Engineering Papers⌕ Search

Engineering topics

Reed, I. S.

Publications and source records attributed to Reed, I. S..

At least 73 records · Page 4
All sourcesNASA NTRSDOE OSTI
NASA NTRS1981

Addendum to 'A new hybrid algorithm for computing a fast discrete Fourier transform'

The reported investigation represents a continuation of a study conducted by Reed and Truong (1979), who proposed a hybrid algorithm for computing the discrete Fourier transform (DFT). The proposed technique employs a Winograd-type algorithm in conjunction with the Mersenne prime-number theoretic transform to perform a DFT. The implementation of the technique involves a considerable number of additions. The new investigation shows an approach which can reduce the number of additions significantly. It is proposed to use Winograd's algorithm for computing the Mersenne prime-number theoretic transform in the transform portion of the hybrid algorithm.

Reed, I. S.↗
NASA NTRS1981

A decoding failure test for the transform decoder of Reed-Solomon code

Using a finite field transform, a transform decoding algorithm is able to correct erasures as well as errors of any (n,k,d) Reed-Solomon code over the finite field GF(q). A pitfall of transform decoding and how to avoid it are discussed. A simple test is given so that the decoder fails to decode instead of introducing additional errors, whenever the received word contains too many errors and erasures.

Miller, R. L.↗
NASA NTRS1981

Fast polynomial transform and its implementation by computer

A fast polynomial transform (FPT) algorithm for computing two-dimensional cyclic convolutions on a general-purpose computer is demonstrated and compared with the FFT approach. An FPT program for two-dimensional convolutions written in FORTRAN is shown to be 20% faster than the conventional FFT algorithm. This higher speed advantage makes the FPT algorithm a candidate for many two-dimensional digital image filtering applications.

Reed, I. S.↗
NASA NTRS1981

Fast transforms for decoding Reed-Solomon codes

In the paper it is shown that the Chinese remainder theorem when coupled with a modification of Winograd's method can be used to compute Fourier-like transforms over GF (s super m), where m = 2, 3, . . . , 8. These new transform techniques are to decode Reed-Solomon codes of block length 2 super m -1. The results are shown to be more efficient than the more conventional method.

Reed, I. S.↗
NASA NTRS1980

A bandpass filter for the enhancement of an X-ray reconstruction of the tissue in the spinal canal

In this communication, a new bandpass reconstruction filter is developed to partially remove the low spatial frequencies of the bone and the soft tissue in an X-ray reconstruction of a lumbar spine. This partial removal of the low frequencies suppresses the bony vertebral body and the soft tissue components within the projections of actual clinical data. It also has the effect of enhancing the sharp edges of the fatty tissue surrounding the spinal cord region. The intent of this effort is to directly visualize the spinal cord without the need for water-soluble contrast (e.g., metrizamide) to be installed through lumbar punctures.

Reed, I. S.↗
NASA NTRS1980

Efficient program for decoding the /255, 223/ Reed-Solomon code over GF/2 to the 8th/ with both errors and erasures, using transform decoding

The paper deals with a method developed for decoding a (255, 223) Reed-Solomon code over GF(2 to the 8th) with both errors and erasures. The matrix of decoding times for correcting errors and erasures of the code using a simplified decoder is presented. It is shown that the algorithm proposed is faster by a factor of from three to seven.

Miller, R. L.↗
NASA NTRS1980

On the application of a fast polynomial transform and the Chinese remainder theorem to compute a two-dimensional convolution

A fast algorithm is developed to compute two dimensional convolutions of an array of d sub 1 X d sub 2 complex number points, where d sub 2 = 2(M) and d sub 1 = 2(m-r+) for some 1 or = r or = m. This algorithm requires fewer multiplications and about the same number of additions as the conventional fast fourier transform method for computing the two dimensional convolution. It also has the advantage that the operation of transposing the matrix of data can be avoided.

