Monday September 14th in the Large Conference Room Mohamad Jaber will present a seminar covering BIP and related research with an open discussion afterwards.

For the next two weeks we have Mohamad Jaber, one of Joseph Sifakis' Ph.D. students, working here at ISIS. On Monday (09/21) starting at 10:30am in the Large Conference Room, he will present a seminar covering BIP and related research with an open discussion afterwards. An abstract for his presentation and a brief CV are below.

There will be donuts and drinks at the talk. If you have any questions about the seminar please feel free to contact Graham at graham.hemingway[at]vanderbilt.edu.

Abstract :

The talk studies efficient code generation and 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. We propose two new implementations:

1- For efficient code generation, we propose a method that transforms the interactions of a component-based program in BIP and generates a functionally equivalent program. The method is based on the successive application of three types of source-to-source transformations:
flattening of components, flattening of connectors and composition of atomic components. By exhaustive application of the transformations, any BIP component can be transformed into an equivalent monolithic component. From this component, efficient C code can be generated.

2- For distributed implementation, we propose a method for generating distributed implementations from BIP models. The method is based on two source-to-source transformations preserving observationalequivalence: 1) A transformation from a global state to a partial state model which replaces strong synchronization primitives in atomic components by send/receive primitives; 2) A transformation which replaces multiparty interaction with protocols based on asynchronous message passing. This transformation may use additional protocols ensuring mutual exclusion between conflicting interactions. We propose different distributed implementations from fully decentralized to fully centralized ones. We study their properties, in particular performance criteria such as degree of parallelism and overhead for coordination. We also provide experimental results including a tool for automatic generation of implementations and benchmarks for different case studies.

Brief CV:

I am currently third year PhD student at Verimag Laboratory under the supervision of Joseph Sifakis. I work on the development of theory, methods and tools for Source-to-Source Transformation in component- based systems decribed in BIP. We have recently developed BIP2BIP toolset which implements a method to flatten a system description specified with hierarchical components in the BIP language into a single monolithic component. This then allows more efficient code generation from the description by avoiding dynamic run-time calls to the BIP execution environment. Currently I'm working in the generation of distributed implemention from BIP models.