Pulse or digital communications – Systems using alternating or pulsating current – Plural channels for transmission of a single pulse train
Reexamination Certificate
1998-03-27
2001-05-15
Chin, Stephen (Department: 2634)
Pulse or digital communications
Systems using alternating or pulsating current
Plural channels for transmission of a single pulse train
C375S260000, C375S261000, C375S262000, C375S340000
Reexamination Certificate
active
06233286
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to data communication and, in particular embodiments, to the communication of data over channel exhibiting intersymbol interference.
An intersymbol interference (ISI) channel is one in which the signal energy of a signal point transmitted in one signaling interval becomes dispersed over a number of adjacent signaling intervals. One example is a channel used to communicate data over a cable TV coaxial cable. Another is a wireless, or cellular, telecommunications channel, the dispersion being principally due to the phenomenon referred to as multi-path. In any such ISI channel, the dispersed energy combines with signal points transmitted in the adjacent intervals and thus constitutes a source of noise in those other interval . When the level of ISI is small, a so-called linear equalizer is effective in mitigating against it. However, if the ISI is severe, other, more powerful techniques must be brought into play. Typically, a decision feedback equalizer (DFE) is used. DFE is an interference cancellation technique. It estimates the amount of ISI in a given received signal point and subtracts the ISI estimate therefrom to arrive as a ISI-compensated signal point from which a decision is made as to the identity of the transmitted signal point.
Even better results can be achieved using a technique referred to as “maximum likelihood decoding for ISI channels.” (See, for example, G. D. Forney, Jr., “The Viterbi Algorithm,” Proc. IEEE, Vol. 61, pp. 268-278, March 1973.). That technique takes advantage of the recognition that the ISI phenomenon can be modeled as a for of convolutional coding within the channel. Therefore known techniques for decoding convolutional codes, such as Viterbi decoding, can be applied to the received ISI-corrupted signal even in a case where the transmitted signal point stream was not processed with any explicit convolutional coding in the transmitter. The underlying theory of this approach is that it provides what I have come to refer to as “conversion gain,” this being the improvement in error immunity that results from the con version of at least a portion of the harmful ISI into useful signal energy. Thus rather than Subtracting the ISI energy from the received signal, the ISI is returned to the signaling interval from which it came. The signal-to-noise ratio, and therefore the receiver error performance, are thereby improved.
Practical application of this approach is quite limited, however. The number of so-called states, S, in the Viterbi decoder is roughly given by S=C
L
, where C is the number of signal points in the transmitter constellation (its “size”), and L is the number of signaling intervals over which there is significant dispersion. Thus except for cases in which C and L are relatively small, the number of states, S, and thus the associated complexity of the Viterbi decoder, will be prohibitively large from an implementational standpoint. Indeed, few present-day communication systems meet the criterion of small C and/or small L. Moreover, if explicit convolutional coding were to be implemented in the transmitter, the complexity would be even far greater, reducing even further the practical applicability of this technique. The principal object the invention, then, is to be able to achieve the performance advantage offered by the above-described maximum likelihood decoding for ISI channels, without suffering the implementational complexity that arises for large values of C and/or L.
SUMMARY OF THE INVENTION
The above and other objects are achieved by a technique for processing transmitted signal points received from a transmitter, each transmitted signal point being from a respective one of a sequence of transmitter subsets of signal points as defined by an N-state transmitter trellis diagram. Responsive to the received signal points, successive sets of M surviving signal point paths through a receiver trellis defined by a receiver trellis diagram are identified and, in accordance with the invention, the identifying is such that more than one path corresponding to an individual sequence of transmitter subsets of signal points can be identified as ones of the surviving paths. The transmitted signal points can then be recovered as a function of the surviving paths.
In preferred embodiments, the received signal points are decoded as though they had been divided in the transmitter into finer subsets-referred to herein as the “receiver subsets”—than they actually were, and using a refined receiver trellis diagram The refined receiver trellis diagram is the same as the trellis diagram used by the transmitter, except that it has more than one branch for at least some of its state transitions. Each of those more-than-one branches is associated with one of the finer receiver subsets which is a part of the transmitter subset associated with the state transition. Calculation of branch metrics in the decoder and subsequent identification of surviving paths are carried out in a way which allows for the possibility that two or more paths, each associated with the same sequence of transmitter subsets, can be identified as surviving paths.
Also in preferred embodiments, each branch metric is calculated based on a respective ISI-compensated received signal point, wherein-as described in my co-pending U.S. patent application Ser. No. 09/023,063 filed Feb. 12, 1998—the ISI estimate is a function of the surviving path from which the branch in question emanates, and the surviving paths are identified using path-oriented, rather than state-oriented, pruning.
The invention is applicable not only to arrangements implementing coded modulation, for which N is greater than 1, but also to arrangements using uncoded modulation, for which N is equal to 1. In the latter arrangements, the decode is as just described (even though uncoded modulation arrangements typically use so-called “slicing” rather than decoding), the transmitter trellis diagram being a default trellis diagram having only one state and only one state transition.
Advantageously, the present invention introduces only a relatively small amount of complexity to the processing carried out in the receiver, but provides a very large amount of processing gain, by which I mean the combined effect of a) the coding gain afforded by the transmitter modulation coding, if any, and b) the aforementioned conversion gain. That processing gain manifests itself in the form of a significant improvement in the overall error rate performance of the communication system.
REFERENCES:
patent: 5056117 (1991-10-01), Gitlin et al.
patent: 5559561 (1996-09-01), Wei
patent: 5570391 (1996-10-01), Lin et al.
patent: 5675590 (1997-10-01), Alamouti
G. D. Forney, Jr., “The Viterbi Algorithm,” Proceedings Of the IEEE, vol. 61, pp. 268-278, Mar. 1973.
A. Duel-Hallen and C. Heegard “Delayed Decision-Feedback Sequence Estimation,” IEEE Transactions on Comm., vol. 37, pp. 428-436, May 1989.
P. R. Chevillat and E. Eleftheriou, “Decoding Of Trellis-Encoded Signals In The Presence of Intersymbol Interference And Noise,” IEEE Trans on Comm., vol. 37, pp. 669-676, Jul. 1989
M. V. Eyuboglu and S. U. H. Qureshi, “Reduced-State Sequence Estimation For Coded Modulation On Intrersymbol Interference Channels,” IEEE Journal on Selected Areas in Comm., vol. 7, pp. 989-995, Aug. 1989.
S. Mohan and A. K. Sood, “A Multiprocessor Architecture For The (M, L)- Algorithm Suitable For VlSI Implementation,” IEEE Trans. on Comm., vol. 34, pp. 1218-1224, Dec. 1986.
F. K. Soong and E. F. Huang., “A Tree-Trellis-Based Fast Search For Finding The N Best Sentence Hypotheses In Continuous Speech Recognition,” ICASSP 91. International Conf. on Acoustics, Speech and Signal Processing, pp. 705-708, 1991.
N. Seshadri and C.-E. W. Sundberg, “List Viterbi Decoding Algorithms With Applications, ” IEEE Transactions on Communications vol. 42, pp. 313-323. Feb./Mar./Apr. 1994.
Chin Stephen
Liu Shuwang
Lucent Technologies - Inc.
McCabe John F.
Slusky Ronald D.
LandOfFree
Path-oriented decoder using refined receiver trellis diagram does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Path-oriented decoder using refined receiver trellis diagram, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Path-oriented decoder using refined receiver trellis diagram will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2518407