The multi-innovation stochastic gradient method for multi-frequency signal models

This paper considers the problem of parameter estimation for the multi-frequency sine signal based on the nonuniform measured data. In accordance with the gradient search and the multi-innovation theory, a multi-innovation stochastic gradient algorithm is developed to estimate the parameters of the multi-frequency sine signal. Through adding a forgetting factor and a convergence index, the estimation accuracy is improved. The effectiveness of the proposed algorithm is illustrated by a numerical example.