Research And Implementation Of The DSP+FPGA Platform On The Power Management

With the development of electronic technology,the demand for the performance of electronic products and reliability is growing, the problem of the electronic products has been more and more people concerned. However, the service life of electronic products is closely related to the integrated circuit temperature. If the temperature increases, the life of the electronic products will be relatively reduced.Reducing the power consumption of the integrated circuit can effectively control the temperature of the system, so as to prolong the life of electronic products. Power consumption, therefore, become a critical problem in the design of integrated circuit,the low power design has gradually become important research direction in the design of integrated circuit.In this paper, I analyzed the power consumption of the original DSP+FPGA platform, and integrated the characteristics that the platform would wait for a period of time after the completion of work, and the existing power optimization method, I adopt the method of the dynamic power management system level and integrate the power supply design, software optimization power optimization method to reduce the power consumption of the platform.This paper established the dynamic power management model, worked pattern division of different modules in the system, defined the idle mode and sleep mode and calculated the switching transition conditions of the working mode. In the dynamic power management strategy, this paper compared the timeout strategy and prediction strategy respectively, and compared the timeout algorithm, the average index prediction algorithm, improving index prediction algorithm, grey prediction algorithm, real-time power management algorithm, and a prediction algorithm dealing with an emergency. Extract the advantages of each algorithm, and then take advantages of the grey prediction algorithm and the prediction algorithm responding to the emergency to deal with the actual situation. On the hardware, this paper used the ultra low power 430 MCU to control the working pattern, and then applied the improving algorithm to the DSP+FPGA platform of power management successfully.At last, in the platform of the DSP+FPGA, this paper measured the power of the normal mode, idle mode, sleep mode and the dynamic power management mode by energy meter to improve the dynamic power management can reduce the power most.