Voice and image recovery from lost samples

In a digital communication system, a common problem is a loss of samples due to channel fading, interfering signal, and additive noise. The authors introduce novel schemes for the interpolation of the missing samples, called the division method. From the nonuniform sampling theorem it is known that, as long as the sampling rate of a signal satisfies the Nyquist rate, the signal is uniquely determined by the set of the nonuniform samples. The authors consider the distorted signals as a set of nonuniform samples and derive the methods based on the theorem. A number of simulations have been performed for both distorted voice and image signals. The advantage of this method is that it is very simple to implement in real time