Inconsistency Checking of UML Sequence Diagrams and State Machines Using the Structure-Behavior Coalescence Method

Abstract

UML offers various models for system behavior. A key concern is efficiently determining that these different models are all consistent. Determining this requires a precise semantic understanding of each model. Two of the most-used UML models are Sequence Diagrams (SDs) and State Machine (or Statechart) Diagrams (SMDs). We propose a metamodel to describe both SDs and SMDs. Such a unified metamodel simplifies verifying behavioral equivalence between models, as each is represented with structural similarities. Our proposed metamodel is a label transition system (LTS), which is a form of finite state machine that is derived from process algebra.A metamodel based on process algebra has several benefits. First, it allows for a high-level plain-text description of the system as a set of process algebra equations. These algebraic equations can also be presented graphically as an informative overview diagram. Second, the use of process algebra simplifies the derivation of the semantic meanings of both the SD and SMD, by converting it into a problem of parsing the syntax of algebraic expressions, then applying transformations to derive an LTS from the parse tree. Because the operands of process algebra are processes, a direct and seamless correspondence exists between the high-level algebraic description and the low-level semantic LTS description. Third, as the SD and SMD are both described in process algebra, an LTS analyzer can simplify verification of behavioral equivalence through the structural similarity between the LTS descriptions of the SD and SMD.