Search

The development methodology for today’s computer systems is lacking in the area of preliminary design. Decisions that greatly influence the product’s life-cycle cost are made during this stage of the design with minimal knowledge. One of the key decisions made during preliminary design is the allocation of the system’s functions to hardware, software, or firmware implementations. This research investigates methods to automate the decision making process during preliminary design. These decisions include the hardware, software, and firmware allocation process and the reduction of the preliminary design space by ordering the design space and determining a cut-off point. A general system model is presented based on the top-down design methodology. A decision process is applied to the lowest level functions of the system model to order the allocation design space. The methods explored to perform decisions are the general linear model and fuzzy logic membership functions. A computer based tool is developed that applies the decision methods to a system model. The output of the tool is a set of optimum hardware/software/firmware function allocations. The application of this tool greatly reduces the overall allocation design space which allows the human designer to make improved decisions during preliminary design.