Research On Pre-Emphasis Compensation Technology And Its Efficient Implementation Algorithm For Digital-to-Analog Converter

The Digital-to-Analog Converter(DAC)has the fading of the SINC function in the amplitude-frequency characteristics.Compensating the DAC with digital filter is more and more concerned.Using digital interpolation to increase the slew rate simply reduces the attenuation effect of the DAC to an acceptable level,and it does not eliminate this phenomenon from the generation mechanism.The frequency domain sampling method has the problem that there is no effective solution to get clear requirements for the pass band resistance.It is necessary and practical to design a filter that can eliminate the attenuation effect in principle and maximize the resource consumption.At the same time,there is a rate mismatch between the hardware platform of the compensation filter and the DAC,which seriously affects its specific application implementation.In this paper,the DAC pre-emphasis compensation technology and its efficient implementation algorithm are studied.The main work is as follows:In Chapter 3,the design scheme of DAC frequency response pre-emphasis compensation filter is studied.Aiming at the problem that the existing compensation method cannot control the pass-stop band flexibly,a least-squares algorithm and a minimax-based algorithm are proposed respectively.Among them,the pre-emphasis compensation filter based on the least-squares algorithm takes the control of error energy as the core index.The simulation shows that the algorithm can realize the flexible adjustment of the pass-stop band.Under the same conditions,compared with the filter band,the compensation filter significantly reduces the error energy.The pre-emphasis compensation filter based on minimax algorithm takes ripple control as the core index.The simulation shows that the algorithm can also realize the flexible adjustment of the pass-stop band.Under the same conditions,compared with the filter bank,the compensation filter significantly reduces the ripple size.Starting from a large number of simulation data,the order estimation function of the filter is derived.Through the curve fitting and function derivation under MATLAB simulation,the order estimation function of the compensation filter under different conditions is obtained.The estimation accuracy of the estimation function is verified by traversing the parameter values,which provides reference and assistance for resource estimation and allocation in top-level design.Chapter 4 studies the efficient parallel structure of pre-emphasis compensation filters.Aiming at the problem that the existing method consumes too much hardware resources,we use the Cook-Toom algorithm and the symmetry of sub-filter coefficients with linear phase to reduce the number of multipliers,a method based on The improved parallel structure of the pre-emphasis compensation filter of the Cook-Toom algorithm.Compared with the existing parallel decomposition algorithm of FIR filter,the structure proposed in this chapter is more hardware efficient.When verifying its performance,it first simulates the integrity of the verification signal before and after parallel,and obtains the proposed error by horizontally comparing the signal error before and after parallel.Corroboration of structural accuracy.The improved structure proposed in this paper is compared with the existing methods to verify the difference between the number of adders and multipliers under different orders and different parallel degrees.When the number of parallel decomposition paths is larger and the filter length is longer,the computational resources saved by the improved structure proposed in this paper are more significant.