Ananda Basu

Modeling Heterogeneous Real-time Components in BIP

We present a methodology for modeling heterogeneous real-time components. Components are obtained as the superposition of three layers: behavior, specified as a set of transitions; Interactions between transitions of the behavior; Priorities, used to choose amongst possible interactions. A parameterized binary composition operator is used to compose components layer by layer. We present the BIP language for the description and composition of layered components as well as associated tools for executing and analyzing components on a dedicated platform. The language provides a powerful mechanism for structuring interactions involving rendezvous and broadcast. We show that synchronous and timed systems are particular classes of components. Finally, we provide examples and compare the BIP framework to existing ones for heterogeneous component-based modeling

Rigorous Component-Based System Design Using the BIP Framework

An autonomous robot case study illustrates the use of the behavior, interaction, priority (BIP) component framework as a unifying semantic model to ensure correctness of essential system design properties.

Distributed Semantics and Implementation for Systems with Interaction and Priority

The paper studies a distributed implementation method for the BIP (Behavior, Interaction, Priority) component framework for modeling heterogeneous systems. BIP offers two powerful mechanisms for describing composition of components by combining interactions and priorities. A system model is layered. The lowest layer contains atomic components; the second layer, describes possible interactions between atomic components; the third layer includes priorities between the interactions. The current implementation of BIP is based on global state operational semantics. An Engine directly interprets the operational semantics rules and computes the possible interactions between atomic components from global states. The implementation method is a translation from BIP models into distributed models involving two steps. The first translates BIP models into partial state models where are known only the states of the components which are ready to communicate. The second implements interactions in the partial state model by using message passing primitives. The main results of the paper are conditions for which the three models are observationally equivalent. We show that in general, the translation from global state to partial state models does not preserve observational equivalence. Preservation can be achieved by strengthening the premises of the operational semantics rules by an oracle. This is a predicate depending on the priorities of the BIP model. We show that there are many possible choices for oracles. Maximal parallelism is achieved for dynamic oracles allowing interaction as soon as possible. Nonetheless, these oracles may entail considerable computational overhead. We study performance trade-offs for different types of oracles. Finally, we provide experimental results illustrating the application of the theory on a prototype implementation.

Using BIP for Modeling and Verification of Networked Systems -- A Case Study on TinyOS-based Networks

We apply a model construction methodology to TinyOS- based networks, using the behavior-interaction-priority (BIP) component framework. The methodology consists in building the model of a node as the composition of a model extracted from a nesC program describing the application, and models of TinyOS components. Models for networks are obtained by composition of models for nodes by using BIP connectors implementing different types of radio chan- nels. This opens the way for enhanced analysis and early error detection by using verification techniques.

Statistical abstraction and model-checking of large heterogeneous systems

We propose a new simulation-based technique for verifying applications running within a large heterogeneous system. Our technique starts by performing simulations of the system in order to learn the context in which the application is used. Then, it creates a stochastic abstraction for the application, which takes the context information into account. This smaller model can be verified using efficient techniques such as statistical model checking. We have applied our technique to an industrial case study: the cabin communication system of an airplane. We use the BIP toolset to model and simulate the system. We have conducted experiments to verify the clock synchronization protocol i.e., the application used to synchronize the clocks of all computing devices within the system.

Priority Scheduling of Distributed Systems Based on Model Checking

Priorities are used to control the execution of systems to meet given requirements for optimal use of resources, e.g., by using scheduling policies. For distributed systems, it is hard to find efficient implementations for priorities; because they express constraints on global states, their implementation may incur considerable overhead. Our method is based on performing model checking for knowledge properties. It allows identifying where the local information of a process is sufficient to schedule the execution of a high priority transition. As a result of the model checking, the program is transformed to react upon the knowledge it has at each point. The transformed version has no priorities, and uses the gathered information and its knowledge to limit the enabledness of transitions so that it matches or approximates the original specification of priorities.

Priority scheduling of distributed systems based on model checking

Priorities are used to control the execution of systems to meet given requirements for optimal use of resources, e.g., by using scheduling policies. For distributed systems it is hard to find efficient implementations for priorities; because they express constraints on global states, their implementation may incur considerable overhead.Our method is based on performing model checking for knowledge properties. It allows identifying where the local information of a process is sufficient to schedule the execution of a high priority transition. As a result of the model checking, the program is transformed to react upon the knowledge it has at each point. The transformed version has no priorities, and uses the gathered information and its knowledge to limit the enabledness of transitions so that it matches or approximates the original specification of priorities.

Rigorous system level modeling and analysis of mixed HW/SW systems

A grand challenge in complex embedded systems design is developing methods and tools for modeling and analyzing the behavior of an application software running on multicore or distributed platforms. We propose a rigorous method and a tool chain that allows to obtain a faithful model representing the behavior of a mixed hardware/software system from a model of its application software and a model of its underlying hardware architecture. The system model can be simulated and analyzed for validation of both functional and extra-functional properties. The tool chain uses DOL (Distributed Operation Layer [1]) as the frontend for specifying the application software and hardware architecture, and BIP (Behavior Interaction Priority [2]) as the modeling and analysis framework. It is illustrated through the construction of system models of MJPEG and MPEG2 decoder applications running on MPARM, a multicore architecture.

Verification of an AFDX infrastructure using simulations and probabilities

Until recently, there was not a strong need for networking inside aircrafts. Indeed, the communications were mainly cabled and handled by Ethernet protocols. The evolution of avionics embedded systems and the number of integrated functions in civilian aircrafts has changed the situation. Indeed, those functionalities implies a huge increase in the quantity of data exchanged and thus in the number of connections between functions. Among the available mechanisms provided to handle this new complexity, one find Avionics Full Duplex Switched Ethernet (AFDX), a protocol that allows to simulate a point-to-point network between a source and one or more destinations. The core idea in AFDX is the one of Virtual Links (VL) that are used to simulate point-to-point communication between devices. One of the main challenge is to show that the total delivery time for packets on VL is bounded by some predefined value. This is a difficult problem that also requires to provide a formal, but quite evolutive, model of the AFDX network. In this paper, we propose to use a component-based design methodology to describe the behavior of the model. We then propose a stochastic abstraction that allows not only to simplify the complexity of the verification process but also to provide quantitative information on the protocol.

Component Assemblies in the Context of Manycore

We present a component-based software design flow for building parallel applications running on top of manycore platforms. The flow is based on the BIP - Behaviour, Interaction, Priority - component framework and its associated toolbox. It provides full support for modeling of application software, validation of its functional correctness, modeling and performance analysis on system-level models, code generation and deployment on target manycore platforms. The paper details some of the steps of the design flow. The design flow is illustrated through the modeling and deployment of two applications, the Cholesky factorization and the MJPEG decoding on MPARM, an ARM-based manycore platform. We emphasize the merits of the design flow, notably fast performance analysis as well as code generation and efficient deployment on manycore platforms.