Browse Prior Art Database

Schedule Based Resource Allocation Disclosure Number: IPCOM000235606D
Publication Date: 2014-Mar-11
Document File: 4 page(s) / 151K

Publishing Venue

The Prior Art Database


In examining the problem of underutilized computer resources at universities, a solution was derived that looked at scheduling resource allocation based on anticipated usage schedules calculated from information provided by the universities. By looking at variables including number of students in each class, computational difficulty per assignment, and the due dates for each assignment, a rough calculation can be made that allows for an accurate prediction of computational needs throughout a given term. This solution provides an advantage to the university in that they no longer pay for computer resources that aren't being utilized at any given time, and an advantage to the cloud service provider in that they can free up unutilized resources for use by other customers.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 4

Schedule Based Resource Allocation

A. Domain and Problem

At a University, computing resources are used across a large population of students and professors. Usage tendencies at a university vary from term to term based on course offerings, student enrollment, project due dates, and a number of other variables. This makes it difficult to estimate the computational resources the university needs from the cloud at any specific point in time.

Take a hypothetical university, for instance, "Watson State University." Say it only offers two courses in a given term, just to simplify the problem. Watson State offers a computer programming class, Computer Vision I, and a computer art class, Intro to 3D Animation. In the computer vision class, assume there is one large project due at the end of the term, and there are no other notable assignments due (notable in the sense that they require computer resources). In the 3D animation class, assume there are several smaller assignments due throughout the term.

If Watson State were to purchase block computer resources at the beginning of the term and keep the same throughout (See Fig. 1), the majority of the time, a lot of those resources would just sit idle and unused. However, by matching resources to the anticipated needs of WSU, they can minimize under utilized computer resources (See Fig. 2).

Figure 1

B. Example


Page 02 of 4

Figure 2

The goal of this invention is to give universities a way to predict their resource usage over the course of a term based on classes, class schedules, and enrollment in the classes. The university can then use this information to purchase the minimum computational resources the university needs from the cloud. This solution provides an important and useful alternative to previous solutions because it does not rely on past data for predictions, which is necessary for the ever-changing university environment. This solution also does not involve finding the best way to allocate a limited amount of resources, rather it determines the requisite resources based on estimated resource requirements.

The following is a general algorithm for determining a logical and predictive way to allocate computer resources for a university. The following inputs are given:


 Class Schedules A list of class schedules for every course offered at the school that will need computer resources. Assignment Dates The date the professor issues the assignment to students. Used to know when to start making resources

available for a given assignment.

Due Dates The date and time at which the assignment is due. Used to know when resources will no longer be

required for a given project, or when to greatly reduce the resource allotment for a given project (in the event of a late turn in date).

Late Turn In Date The lates...