Coding scheme for cable modems

Details Coding scheme for cable modems Coding scheme for cable modems

1999-06-30

2003-04-15

Ghayour, Mohammad H. (Department: 2634)

Pulse or digital communications

Systems using alternating or pulsating current

Plural channels for transmission of a single pulse train

C714S755000

Reexamination Certificate

active

06549584

ABSTRACT:

CROSS-REFERENCE TO RELATED APPLICATIONS
Not applicable.
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT
Not applicable.
BACKGROUND OF THE INVENTION
This invention is in the field of data communications, and is more specifically directed to coding techniques for high data rate transmissions via cable modems.
In the field of data communications, particularly in the provision of Internet access to homes, the use of the cable television (CATV) network as the medium of communication has become attractive. The attractive features of the CATV network include the relatively large installed base of homes having cable television service, and also the inherently high data rates that may be carried by the coaxial cable with which cable television programs are delivered, especially when compared with twisted-pair copper wiring commonly used in telephone service. As such, significant effort has been undertaken to develop the appropriate technology for two-way broadband communication over the CATV network. Attention is directed, in this regard, to Perkins and Gatherer, “Two-way broadband CATV-HFC networks: state-of-the-art and future trends”,
Computer Networks,
Vol. 31 (Elsevier Science B.V., 1999), pp. 313-326, for a survey of the current state of technology in this field.
An important measure of any high-data-rate communications approach is referred to as “spectral density”, or synonymously as “spectral efficiency”. Spectral density refers to the number of bits that may be communicated per second for a given frequency. Of course, the higher the spectral density, the more information that can be communicated at the transmission frequency. In recent -years, relatively complex modulation techniques have been developed to improve the spectral density of data communications. It is now commonplace for communications to be carried out using phase and amplitude modulation, in combination, to modulate digital data into the transmitted signal. As is fundamental in such modulation, a “constellation” is defined by a number of discrete points in complex Euclidean space, each point representative of a value of a data symbol. An example of a complex modern constellation, referred to in the art as 256 QAM, has 256 points mapped into a sixteen-by-sixteen array in complex space; this constellation thus. permits the transmission of eight-bit symbols, and provides excellent spectral density.
As is also well-known in the art, the spectral density of a transmission scheme is limited by the noise in the transmission channel. If the noise is excessive, constellation points may be mistaken for one another, resulting in. an error in the transmission. As spectral density increases (i.e., with more points in the complex space constellation), the distance between adjacent constellation points decreases; for a given level of noise, a smaller distance between adjacent points directly relates to the likelihood of a transmission error.
Redundant encoding techniques have been used in combination with phase and amplitude modulation to ease the tradeoff between spectral density and error rates. One encoding technique is referred to as “trellis” coding, which is a species of convolutional encoding. A survey of trellis coding is described in Ungerboeck, “Trellis-Coded Modulation with Redundant Signal Sets Part I: Introduction”,
IEEE Communications Magazine,
Vol. 25, No. 2 (1987), pp. 5-11. According to trellis coding techniques, the constellation is divided into subsets, but the signal transitions are limited by way of a finite state machine. In this way, depending upon the particular implementation, transitions are not permitted to neighboring points in the constellation in every state, while transitions from state to state are permitted so that all points in the available constellation are used with equal frequency. The use of the finite state machine thus includes the value of prior symbols into the determination of the output signal, providing controlled redundancy in the encoding scheme. In trellis coding, only a portion of the bits of the symbol are encoded while the remaining symbol bits are transmitted in an uncoded fashion. The coded bits select the constellation subset to which the symbol relates, while the uncoded bits indicate the particular point within the selected constellation subset corresponding to the symbol. Because of the subdivision of the constellation, however, the Euclidean distance between the uncoded bits is increased. In other words, for a given bit error rate, the signal-to-noise ratio required using trellis coding may be a few dB less than for the uncoded case.
By way of further background, U.S. Pat. No. 5,511,082 describes a method and apparatus for encoding digital data using a convolutional code to encode part of the transmitted symbols. The convolutional code in this example has a coding rate of 4/5, which indicates that one redundant bit is inserted for every four symbol bits to be encoded. An extension of this approach is used in the trellis coding specified for cable modems according to ITU-T Recommendation J.83 Annex B, and in other cable modems specifications. In this approach, four constellation subsets are used, with the two least significant bits being coded according to a convolutional code, with the coded result used to select among the constellation subsets.
FIG. 1
illustrates an example of this methodology, as utilized in a simple 16-QAM context. As evident from
FIG. 1
, the least significant bit of the codeword alternates along the imaginary axis of the constellation, and the second least significant bit alternates along the real axis. Considering the two LSBs to indicate the selection of the sub-constellation (e.g., all points xx01 correspond to one sub-constellation or subset), it is evident from
FIG. 1
that adjacent points within a sub-constellation are significantly separated from one another than are adjacent points in the overall 16-QAM constellation. In this example, this approach doubles the Euclidean distance between adjacent uncoded constellation bits (providing a 6 dB improvement in error performance).
FIG. 2
illustrates, in block diagram form, the coder functions used to implement this LSB encoding approach described in U.S. Pat. No. 5,511,082. Parser
3
receives a bitstream B, and parses the incoming bitstream B into m-bit symbols, considering the coding. In this example, where 4/5 convolutional coders
5
I
and
5
Q
are used, five m-bit symbols are to be generated at a time; as such, parser
3
forwards 5(m−2) bits in uncoded fashion to QAM map function
7
. Parser
3
also forwards four bits to 4/5 convolutional coder
5
I
, and four bits to 4/5 convolutional coder
5
Q
.
FIG. 3
illustrates the functional construction of one of 4/5 convolutional coders
5
I
and
5
Q
which, of course, are similarly constructed. As shown in
FIG. 3
, four input bits x
0
through x
3
are serially presented (bit x
3
being earliest in time, and bit x
0
latest) to convolutional coder
5
. Convolutional coder
5
includes four delay stages S
0
through S
3
, the outputs of which are applied to exclusive-OR functions
9
0
and
9
1
according to the desired encoding. In this example, exclusive-OR function
9
0
receives the outputs of all four delay stages S
0
through S
3
and also the currently applied input bit, and exclusive-OR function
9
1
receives the currently applied input bit, and the outputs of delay stages S
1
and S
3
. As described in the above-referenced U.S. Pat. No. 5,511,082, this arrangement corresponds to octal generators of 25
8
(10101
2
) and 37
8
(11111
2
), for exclusive-OR functions
9
0
and
9
1
, respectively.
The outputs of exclusive-OR functions
9
are applied to puncture logic
11
, which selects the bits to be output by coder
5
according to the desired puncture scheme. In this example, puncture logic
11
selects each of the four bits output by exclusive-OR function
9
0
from the sequence of four input bits x
0
through x
3
, and only the fourth (last) bit output by exclusive-OR function
9
1
. As described in t

Affiliated with

Also associated with

Rating

Coding scheme for cable modems does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Coding scheme for cable modems, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Coding scheme for cable modems will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3072654