9 Requirements Engineering II

R. Baskaran

 

When it comes to any type of project, requirement collection plays a key role. Requirements collection is not only important for the project, but it is also important for the project management function.

 

For the project, understanding what the project will eventually deliver is critical for its success. Through requirements, the project management can determine the end deliveries of the project and how the end deliveries should address client’s specific requirements.

 

Although requirements collection looks q uite straightforward, surprisingly, this is one of the project phases where most of the projects start with the wrong foot. In general, majority of the failed projects have failed due to the wrong or insufficient requirements gathering. We will discuss on this in the following section.

 

Following is an illustration indicating where the requirements collection comes in a project:

 

The Importance of Requirements

 

Let’s take a software development project as an example. O nce the project initiation is over, the business analyst team is in a hurry to collect requirements. The BA (business analysts) team use various methods to capture project requirements and then pass the requirements to the project team. Once business requirements are converted into technical requirements, the implementation starts.

 

Although the above cycle looks quite normal and trouble-free, the reality is somewhat different. In most of the cases, the BA team is unable to capture all the requirements related to the project. They always overlook a portion of requirements. During the construction of the project, usually the client recognizes the requirements gaps of the project.

 

The project team will have to implement these missing requirements with no additional client payments or with client approved change requests. In case if it was BA team’s fault, the service provider may have to absorb the cost for implementing the missing requirements. In such instances, if the effort for missing requirements has a significant impact on the cost of the project, the project may be a financial loss for the service provider.

 

Therefore, the requirement collection process is the most important phase of any project. The Process for Requirements Collection:

 

For the purpose of requirements collection, there are a few methods used by the business analysts.

 

These methods usually differ from one project to another and one client organization to another.

 

Usually, requirements for a new system are gathered from the potential end-users of the system. The methods used for gathering requirements from these potential end- users vary depending on the nature of the end-users. As an example, if there is a large number of end-users, then the workshop method can be used for requirements collection.

 

In this method, all the potential end- users are asked to participate for a workshop. In this workshop, the business analysts do engage with the users and collect the requirements for the new system. Sometimes, the workshop session is video recorded in order to review and capture any user feedback.

 

If the user base is quite small in number, the business analysts can carry out face-to-face interviews. This is the most effective way of finding all the necessary requirements as the business analyst can have all their questions asked and cross questioned as well.

 

Questioners can be used effectively for requirements collection process, but this should not be the only method of interacting with the end-users. Q uestioners should be used as a supporting feature for interviews or a workshop.

 

In addition to the above methods, there are many other specific methods that can be used for specific conditions.

 

Tips for Successful Require ments Collection:

 

Following are some of the tips for making the requirements collection process successful:

 

1. Never assume that you know customer’s requirements. What you usually think could be quite different to what the customer wants. Therefore, always verify with the customer when you have an assumption or a doubt.

 

2. Get the end-users involved from the start. Get their support for what you do.

 

3. At the initial levels, define the scope and get customer’s agreement. This helps you to successfully focus on scope of features.

 

4. When you are in the process of collecting the requirements, make sure that the requirements are realistic, specific and measurable.

 

5. Focus on making the requirements document crystal clear. Requirement document is the only way to get the client and the service provider to an agreement. Therefore, there should not be any gray area in this document. If there are gra y areas, consider this would lead to potential business issues.

 

6. Do not talk about the solution or the technology to the client until all the requirements are gathered. You are not in a position to promise or indicate anything to the client until you are clear about the requirements.

 

7. Before moving into any other project phases, get the requirements document signed off by the client.

 

8. If necessary, create a prototype to visually illustrate the requirements.

 

A high- level requirements specification is required. The purpose of the requirements analysis is to identify requirements for the proposed system. The emphasis is on the discovery of user requirements. Each requirement (or problem) must be defined and documented in the requirements catalogue. Each requirement is recorded in the requirements catalogue on a requirements catalogue entry form. A copy of the form is in the appendix section of the standards manual. The form should be completed as follows:

