Automatic compensation of nonlinear distortions

When a signal passes a system having a static, nonlinear transfer function, the output signal will be distorted. In addition to it, the defected output signal is usually contaminated by noise, due to the noisy environment. If the distortion level of the signal is not acceptable, we have to compensate it. This should be done carefully, because an improper compensation function can extremely amplify the noise. Our aim is to find a compensation function in the knowledge of the original nonlinear transfer function, which can keep both the distortion and the noise level low. In this paper we will show the further developments of our method, which can estimate the optimal level of noise suppression of the compensation function, based on Tikhonov's regularization technique. We will show a method that can find the optimal value of the regularization parameter.