Correctly modeling software requirements is one of the grand challenges of current ECU (Electronic control Unit) development. To ensure the correctness of the requirements, formal modeling techniques are usually used because they allow analyzers to simulate, verify and even conduct performance analysis in the requirement level. In this paper, we propose a requirements modeling framework, based on the philosophy of separation of concerns and the formal modeling techniques. The main contributions of this paper are two-fold: (1) We divide a complicated automotive software as several concerns, each of which is modeled by different formal techniques, thus the descriptive complexity of the requirements is decreased, and accordingly the models’ understandability is enhanced; (2) The adoption of formal techniques allows us to simulate the execution of the software and calculate the performance in the early stage of development, therefore the correctness of requirements can be improved.

Requirements modeling is a critical stage in the development of ECUs (Electronic Control Unit) of automotive systems. A typical automotive system usually involves large amount of requirement items, and possesses restricted timing constraints and complicated interactions, therefore a single modeling language only does not cover the whole characteristics of a system. In this paper, we extend SaveCCM, a component model for automotive software development, with timed automata to form a modeling framework for automotive requirements. The integrated model can be used to analyze the architecture and timed behavior of an automotive system. The proposed approach takes the following advantages that: (1) It can effectively and completely express the architecture and behavior requirements of automotive systems; (2) It can describe the interactions between an ECU and its environment in terms of synchronization channels.