  • Project/System – the proposed system name or an abbreviation of it.
  • Analyst – your name as the analyst on this project.
  • Date – the date the entry was made in the catalogue, whether it is a new entry or an amended entry, i.e. another version of an existing entry.
  • Version – a version number is assigned to a requirement. The initial version of a requirement is number one. However, the requirement may need to be updated in the course of the development of the project, so then the requirements catalogue entry will be replaced with the updated version of the requirement and the updated version number will reflect this change.
  • Status – the status of the requirement will be either ongoing or complete. When the status is ongoing the status box will be empty, and when the status is complete the status box will contain a tick (().
  • Page – page numbering will be maintained within the catalogue.
  • Source/Origin – is the originator of the requirement; the person with the responsibility for negotiation about the requirement.
  • Requirement Number – a unique requirement number is assigned to each requirement. The requirement number is an incremental number starting with one.
  • Priority – a priority is assigned to the requirement. The priority given is agreed between the originator and the analyst. There are three priority levels: high (H), medium (M) and low (L). High priority is assigned when the requirement is mandatory. Medium priority is assigned when the requirement is desirable. Low priority is assigned when the requirement is optional.
  • Functional Requirements and Non-Functional Requirements – see next page.
  • Related Documents – reference to any related documents, eg user documentation, data flow diagrams.
  • Proposed Solution – any possible solution or general comments.
  • Actual Solution – records how the requirement is resolved, eg implemented or abandoned. This section will be completed at a later stage when the requirement has reached the completed status.

 

Entity Relations hip Diagrams (ER)

 

Now that we have the DFDs, we have the system view of the main functions and processes in the new system. Another important stage in the development is creating Entity Relationship diagrams, which will help us understand the data model view of the system.

 

Data modelling is a technique that is widely used in the world of business and information technology to show how information is, or should be, stored and used within a system. The success of any organisation relies on the efficient flow and processing of information.

 

Let’s look at an example in which information flows around the various departments in the organisation. This information can take many forms; for example it could be written, oral or electronic.

The General Manager regularly communicates with staff in the sales and marketing and accounts departments by using email. Orders received by sales and marketing are forwarded to the production and accounts departments for completion and invoicing. The accounts department forwards regular written reports to the General Manager. It also raises invoices and sends these to the customers.

 

Data modelling is a technique aimed at making the most of the way that information is stored and used in an organisation. It begins with the identification of the main data groups, for example the invoice, and continues by defining the detailed content of each of these groups. This results in structured definitions for all of the information that is stored and used within a system.

 

The technique provides a solid foundation for systems design and a universal standard for system documentation. Data modelling is an essential precursor to analysis and design, maintenance and documentation, and improving the performance of an existing system.

 

1.  Introduction

 

The following subsections of the Software Requirements Specifications (SRS) document should provide an overview of the entire SRS.   The thing to keep in mind as you write this document is that you are telling what the system must do – so that designers can ultimately build it.  Do not use this document for design!!!

 

1.1  Purpose

 

Identify the purpose of this SRS and its intended audience. In this subsection, describe the purpose of the particular SRS and specify the intended audience for the SRS.

 

1.2  Scope

 

In this subsection:

 

(1)  Identify the software product(s) to be produced by name

(2) Explain what the software product(s) will, and, if necessary, will not do

(3) Describe the application of the software being specified, including relevant benefits, objectives, and goals

(4) Be consistent with similar statements in higher- level specifications if they exist

 

This should be an executive-level summary.  Do not enumerate the whole requirements list here.

 

1.3 Definitions, Acronyms, and Abbreviations.

 

Provide the definitions of all terms, acronyms, and abbreviations required to properly interpret the SRS. This information may be provided by reference to one or more appendices in the SRS or by reference  to documents. This information may be provided by reference to an Appendix.

 

1.4 References

 

In this subsection:

 

(1)  Provide a complete list of all documents referenced elsewhere in the SRS

(2) Identify each document by title, report number (if applicable), date, and publishing organization

(3) Specify the sources from which the references can be obtained.

 

This information can be provided by reference to an appendix or to anot her document. If your application uses specific protocols or RFC’s, then reference them here so designers know where to find them.

 

1.5  Overview

 

In this subsection:

 

(1)  Describe what the rest of the SRS contains

(2) Explain how the SRS is organized

 

Don’t rehash the table of contents here.  Point people to the parts of the document they are most concerned with.  Customers/potential users care about section 2, developers care about section 3.

 

2. The Overall Description 

 

Describe the general factors that affect the product and its requirements. This section does not state specific requirements.  Instead, it provides a background for those requirements, which are defined in section 3, and makes them easier to understand. In a sense, this section tells the requirements in plain English for the consumption of the customer. Section3 will contain a specification written for the developers.

 

2.1  Product Perspective 

 

Put the product into perspective with other related products.  If the product is independent a nd totally self-contained, it should be so stated here.  If the SRS defines a product that is a component of a larger system, as frequently occurs, then this subsection relates the requirements of the larger system to functionality of the software and identifies interfaces between that system and the software.  If you are building a real system,compare its similarity and differences to other systems in the marketplace.  If you are doing a research-oriented project, what related research compares to the system you are planning to build.

 

A block diagram showing the major components of the larger system, interconnections, and external interfaces can be helpful.  This is not a design or architecture picture.  It is more to provide context, especially if your system will interact with external actors.  The system you are building should be shown as a black box.  Let the design document present the internals.

