The problem of hardware-software partitioning in the design ofembedded systems is addressed. Uncertainties about the performance ofthe options for realization are expressed in triangular possibilisticnumbers. To handle such numbers an integer programming formulation ofthe partitioning problem is derived. This formulation can be convertedinto a possibilistic program without changing the asymptoticcomputational complexity. The approach is illustrated with resultsobtained with the receiver part of a transceiver of a wireless indoorspread spectrum system. This example and several other experiments haveshown that these optimizations can reach solutions within seconds fordesigns of that complexity and above
展开▼