No abstract

This paper presents our research in the domain of automated generation of a model from an SDL (Specification and Description Language) system specification. We use the Spin (Simple Promela Interpreter) formal verification tool and the Promela (Process Meta-Language) language for the description of the model. If the model is prepared manually, we need an expert with the detailed knowledge of the system, specification language, and modelling language. The quality of the model is directly influenced by the expert and is prone to the incorrect modelling of the system's properties due to the human error. Therefore, automatic generation of the model from the SDL specification is desired. In this paper we present our approach to the automated generation of the model in Promela. Additionally, we present challenges and future research directions.

No abstract

Abstract-This paper presents our research in the domain of mechanical extraction of a model from an SDL (Specification and Description Language) specification of a system. We use formal verification tool Spin (Simple Promela Interpreter) and Promela (Process Meta-Language) language for the description of the model. With the model checking technique the model's accordance with the system correctness requirements can be established with mathematical accuracy. The model can be generated manually or mechanically. If it is to be prepared manually, we will need an expert with the detailed knowledge of the system and both languages. The quality of the model is directly influenced by the expert. The process is prone to the incorrect modelling of the system's properties. In this paper we present the most critical parts of mechanical creation of the models in Promela. Additionally, we present challenges, research directions and some solutions for the automatic generation of models from an SDL specification.

SUMMARYInput/Output (I/O) automata and the Temporal Logic of Actions (TLA) are two well-known techniques for the specification and verification of concurrent systems. Over the past few years, they have been extended to the so-called dynamic I/O automata and, respectively, Mobile TLA (MTLA) in order to be more appropriate for mobile agent systems. Dynamic I/O automata is just a mathematical model, whereas MTLA is a logic with a formally defined language. In this paper, therefore, we investigate how MTLA could be used as a formal language for the specification of dynamic I/O automata. We do this by writing an MTLA specification of a travel agent system which has been specified semi-formally in the literature on that model. In this specification, we deal with always existing agents as well as with an initially unknown number of dynamically created agents, with mobile and non-mobile agents, with I/O-automata-style communication, and with the changing communication capabilities of mobile agents. We have previously written a TLA specification of this system. This paper shows that an MTLA specification of such a system can be more elegant and faithful to the dynamic I/O automata definition because the agent existence and location can be expressed directly by using agent and location names instead of special variables as in TLA. It also shows how the reuse of names for dynamically created and destroyed agents within the dynamic I/O automata framework can be specified in MTLA. key words: dynamic input/output automata, mobile agent system, formal specification, spatio-temporal logic