SDLC and Development Approaches in System Analysis and Design: Models, Selection Rules, and a 16-Week Worked Example

Learn how analysis, design, traceability, and major SDLC models fit together. Then use project conditions and risk exposure to plan a 16-week examination portal.

KnowledgeGate Team

Exam prep & CS education

Updated 6 Nov 20265 min read
SDLC and Development Approaches in System Analysis and Design: Models, Selection Rules, and a 16-Week Worked Example

A memorised SDLC phase list does not show why Waterfall, V-model, prototyping, Spiral, Agile, and DevOps differ, or how to choose. The approaches differ in how they handle sequence, feedback, risk, and release, as shown by the college examination portal example.

SDLC in system analysis and design: lifecycle, model, and methodology

The systems development life cycle governs an information system from problem recognition to retirement. A process model orders activities; a methodology supplies practices, roles, and artefacts. The lifecycle tells us what work exists; the model tells us when; the methodology tells us how.

Textbooks may merge or rename these eight activities:

  1. Preliminary investigation and planning

  2. Feasibility study

  3. Requirements analysis

  4. Logical and physical design

  5. Implementation

  6. Verification and validation

  7. Deployment or conversion

  8. Operation and maintenance

Feedback and change control make this a lifecycle, not a checklist. Place it in the wider syllabus through GATE CS Exam Preparation.

System analysis and system design: questions, artefacts, and traceability

System analysis asks what problem must be solved and why. It covers scope, stakeholders, functional and non-functional requirements, constraints, and TELOS feasibility: technical, economic, legal, operational, and schedule. Outputs include a scope statement, context diagram or DFDs, use cases, data dictionary, feasibility note, prioritised requirements, and SRS.

System design asks how approved requirements will be realised. Logical design covers processes, data flows, entities, and controls. Physical design chooses architecture, database schemas, interfaces, UI, deployment, security, backup, and recovery. Implementation, testing, conversion, and maintenance follow.

Traceability connects them. R17, "A candidate can download the hall ticket only after payment is confirmed," maps to UC-07, PaymentService, HallTicketService, and acceptance test T-42. A later admin-fee-waiver change must trigger an impact check across every record, not an isolated code edit.

SDLC development approaches: predictive, iterative, incremental, and adaptive

Model or approach

Core flow

Strong fit

Main limitation

Waterfall

Sequential phase gates

Stable requirements

Late change is costly

V-model

Stages paired with test levels

Early test planning

Rigid under change

Iterative

Revisit and improve

Repeated learning

Full scope arrives later

Incremental

Release usable slices

Independent feature value

Integration needs planning

Prototyping

Build to clarify uncertainty

Unclear interaction

Rough code may be misused

Spiral

Objective, risk, build, evaluate, plan loops

High technical risk

Needs risk-analysis skill

RAD

Time-boxed component construction

Modular scope and available users

Weak for tight coupling

Agile

Short adaptive iterations

Frequent reprioritisation

Needs active feedback

DevOps

Automate delivery and operations feedback

Repeatable releases

Does not replace the lifecycle

An iteration improves a slice; an increment adds usable scope. A throwaway prototype answers a question and is discarded; an evolutionary prototype is hardened. Each Spiral loop sets objectives, analyses risk, develops, evaluates, and plans.

In the V-model, verification checks a work product against its predecessor, such as design against the SRS. Validation checks the running system against user needs, such as T-42 against UC-07.

Choosing an SDLC model from project conditions

Project condition

Initial choice

Stable, regulated requirements

Predictive phase-gated model

Early test planning

V-model

Unclear user interaction

Prototyping

High technical or integration risk

Spiral

Independently valuable functions

Incremental

Frequent reprioritisation with user feedback

Agile

Repeatable deployment and production feedback

DevOps practices around the chosen process

Also assess stability, late-change cost, early-software need, user availability, compliance, integration, release frequency, and team maturity. In a hybrid, each practice must solve a named condition. Link conditions to model features and acknowledge limitations.

