Learning Objectives

Cohesion refers to how closely related and focused the responsibilities of a single module are within a software system. In contrast, coupling describes the degree of interdependence between different modules. When a module has high cohesion, it performs a well-defined task effectively, leading to easier maintenance and understanding. Low coupling means that modules can operate independently of one another, reducing the impact changes in one module have on others.

Cohesion can be classified into several types: Functional Cohesion (high and ideal, where all elements contribute to a single task); Sequential Cohesion (high, where outputs of one part serve as inputs to another); Communicational Cohesion (high, where all elements operate on the same data); Procedural Cohesion (medium, where activities follow a sequence); Temporal Cohesion (low, where actions are related by time); Logical Cohesion (very low, where unrelated tasks are executed based on a parameter); and Coincidental Cohesion (worst, where tasks are grouped with no relationship). Understanding these types helps in designing stronger modules.

Coupling can also be categorized: Data Coupling (ideal, where modules share only necessary data); Stamp Coupling (low, where modules share data structures but only part is used); Control Coupling (medium, where one module controls the execution path of another); External Coupling (medium, where modules depend on external entities); Common Coupling (high, where modules share global data); and Content Coupling (worst, where one module depends directly on the internal data of another, violating encapsulation). Recognizing these types is crucial for managing module dependencies and reducing complexity.

The principles of high cohesion and low coupling are foundational to good software design. High cohesion within modules means they are easier to understand, maintain, and reuse since they focus on a single responsibility. Low coupling between modules increases the flexibility of the software system, making changes to one module less likely to impact others. Together, these principles lead to systems that are easier to modify and extend.

Structured Analysis and Structured Design (SASD) provides a methodological foundation for developing software systems. It organizes the development process into phases, including requirements gathering (Structured Analysis) and design (Structured Design), ensuring that both the functionality and structure of the software are well-defined. SASD emphasizes clarity, documentation, and a systematic approach to handle complexity.

Data Flow Diagrams (DFDs) are essential tools used in Structured Analysis that visually represent data movement within a system. They illustrate how data enters the system, how it is processed, stored, and subsequently given output. DFDs help stakeholders to understand system functionalities without delving into technical details and assist in identifying requirements and inefficiencies.

DFDs utilize four basic symbols: Processes (represented as circles or ovals), which show transformations of data; Data Flows (indicated by arrows) that represent data movement; Data Stores (depicted by parallel lines or open rectangles) for where data is held; and External Entities (shown as rectangles) that signify sources or destinations of data outside the system. Understanding these symbols is critical for accurately modeling data flows.

DFDs are organized in levels, starting with the Context Diagram that provides a high-level overview of the entire system. Level 1 DFDs break this down into major subprocesses. Each subsequent level (Level 2, Level N) offers a more detailed look into specific processes. Balancing DFDs ensures that the data flows into and out of a parent process match those in its child processes, maintaining consistency across levels.

Developing DFD models involves starting with a Context Diagram and progressively breaking down processes into more detailed Level 1, Level 2 DFDs, etc. Each level illustrates data flows and relationships among internal processes and data stores. This systematic development is crucial for accurately capturing the functioning of systems, managing complexity, and identifying data flow issues.