Universal parallel processing decoder

Details Universal parallel processing decoder Universal parallel processing decoder

2000-05-30

2004-07-06

Chin, Stephen (Department: 2734)

Pulse or digital communications

Systems using alternating or pulsating current

Plural channels for transmission of a single pulse train

C375S340000, C332S106000, C329S311000

Reexamination Certificate

active

06760385

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates, in general, to data transmission systems and, more particularly, to a universal parallel processing Viterbi decoder for decoding both feedforward and feedback trellis or convolutional codes utilized in high speed data transmission systems, such as in digital subscriber line systems.
BACKGROUND OF THE INVENTION
With the advent of high-speed data transmission systems, various technologies have emerged to provide increased reliability and accuracy of the data transmitted and received. Typically, such data is encoded prior to modulation and transmission, to increase error detection and error correction when the data is received and decoded. The encoded data stream is then mapped or modulated onto any of various, standardized signaling formats or constellations for transmission, such as pulse amplitude modulation (“PAM”) or quadrature amplitude modulation (“QAM”).
Different encoding schemes have been utilized or are being proposed for encoding in various high-speed data transmission systems, and are being included as recommendations or standards, such as the proposed T1.418 standard promulgated by the American National Standards Institute (“ANSI”). For example, for communication systems such as the next or second generation (“HDSL2”) of high bit rate digital subscriber line (“HDSL”), a 512-state, rate ½ feedforward convolutional encoder may be utilized, while other encoding schemes may be utilized for systems such as symmetric digital subscriber line (“SDSL”), asymmetric digital subscriber line (“ADSL”), and other forms of digital subscriber line systems generally referred to as “xDSL”. In addition, various older or legacy systems utilize other encoding systems, such as an 8-state, rate ½ feedback convolutional encoding and a 32-state, rate ½ feedback convolutional encoding.
With the use of these multiple and different encoding schemes for data transmission, corresponding decoding following data reception is increasingly difficult and complicated. In the prior art, for systems implementing codes having differing numbers of states and having different feedback or feedforward methodologies, entirely separate and independent decoders have been utilized for compatibility among these systems.
As a consequence, a need remains to provide for a single decoder which is capable of decoding both feedforward and feedback codes. Such a universal decoder should be able to decode codes having differing numbers of states, such as 512 states or 32 states. In addition, such a decoder should provide for parallel processing and should be programmable, allowing the same hardware to provide such universal decoding, rather than utilizing separate hardware to provide independent and redundant decoding systems.
SUMMARY OF THE INVENTION
The present invention provides a single, parallel processing decoder which is capable of decoding both feedforward and feedback codes. The universal decoder of the present invention not only decodes both feedforward and feedback codes, but also decodes codes having differing or variable numbers of states, such as 512 states or 32 states. The universal decoder of the present invention is capable therefore of decoding a 512-state feedforward code which may be utilized for HDSL2, and various legacy codes such as an 8-state feedback code and a 32-state feedback code. The preferred decoder is also programmable, allowing the same hardware to provide such universal decoding, without requiring separate hardware for independent and redundant decoding systems. Lastly, the preferred universal decoder is capable of utilizing parallel processing for both feedback and feedforward codes, providing an increase in the speed of decoding.
The data for decoding will have been typically encoded utilizing a plurality of encoder coefficients and transmitted as a signal from a transmitter to a receiver to form a received signal. In the preferred embodiment, the transmitter and receiver exchange information pertaining to whether the data was encoded utilizing a feedback or feedforward methodology, and exchange information concerning the values of the encoder coefficients utilized in that encoding.
In the decoding process, for each current state of the plurality of states, the method determines a most likely previous state to form a plurality of previous states. This is preferably accomplished by determining a branch metric, as a distance, such as a Euclidean distance, between the received, equalized signal and a closest point of a subset of a signaling constellation. With the various state transitions, the branch metrics are accumulated, forming path metrics, with the shortest path metric retained for each current state. Based upon the various path metrics, the method determines a terminating state and a penultimate terminating state (the next state following the terminating state), from the plurality of previous states. Next, the method determines a plurality of subset bits (in which the plurality of subset bits has a most significant subset bit). In the preferred embodiment, this is accomplished by re-encoding the most significant bit (“MSB”) of the penultimate terminating state, with the state of an encoder set to the values of the terminating state, to produce the subset bits corresponding to a state transition from the terminating state to the penultimate terminating state.
These subset bits are utilized to select a corresponding subset of the signaling constellation, and with the equalized output, to determine a comparatively closest signaling point. The comparatively closest signaling point, in turn, is mapped to a corresponding index. The preferred method then determines the higher significant decoded bits as equal to the higher significant bits of the corresponding index. When the data has been feedforward encoded, the decoded least significant input bit is determined as equal to the MSB of the penultimate terminating state. When the data has been feedback encoded, the decoded least significant input bit is determined as equal to the most significant subset bit.
Numerous other advantages and features of the present invention will become readily apparent from the following detailed description of the invention and the embodiments thereof, from the claims and from the accompanying drawings.


REFERENCES:
patent: 4520490 (1985-05-01), Wei
patent: 4761784 (1988-08-01), Srinivasagopalan et al.
patent: 4998253 (1991-03-01), Ohashi et al.
patent: 5432803 (1995-07-01), Liu et al.
patent: 5872817 (1999-02-01), Wei
patent: 5920599 (1999-07-01), Igarashi
patent: 5946361 (1999-08-01), Araki et al.
patent: 6141391 (2000-10-01), Morelos-Zaragoza et al.
patent: 6151370 (2000-11-01), Wei
patent: 6212654 (2001-04-01), Lou et al.
patent: 6259749 (2001-07-01), Andoh
patent: 6330684 (2001-12-01), Yamanaka et al.

Affiliated with

Also associated with

Rating

Universal parallel processing decoder does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Universal parallel processing decoder, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Universal parallel processing decoder will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3247374