SDLC worked example: a 16-week college examination portal

A six-person team has 16 weeks and 48 requirements: 30 stable, 12 likely to change after staff feedback, and 6 technically high-risk. The non-stable set is 12 + 6 = 18, and 18 / 48 x 100 = 37.5%. One frozen specification and a late release fit poorly, but the stable core supports usable increments.

Risk exposure is probability x impact in person-weeks:

  1. Payment-gateway failure: 0.30 x 8 = 2.4 person-weeks.

  2. Remote-proctoring integration: 0.50 x 12 = 6.0 person-weeks.

  3. Reporting-UI misunderstanding: 0.20 x 5 = 1.0 person-week.

  4. Total initial exposure: 2.4 + 6.0 + 1.0 = 9.4 person-weeks.

Remote proctoring ranks first. Prototype evidence lowers its estimated failure probability from 0.50 to 0.15 and its estimated impact from 12 to 8 person-weeks. Its revised exposure is 0.15 x 8 = 1.2 person-weeks. The new total is 2.4 + 1.2 + 1.0 = 4.6; the reduction is 9.4 - 4.6 = 4.8 person-weeks. Expected risk guides a decision, it does not promise saved effort.

A Spiral-plus-incremental hybrid gives this calendar:

Weeks

Work

1-2

Planning, TELOS feasibility, scope, and requirements baseline

3-4

Proctoring risk prototype

5-8

Increment 1: registration

9-12

Increment 2: payment confirmation and hall-ticket generation

13-15

Increment 3: proctoring and reports

16

Acceptance, conversion, and deployment

Each increment feeds findings into the backlog. Any R17 change is traced through UC-07, both services, and T-42.

Sixteen-week portal lifecycle from planning through increments, acceptance, and deployment.
Portal risk exposure before and after the proctoring prototype, with totals and reduction.

SDLC traps: similar terms and wrong shortcuts

Wrong shortcut

Correct reasoning

Iterative means incremental

Iteration improves; an increment adds usable scope

Every prototype is a release

Throwaway prototypes are discarded; increments are usable

Spiral means repeat until done

Every loop has objectives and risk analysis

V-model is Waterfall

V-model pairs work products with test levels

Verification means acceptance

Verification checks artefacts; validation checks user need

Logical equals physical design

Logical models; physical chooses implementation

Maintenance only fixes faults

Corrective fixes, adaptive responds, perfective improves, preventive protects

Agile still plans and documents, but closer to changing work. DevOps retains analysis, design, and testing while shortening development and operations feedback.

Pure Waterfall ignores 18 non-stable requirements. Choosing Spiral because the project seems "large" ignores the 6.0 person-week exposure. Prototyping every module wastes effort on 30 stable requirements.

GATE-style questions and interviews on SDLC approaches

Questions ask you to arrange phases or artefacts, match scenarios to models, distinguish paired concepts, or rank risk exposure. Here, proctoring starts highest at 6.0 person-weeks, then falls below payment because 1.2 < 2.4.

In an interview, state volatility, rank risk, choose feedback cadence, name artefacts, then explain testing and change control. Here: prototype proctoring, deliver stable scope incrementally, and trace R17 through T-42.

For SDD structure, requirement-to-design traceability, test documentation, and QA metrics, continue with the live exam hub for related study. This post owns lifecycle sequencing and model selection. Use GATE CS Subject Weightage: Where Hours Pay Off for revision-priority context.

SDLC and development approaches: the short version

Use this five-step decision ladder:

  1. Define the problem and test TELOS feasibility.

  2. Baseline requirements and create traceability.

  3. Classify requirement stability, technical risk, and feedback needs.

  4. Select or combine models for named reasons.

  5. Validate each increment and control change through maintenance.

For this portal, a risk-first Spiral loop followed by usable increments beats one late release. For the full sequence, use GATE Guidance by Sanchit Sir. For practice, use the GATE Test Series. If you only need this concept, redraw both diagrams and justify the choice.

Go deeper with the full courseGATE Guidance by Sanchit Sir

Keep learning

Discussion