lundi 30 août 2010
Consumer electronic devices are made of hardware (the physical part) and software. In the past, the hardware part had to be fabricated first (e.g. prototype boards), then used for developing the software. Transaction-level modeling (TLM) is an attempt to avoid waiting for the hardware to physically exist before starting the development of software.
TLM models are abstractions of pieces of hardware. These models can be assembled together, but then several questions arise : if two models are independently "faithful" (w.r.t. the real hardware they represent), is the assembly still "faithful" ? This is one of the questions I am investigating.