Affiliations: Computer Science Department, Texas Tech University,
Lubbock, Texas, USA | Department of Computing, University of South Africa,
Pretoria, South Africa | Electrical and Computer Engineering Department,
University of Alabama at Birmingham, Birmingham, Alabama, USA | Department of Defense, USA
Abstract: Component-Based Software (CBS) engineering is envisioned to address
the issues related to the increasing size and complexity of software systems.
In CBS development, the designer designs systems by using readily available
(possibly third party) software components without needing the source code for
the components. Lack of source code, in general, renders the classical metrics
cumbersome to use, if not useless. Coupling and cohesion aspects of a
system/subsystem are the quality attributes that can seriously impact the
maintenance, evolution, and reuse. We present an information-theoretic approach
based on the notion of Shannon Languages for helping the system designer in the
assessment of coupling and cohesion early in the design phase. The proposed
methodology is most beneficial for CBS (where the source code is in general
absent) however it is applicable in other development methodologies in which
the source code for the software components is available.
