Research On Key Technologies Of Dynamic Hardware/software Partitioning

With the continuous development of the embedded system in people's life,the complexity of the system is further improved and the research of Hardware/software partitioning method is increasingly concerned by many researchers and experts.Hardware/software partitioning plays a very important role in the design of Hardware/software co-design.In the research of Hardware/software partitioning techniques,the incremental learning algorithm based on population has the advantages of simple concept,few parameters and fast convergence rate which is a feasible research method to solve the problem of Hardware/software partitioning.The PBIL algorithm is used as the main tool to study the problem of Hardware/software partitioning and the application of PBIL algorithm in Hardware/software partitioning is discussed,which includes the following aspects:First of all,the paper introduces the basic knowledge of Hardware/software partitioning,including the basic concepts,methods,target architecture,objectives,system cost,optimization objectives,constraints and so on.And in the process of comparison,select the most appropriate model and parameter definition,and provide a scientific test platform for the analysis and research of the performance of the Hardware/software partitioning algorithm.Secondly,the classification and main characteristics of dynamic environment are introduced and a new dynamic environment generation method is proposed.This method can simulate various changes of dynamic environment.The detection method of dynamic environment change is systematically arranged and a nested detection method is proposed to solve the low accuracy of the dynamic environment detection method.The experimental results prove the accuracy of the nested environmental detection method.Moreover,in the light of the efficiency of the standard PBIL algorithm,the learning factor and domain search strategy is improved.A new Hardware/software partitioning method based on improved PBIL algorithm is proposed.Then,the improved algorithm is compared with the other six kinds of artificial intelligence optimization algorithm.The experimental results show the effectiveness of the proposed algorithm.At last,in the light of dynamic Hardware/software partitioning,a new Hardware/software partitioning method based on improved PBIL algorithm is proposed.Adopt the method of generating the dynamic environment the paper presents in the third chapter as the test problem.The improved PBIL algorithm and the original PBIL algorithm are compared with the experimental results.The experimental results show that the improved PBIL algorithm has better performance and stability.