A Practical Model for Dynamic Software Measurement

Usually, software measurement model is considered can not be used in practice; however, in this paper, we present a dynamic software measurement model, which performs very well in practice. In this model, the CPU utilization, memory occupied, threads and execution time are encapsulated into a multi-dimensional vector, viz. software state vector x; further, a state space method is incorporated to compute the software total complexity according to the relations between the state vector x and the program scale, difficulties, program structures, and development expenses. By such technique (feedback alike), the complexity of the developing software could be dynamically obtained and regulated, so that the software development life cycle and software reliability are affected by the regulated factors, recursively. A prototype has been implemented to demonstrate the feasibility of the proposed model through a funded software project. We also present the initial findings and results concerning the software project.