst requirement of the coming process may vary by a large extent and if the burst time requirement of all the process is approximately same then there is no advantage of using this scheme.' This highlights the practical limitations of the exponential averaging technique in real-world scenarios.
The video provides a comprehensive overview of how to handle the unpredictability of CPU burst times in the Shortest Job First scheduling algorithm. It moves from identifying the problem to introducing the exponential averaging formula, explaining the role of the alpha parameter, and demonstrating calculations through both whiteboard examples and a process table. The lesson concludes by contextualizing the technique as theoretically important but practically limited when process burst times are inconsistent or uniform.