Truong, T. K.↗
NASA NTRS1980

Wiener filtering of successive overlapping sections of an X-ray reconstruction

In this paper, a technique is developed to improve the image quality of an object in three-dimensional reconstruction by a weighted average of successive overlapping two-dimensional sections of the object. It is demonstrated that the signal-to-noise ratio of a two-dimensional picture can be improved by roughly a factor of 2 for typical X-ray beam shapes.

Reed, I. S.↗
NASA NTRS1979

Simplified algorithm for correcting both errors and erasures of Reed-Solomon codes

Using a finite-field transform, a simplified algorithm for decoding Reed-Solomon codes is developed to correct erasures as well as errors over the finite-field GF(q to the m power), where q is a prime and m is an integer. If the finite-field transform is a fast transform, this decoder can be faster and simpler than a decoder that uses more conventional methods.

Reed, I. S.↗
NASA NTRS1979

Decoding of B.C.H. and R.S. codes with errors and erasures using continued fractions

Through the use of continuing fractions, a simplified algorithm for decoding B.C.H. and R.S. codes is developed that corrects both erasures and errors on a finite field GF(q to the m). It is noted that the decoding method is a modification of the Forney-Belekamp technique. Finally, it is believed that the present scheme is both simpler to understand and to implement than more conventional algorithms.

Reed, I. S.↗
NASA NTRS1979

A fast technique for computing syndromes of BCH and RS codes

A combination of the Chinese Remainder Theorem and Winograd's algorithm is used to compute transforms of odd length over GF(2 to the m power). Such transforms are used to compute the syndromes needed for decoding CBH and RS codes. The present scheme requires substantially fewer multiplications and additions than the conventional method of computing the syndromes directly.

Reed, I. S.↗
NASA NTRS1979

A new hybrid algorithm for computing a fast discrete Fourier transform

In this paper for certain long transform lengths, Winograd's algorithm for computing the discrete Fourier transform (DFT) is extended considerably. This is accomplished by performing the cyclic convolution, required by Winograd's method, with the Mersenne prime number-theoretic transform developed originally by Rader. This new algorithm requires fewer multiplications than either the standard fast Fourier transform (FFT) or Winograd's more conventional algorithm. However, more additions are required.

Reed, I. S.↗
NASA NTRS1978

A simplified algorithm for correcting both errors and erasures of R-S codes

Using the finite field transform and continued fractions, a simplified algorithm for decoding Reed-Solomon (R-S) codes is developed to correct erasures caused by other codes as well as errors over the finite field GF (q(m), where q is a prime and m is an integer. Such an R-S decoder can be faster and simpler than a decoder that uses more conventional methods.

Reed, I. S.↗
NASA NTRS1978

A fast computation of complex convolution using a hybrid transform

The cyclic convolution of complex values was obtained by a hybrid transform that is a combination of a Winograd transform and a fast complex integer transform. This new hybrid algorithm requires fewer multiplications than any previously known algorithm.

Reed, I. S.↗
NASA NTRS1978

On the fundamental structure of Galois switching functions

In connection with investigations conducted by Menger (1969), Benjauthrit and Reed (1976), and Pradhan (1978) two methods have been reported for deriving a unique Galois switching function from a given truth table description of the function. However, neither of the two methods appears satisfactory. One method often contains many redundant terms in its formulation, whereas the second method requires a great number of multiplications and additions. By some algebraic manipulations, an expanded formula is obtained which combines the best features of both methods. This formula makes it possible to compute the coefficients of the desired function more directly and probably with less effort. The conduction of calculations by means of the new formula is illustrated with the aid of examples.

Benjauthrit, B.↗
NASA NTRS1978

A new hybrid algorithm for computing a fast discrete Fourier transform

For certain long transform lengths, Winograd's algorithm for computing the discrete Fourier transform is extended considerably. This is accomplished by performing the cyclic convolution, required by Winograd's method, with the Mersenne-prime number theoretic transform. This new algorithm requires fewer multiplications than either the standard fast Fourier transform or Winograd's more conventional algorithm.

Reed, I. S.↗