A New Software Complexity Metrics for Computer Communication Network

H. Singh, A. Mustapha, A.M. Dixit, L. Hua, and G.R. Gerhart (USA)


Software Metric, Software complexity, software reliability, Boolean algebra, Fuzzy Logic


A number of definitions for determining software complexity can be found in numerous papers. These definitions are applicable for large software. However, to the best of author’s knowledge, no definition of software complexity has been proposed for computer communication network. While software reliability has generally been defined from zero to one, the software complexity definition has no standard and has been defined by different authors in different ways. These software complexity definitions do not give an idea while comparing their software, especially when the range of these definitions are different (e.g. the number of line of code may be entirely different from the number of loops. Lines of code may be in the thousands when loops may be few in number). In this paper, software complexity is defined to have a range between zero and one. This premise is based on the notion that more complex software is less reliable and vice versa. The software reliability has a range of zero to one; hence software complexity is also defined as to have a range of zero to one. Further, these definitions are extended so that the complexity of interconnected network can be defined. Boolean Algebra approach for the reliability communication network is utilized. These definitions are especially important because of different software used in different parts of the internet. It is hoped that these definitions will become standard for the software engineers.

Important Links:

Go Back