A DFT-Based method (DBM) has been proposed to compensate for the performance degradation caused by clipping distortion at the expense of bandwidth expansion. On the other hand, in any communication systems, conventional channel coding methods can be employed to improve performance. In this letter, the performance of the DBM and the channel coding methods (CCM) are compared. Furthermore, we introduce a hybrid system which outperforms both the DBM and the CCM.

In this paper we analyze the samples of a signal in time and frequency from a unique point of view. Besides the unified treatment of the subject and the insight gained from this approach, the following original results are claimed. A new method is proposed to estimate the bandwidth and the shape of the spectrum of a signal. The most important part of the paper is the contribution on the frequency spectrum of the interpolating functions of the samples of a signal. As a consequence a novel method is proposed for equalizing the distortion of an interpolating function such as sample and hold signal. The realization is fairly simple and modular in concept.

The reconstruction of the modulating signal from the zero crossings has been treated extensively in the literature. However, the analysis on the zero crossings is only an approximation based on the average pulse counts. The relating reconstruction methods or zero-crossing discriminations, are also elaborate and costly. It is the purpose of this letter to analyze the zero crossings of an FM signal exactly in the frequency domain. From this analysis, a very simple method is proposed to demodulate an FM signal.

In this paper, we study a new hybrid speech coder which employs a modified version of the harmonic sinusoidal analysis to encode the periodic contents of speech waveform and to split the speech spectrum into two frequency regions of harmonic and random components. A reliable fundamental frequency is estimated for the harmonic region using both speech and its linear predictive (LP) residual spectrum. The peak envelope of speech spectrum is encoded in terms of the coefficients of an all-pole spectrum. A harmonic tracking algorithm appropriately interpolates the sinusoidal parameters to achieve a smooth transition between the parameter update points and to reconstruct an essential level of periodicity in the synthetic voiced speech. The random part of spectrum and unvoiced speech are coded using the conventional CELP algorithm. The individual components are then combined at the decoder to obtain the synthetic speech. The proposed hybrid coder which combines the powerful features of the sinusoidal and CELP coding algorithms yeilds a high quality synthetic speech at 4.05 kbps.

Redundancy is introduced by sampling a bandlimited signal at a higher rate than the Nyquist rate. In the cases of erasures due to fading or jamming, the samples are discarded. Therefore, what we get at the output of the receiver is a set if nonuniform samples obtained from a uniform sampling process with missing samples. As long as the rate of nonuniform samples is higher than the Nyquist rate, the original signal can be recovered with no errors. The sampling theorem can be shown to be equivalent to the fundamental theorem of information theory. This oversampling technique is also equivalent to a convolutional code of infinite constraint length is the Field of real numbers. A DSP implementation of this technique is through the use of a Discrete Fourier Transform (DFT), which happens to be equivalent to block codes in the field of real numbers. An iterative decoder has been proposed for erasure and impulsive noise, which also works with moderate amount of additive random noise. The iterative method is very simple and efficient consisting of modules of Fast Fourier Transforms (FFT) and Inverse FFT's. We also suggest a non-linear iterative method which converges faster than the successive approximation. This iterative decoder can be implemented in a feedback configuration. Besides FFT, other discrete transforms such as Discrete Cosine Transform, Discrete Sine Transform, Discrete Hartley Transform, and Discrete Wavelet Transform are used. The results are comparable to FFT with the advantage of working in the field of real numbers.

One of the categories of decoding techniques for DFT codes in erasure channels is the class of iterative algorithms. Iterative algorithms can be considered as kind of alternating mapping methods using the given information in a repetitive way. In this paper, we propose a new iterative method for decoding DFT codes. It will be shown that the proposed method outperforms the well-known methods such as Wiley/Marvasti, and ADPW methods in the decoding of DFT codes in erasure channels.

Discrete Fourier Transform (DFT) is used for error detection and correction. An iterative decoder is proposed for erasure and impulsive noise which also works with moderate amount of additive random noise. The iterative method is very simple and efficient consisting of modules of Fast Fourier Transforms (FFT) and Inverse FFT's. This iterative decoder can be implemented in a feedback configuration.