Application of Uncertain Programming in Hardware/Software Partitioning: Model and Algorithm.
Si Chen,Lida Huang,Guoqi Xie,Renfa Li,Keqin Li
DOI: https://doi.org/10.1142/s0218126623501050
2022-01-01
Journal of Circuits Systems and Computers
Abstract:Hardware/software partitioning is a typical multi-stage decision optimization problem; most existing hardware/software partitioning methods ignore a fact that real-life decisions are usually made in an uncertain state. We should model the hardware/software partitioning problem in uncertain environments and deal with uncertainty. The state-of-the-art work proposed an uncertainty conversion method for hardware/software partitioning, but this method does not include the equivalent deterministic model and is not suitable for dealing with different types of uncertainties. In order to cope with different situations with various uncertainties, we should apply uncertain programming to build a model in uncertain environments and give different equivalent deterministic models to convert different uncertainties theoretically. In this paper, we present the process of applying uncertain programming to solve the hardware/software partitioning problem, including the model and algorithm. We convert the uncertain programming model into its equivalent deterministic models, including the expected value model and the chance-constrained programming model; we give details for the conversion methods of these two models. We present the custom genetic algorithm to solve the converted model, by incorporating a greedy idea in two steps of the genetic algorithm. Experimental results show that the custom genetic algorithm can find a high-quality approximate solution while running much faster for large input scales, compared with the exact algorithm.