Pulse or digital communications – Transceivers – Modems
Reexamination Certificate
2000-03-21
2004-07-20
Chin, Stephen (Department: 2734)
Pulse or digital communications
Transceivers
Modems
Reexamination Certificate
active
06765956
ABSTRACT:
CROSS-REFERENCES TO RELATED APPLICATIONS
Not Applicable.
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT
Not Applicable.
BACKGROUND OF THE INVENTION
The present embodiments relate to modems, and are more particularly directed to frame synchronization in wireline modems.
The high-speed exchange of digital information between remotely located computers is now a pervasive part of modem computing in many contexts, including business, educational, and personal computer uses. It is contemplated that current and future applications of high speed data communications will continue the demand for systems and services in this field. For example, video on demand (“VOD”) is one area which has for some time driven the advancement of technology in the area of digital information exchanges. More recently, the rapid increase in use and popularity of the Global Internet has further motivated research and preliminary development of systems directed to advanced communication of information between remotely located computers, particularly in accomplishing higher bit rates using existing infrastructure.
Various types of modems have been and continue to be developed to achieve the high speed data communication arising from matters such as those described above. For example, ISDN modems typically transmit and receive data at speeds of 64 Kbps and 128 Kbps. As another example, cable modems are currently under development with the promise of data connections of much higher speeds than ISDN. More particularly, cable modems are anticipated to receive data at up to 10 Mbps and send data at speeds up from 2 to 10 Mbps. Still other modems are also known in the art.
Given the proliferation of wireline modems, many such modems use frame structures to communicate information. By way of example, therefore,
FIG. 1
illustrates such a frame designated generally at
10
. By way of example, frame
10
is a quadrature amplitude modulation (“QAM”) frame, where it is known in the art that QAM frames encode data in an analog signal which includes one of a different available combination of phases and amplitudes to represent different bit patterns. Within frame
10
is provided training data
12
to tune an equalizer in the receiving modem. In order to locate training data
12
, frame
10
also includes a synchronization or “sync” sequence
14
placed at the start of frame
10
. As a result, a receiving modem must recognize sync sequence
14
at some point during the receipt of frame
10
. Once this recognition occurs, it may be determined where the beginning of the frame is located, and it thus will be known where the end of sync sequence
14
occurs. Knowing the location of the end of sync sequence
14
thereby identifies the location of the beginning of training data
12
. Additionally, frame
10
includes user data
16
located after training data
12
and, thus, by locating the position of training data
12
, the location of user data
16
also may be determined.
By way of further background,
FIG. 2
illustrates a block diagram of a receiver path in a modem
18
. The block diagram of modem
18
is a general representation and, thus applies in general to the prior art but also may be modified as described later to form an inventive embodiment. Modem
18
receives frame data as an analog signal from a wireline (e.g., a telephone line or a cable, such as a coax cable), and that data is input to an analog-to digital converter (“ADC”)
20
where it is converted to a digital form. The digitally converted signal then passes to a timing recovery block
22
that re-times the sampling of the input waveform so that the receive sampling frequency tracks that of the transmitter in frequency. Next, the signal passes to a demodulator
24
that removes the data from its modulated form, thereby producing the baseband values of the data. Note that the frequency of the baseband value signal output from demodulator
24
is typically at some integer multiple (or other fraction greater than one) of the symbol rate; commonly, therefore, the output of demodulator
24
is at two times the symbol rate. From the output of demodulator
24
, the demodulated data passes to both a sync block
26
as well as an equalizer and carrier recovery block
28
. Sync block
26
locates sync sequence
14
in each frame as detailed below, and when this location occurs sync block
26
asserts a SYNC signal to equalizer and carrier recovery block
28
so that it may synchronize itself to the incoming signal and perform training. Equalizer and carrier recovery block
28
outputs the equalized signal to a symbol decision block
30
. Symbol decision block
30
performs the function of estimating the transmitted data from the output of equalizer and carrier recovery block
28
. This is usually performed by finding the nearest point in the signal constellation to each received sample. This result is output to a deframer
32
. In addition, symbol decision block
30
feeds back a signal to equalizer and carrier recover block
28
in order to provide decision-directed tracking of changes in the channel during the data portion of the frame. Lastly, note that, the SYNC signal from sync block
26
is also connected to deframer
32
so that it too may synchronize itself to the incoming signal. Further, deframer
32
performs the function of removing training data
12
and sync sequence
14
from frame
10
, thereby leaving only user data
16
.
Looking now to sync block
26
in greater detail, it locates sync sequence
14
in each frame
10
by taking periodic samples, where this approach is now described with the benefit of a general timing illustration in
FIGS. 3
a
and
3
b
. Specifically, in
FIG. 3
a
, let the points P
0
through P
3
represent successive ideal sample locations in sync sequence
14
, with a common time period T between each location. In other words, in an ideal situation, sync block
26
would sample the incoming signal at the exact point in time corresponding to point P
0
; in the art, this point is sometimes referred to as the center of a so-called eye diagram, with it understood that an actual sample taken at this ideal point is most likely to result in proper synchronization, and any increase in time between this ideal point and the actual sample point correspondingly decreases the synchronization performance (i.e., decreases the chance of successful synchronization). Additionally, given the sampling period T, sync block
26
then also ideally samples at each interval of T thereafter, thereby sampling exactly at the points P
1
, P
2
, and P
3
illustrated in
FIG. 3
a
. However, various factors cause sync block
26
to take actual samples at a phase shifted point in time which is away from that of each point in
FIG. 3
a
. Such factors include the fact that there is no common clock or timing signal for synchronization between the transmitter and the receiver, and also may include other factors such as channel distortion and carrier errors. By way of example, therefore,
FIG. 3
b
again illustrates points P
0
through P
3
, and further illustrates a first scenario where a first actual sample S
0
is taken, followed thereafter by additional samples at each period of T thereafter. Thus, samples are taken at times represented as S
0
, S
1
, S
2
, and S
3
. As the samples are taken, a technique is used whereby the samples are convolved with a filter correlation sequence that represents a time reversed, complex conjugate of sync sequence
14
. As a result, the convolution determination will peak when sync sequence
14
is aligned with the filter correlation sequence. Also in this regard, in an effort to produce the greatest possible peak, note that sync sequence
14
is typically formed by selecting from the four highest energy points of the symbol constellation and, indeed, using only the two of those four points that have the greatest spectral distance between them (i.e., −15−j
max
+15+j
max
for QAM). Given these considerations, the convolution peak may be detected by comparing the convolution result against
Brady III W. James
Chin Stephen
Telecky , Jr. Frederick J.
Texas Instruments Incorporated
Williams Lawrence
LandOfFree
Multiple sampling frame synchronization in a wireline modem does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Multiple sampling frame synchronization in a wireline modem, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Multiple sampling frame synchronization in a wireline modem will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3250389