 

The following subsections describe how the software operates inside various constraints.

 

2.1.1  System Interfaces

 

List each system interface and identify the functionality of the software to accomplish the system requirement and the interface description to match the system. These are external systems that you have to interact with.  For instance, if you are building a business application that interfaces with the existing employee payroll system, what is the API to that system that designer’s will need to use?

 

2.1.2  Interfaces

 

Specify:

 

(1) The logical characteristics of each interface between the software product and its users.

(2) All the aspects of optimizing the interface with the person who must use the system

 

This is a description of how the system will interact with its users.  Is there a GUI, a command line or some other type of interface? Are there special interface requirements? If you are designing for the general student population for instance, what is the impact of ADA (American with Disabilities Act) on your interface?

 

Specify the logical characteristics of each interface between the software product and the hardware components of the system. This includes configuration characteristics.  It also covers such matters as what devices are to be supported, how they are to be supported and protocols.  This is not a description of hardware requirements in the sense that “This program must run on a Mac with 64M of RAM”. This section is for detailing the actual hardware devices your application will interact with and control. For instance, if you are controlling X10 type home de vices, what is the interface to those devices? Designers should be able to look at this and know what hardware they need to worry about in the design.  Many business type applications will have no hardware interfaces.  If none, just state “The system has no hardware interface requirements” If you just delete sections that are not applicable, then readers do not know if:  a. this does not apply or b.  you forgot to include the section in the first place.

 

2.1.4  Software Interfaces

 

Specify the use of other required software products and interfaces with other application systems. For each required software product, include:

 

(1) Name

(2) Mnemonic

(3) Specification number

(4) Version number

(5) Source

 

For each interface, provide:

(1)  Discussion of the purpose of the interfacing so ftware as related to this software product

(2)  Definition of the interface in terms of message content and format

 

Here we document the APIs, versions of software that we do not have to write, but that our system has to use.  For instance if your customer uses SQ L Server 7 and you are required to use that, then you need to specify i.e.

 

2.1.4.1 Microsoft SQ L Server 7. The system must use SQ L Server as its database component.

 

Communication with the DB is through ODBC connections. The system must provide SQ L data table definitions to be provided to the company DBA for setup.

 

A key point to remember is that you do NOT want to specify software here that you think would be good to use.  This is only for customer-specified systems that you have to interact with.  Choosing SQL Server 7 as a DB without a customer requirement is a Design choice, not a requirement. This is a subtle but important point to writing good requirements and not over-constraining the design.

 

2.1.5  Communications Interfaces

 

Specify the various interfaces to communications such as local network protocols, etc. These are protocols you will need to directly interact with. If you happen to use web services transparently to your application then do not list it here.  If you are using a custom protocol to communicate between systems, then document that protocol here so designers know what to design.  If it is a standard

 

2.1.6  Memory Constraints

 

Specify any applicable characteristics and limits on primary and secondary memory. Don’t just make up something here.  If all the customer’s machines have only 128K of RAM, then your target design has got to come in under 128K so there is an actual requirement.  You could also cite market research here for shrink-wrap type applications “Focus groups have determined that our target market has between 256-512M of RAM, therefore the design footprint should not exceed 256M.” If there are no memory constraints, so state.

 

2.1.7  Operations

 

Specify the normal and special operations required by the user such as:

 

(1)   The various modes of operations in the user organization

(2)   Periods of interactive operations and periods of unattended operations

(3)   Data processing support functions

(4)   Backup and recovery operations

 

(Note: This is sometimes specified as part of the User Interfaces section.) If you separate this from the UI stuff earlier, then cover business process type stuff that would impact the design. For instance, if the company brings all their systems down at midnight for data backup that might impact the design. These are all the work tasks that impact the design of an application, but which might not be located in software.

 

2.1.8  Site Adaptation Requirements

 

In this section:

 

(1) Define the requirements for any data or initialization sequences that are specific to a given site, mission, or operational mode

(2) Specify the site or mission-related features that should be modified to adapt the software to a particular installation

 

If any modifications to the customer ’s work area would be required by your system, then document that here.  For instance, “A 100Kw backup generator and 10000 BTU air conditioning system must be installed at the user site prior to software installation”.

 

This could also be software-specific like, “New data tables created for this system must be installed on the company’s existing DB server and populated prior to system activation.” Any equipment the customer would need to buy or any software setup that needs to be done so that your system wil l install and operate correctly should be documented here.

 

2.2  Product Functions 

 

Provide a summary of the major functions that the software will perform. Sometimes the function summary that is necessary for this part can be taken directly from the sect ion of the higher-level specification (if one exists) that allocates particular functions to the software product.

