Rate-distortion function for nonstationary time-varyingautoregressive (TVAR) processes

摘要

The Gaussian TVAR process is of primary interest because of itswide applicability to the modeling real-world sources, e.g., speechsignals and rasterized image signals. We then investigated andconstructed the rate-distortion function R(D) for the class oftime-varying nonstationary Gaussian signals. The numerical evaluationsof the R(D) function for 1D Gaussian TVAR processes are also presented