Software Cost Estimation Models in Software Planning

In starting a new software project, it is important to know how much will it cost to develop and how much development time will it take to complete. These evaluations are needed before development is started and conveyed to the team. The software industry has inconsistently defined and explained metrics or atomic units of measure, data from real and actual projects are largely and highly suspect in terms of consistency and comparability. There are many questions as debates among developers and vendors of software cost estimation models and tools. The various estimation procedures that have been developed and have the following common characteristics are given below:

• The scope of the software project must be created before starting to develop the software.
• Metrics like FP or LOC are used for assessing the software.
• To achieve the targeted cost & schedule estimate, several things arise.
• Obtain one or more cost and effort of a project.

Below are the two models for estimating the cost of a software project:

In a static model, a single variable is grabbed as a key element for calculating the cost and effort whereas, In a dynamic model, all variables are connected, and there is no primary variable.

1. Static Single Variable model- The Methods using this model utilize an equation to get the desired values such as cost, time, effort, etc. And these all depend on the same variable used as a predictor like, size. Below is an example of the most common equation:

C = aL^b

Where C is cost, L is size and a, b are constants.

We have an example of the static single-variable mode. e. i.e. SEL model which is used for estimating software production. The equation of this model is given below:

E = 1.4L^0.93

               DOC = 30.4L^0.90

               D = 4.6L^0.26

Where E is in Person-months, DOC i.e, documentation is in the number of pages and is duration which is months.

2. Static Multivariable model- These models are also known as multivariable models. This model is often based on the first equation and depends on several variables representing different aspects of the software development environment.

Equations are: 

 E = 5.2L^0.91

 D = 4.1L^0.36

Where E is in Person-months, D is duration which is months.

More Cost Estimation Models:

1. 3-point estimation: It is a statistical technique that employs three estimates: optimistic, most likely, and pessimistic, for every task, and using these estimates, it determines how long each task should take.
2. Cost Estimation via Analogy: This method calculates the current project’s cost by using historical data from related projects.
3. Monte Carlo Methodology: This is a statistical technique that simulates various project outcomes by using random sampling and facilitates the estimation of the cost’s probability distribution.
4. Top-Down Estimation: In top-down estimation, the total project cost is calculated first, and then, according to a specified distribution, this total cost is distributed among different tasks or components.
5. Bottom-Up Estimation: Bottom-up estimation, as compared to top-down estimation, is calculating the cost of individual activities or components first, then combining the estimates to get the overall project cost.