 

For clarity:

 

(1) The functions should be organized in a way that makes the list of functions understandable to the customer or to anyone else reading the document for the first time.

(2) Textual or graphic methods can be used to show the different functions and their relationships. Such a diagram is not intended to show a design of a product but simply shows the logical relationships among variables.

 

AH, Finally the real meat of section 2. This describes the functionality of the system in the language of the customer. What specifically does the system that will be designed have to do? Drawings are good, but remember this is a description of what the system needs to do, not how you are going to build it. (That comes in the design document).

 

2.3  User Characteristics 

 

Describe those general characteristics of the intended users of the product including educational level, experience, and technical expertise. Do not state specific requirements but rather provide the reasons why certain specific requirements are later specified in section 3.

 

What is it about your potential user base that will impact the design?  Their experience and comfort with technology will drive UI design.  Other characteristics might actually influence internal design of the system.

 

2.4  Constraints 

 

Provide a general description of any other items that will limit the developer’s options.  These can include:

 

(1)    Regulatory policies

(2)    Hardware limitations (for example, signal timing requirements)

(3)    Interface to other applications

(4)    Parallel operation

(5)    Audit functions

(6)    Control functions

(7)    Higher-order language requirements

(8)    Signal handshake protocols (for example, XON-XOFF, ACK-NACK)

(9)    Reliability requirements

(10)   Criticality of the application

(11)    Safety and security considerations

 

This section captures non- functional requirements in the customers language. A more formal presentation of these will occur in section 3.

 

2.5  Assumptions and Dependencies 

 

List each of the factors that affect the requirements stated in the SRS.  These factors are not design constraints on the software but are, rather, any changes to them that can affect the requirements in the SRS.  For example, an assumption might be that a specific operating system would be available on the hardware designated for the software product.  If, in fact, the operating system were not available, the SRS would then have to change accordingly.

 

This section is catch-all for everything else that might influence the design of the system and that did not fit in any of the categories above.

 

2.6  Apportioning of Requirements. 

 

Identify requirements that may be delayed until future versions of the system. After you look at the project plan and hours available, you may realize that you just cannot get everything done. This section divides the requirements into different sections for development and delivery.  Remember to check with the customer – they should prioritize the requirements and decide what does and does not get done.  This can also be useful if you are using an iterative life cycle model to specify which requirements will map to which interation.

 

3. Specific Requirements 

 

This section contains all the software requirements at a level of detail sufficient to enable designers to design a system to satisfy those requirements, and testers to test that the system satisfies those requirements.  Throughout this section, every stated requirement should be externally perceivable by users, operators, or other external systems.  These requirements should include at a minimum a description of every input (stimulus) into the system, every output (response) from the system and all functions performed by the system in response to an input or in support of an output. The following principles apply:

 

(1) Specific requirements should be stated with all the characteristics of a good SRS

· correct

· unambiguous

· complete

· consistent

· ranked for importance and/or stability

· verifiable

· modifiable

· traceable

(2) Specific requirements should be cross-referenced to earlier documents that relate

(3) All requirements should be uniquely identifiable (usually via numbering like 3.1.2.3)

(4) Careful attention should be given to organizing the requirements to maximize readability (Several alternative organizations are given at end of document)

 

Before examining specific ways of organizing the requirements it is helpful to understand the various items that comprise requirements as described in the following subclasses.  This section reiterates section 2, but is for developers not the customer.  The customer buys in with section 2, the designers use section 3 to design and build the actual application.

Remember this is not design.  Do not require spec ific software packages, etc unless the customer specifically requires them. Avoid over-constraining your design.  Use proper terminology:

 

The system shall… A required, must have feature

The system should… A desired feature, but may be deferred til later

The system may…  An optional, nice-to-have feature that may never make it to implementation.

 

Web Links

  1. https://en.wikipedia.org/wiki/Requirements_engineering
  2. http://www.tutorialspoint.com/software_engineering/software_requirements.htm
  3. http://www.cs.toronto.edu/~sme/papers/2004/FoRE-chapter01-v7.pdf
  4. http://courses.cs.washington.edu/corses/cse403/06sp/assgn/requirements_spec.doc

 

Supporting & Reference Materials

  1. Roger S. Pressman, “Software Engineering: A Practitioner’s Approach”, Fifth Edition, McGraw Hill, 2001.
  2. Pankaj Jalote, “An Integrated Approach to Software Engineering”, Second Edition, Springer Verlag, 1997.
  3. Ian Sommerville, “Software Engineering”, Sixth Edition, Addison Wesley, 2000.