Scheduling Criteria of CPU

Scheduling Criteria of CPU for a scheduler varies from one scheduler to another. There are many scheduling algorithms. Different scheduling algorithms have different properties. The selection of a proper scheduling algorithm may improve the system performance. We must consider the properties of various scheduling algorithm and the computer system for selecting a particular scheduling algorithm.

Many criteria have been suggested for evaluating the scheduling algorithm. Some commonly used scheduling criteria are described below.

1.      CPU Utilization Scheduling Criteria:

The CPU must be busy as much as possible to perform different activities. The percentage of time, the CPU is executing a process may range from 0 to 100 percent. CPU utilization is very important in real time and multiprogramming system. In a real time system the CPI utilization should be 50 percent (lightly loaded system) to 95 percent (heavily loaded system). It means that load on a system affects the CPU utilization. The high CPU utilization is achieved on heavily loaded system.

2.      Balanced Utilization Scheduling Criteria:

Balanced utilization represents the percentage of the time al the resource utilized. In addition to considering he CPU utilization the utilization of memory, I/O devices and other system resources are also considered.

3.      Through Put Scheduling Criteria:

The number of process executed by the system in a specific period of time this time unit is called through put. For long process this rate may be one process per minute. Similarly for short process, it may be 100 processes per minute. The evaluation of through must be considered on the basis of average length.

4.      Turnaround  Time Scheduling Criteria:

Turnaround time represents the average period of time taken by a process executes. The turnaround time is computed by subtracting the time, when the process was created from the time is terminated. The turnaround time is inversely proportional to through put.

5.      Waiting Time Scheduling Criteria:

Waiting time represents the average period of time, a process spends waiting in the ready queue to get a chance for execution. It does not include the time, a prcess is executing on the CPU or performing I/O. waiting time is also very important factor to measure the performance of the system.

6.      Response Time Scheduling Criteria:

Response time represents the average time take by the system to start responding to user request. The response time is considered in interactive systems. For example, ATM is an interactive system, which is used in banks for withdrawal of money. The user expects that the system should response quickly. In interactive system the turnaround time is not a best criterion and this mostly depends on the speed of the users responses to the turnaround time in interactive system has no importance. Therefore the response time in an interactive system should be very less.

7.      Predictability Scheduling Criteria:

Predictability represents the consistency in the average response time in interactive system. It is another measure of performance of a system because users prefer consistency. Suppose an interactive system that normally responds within a microsecond, but on some occasions, it takes 5 to 15 milliseconds or more. In this case the user may be confused. Mostly the users prefer the system with reasonable and predicable response time, than a system that is faster but is highly variable is response time.

8.      Fairness Scheduling Criteria:

Fairness represents the degree to which all processes are given equal opportunity of execution. This criterion is codified in time shared system.

9.      Priorities Scheduling Criteria:

The process with higher priorities must be given preference for execution.