Pulse or digital communications – Pulse code modulation – Differential
Reexamination Certificate
2000-02-07
2002-07-09
Chin, Stephen (Department: 2634)
Pulse or digital communications
Pulse code modulation
Differential
C375S222000, C375S286000
Reexamination Certificate
active
06418170
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to a telecommunications system and more particularly to a method and apparatus for eliminating problems of 180-degree phase reversal in the communications channel when transmitting data utilizing a PCM modulation system.
BACKGROUND OF THE INVENTION
PCM modulation systems are those in which data is transmitted from an analog modem, to an analog loop and a central office to a digital modem. The modulation scheme utilized in such systems includes mapping the incoming data to so-called equivalence classes. These equivalence classes are utilized to allow a larger minimum distance between constellation points utilized in encoding the incoming data stream. Equivalence classes are often used with pre-coding as described in U.S. patent application Ser. No. 08/999,249 entitled Device and Method for Pre-Coding Data Signals for PCM Transmission, assigned to the assignee hereof and incorporated herein by reference. “Constellation points” refers to numbers that ultimately translate to the voltage levels with which the particular incoming symbols are represented in the analog loop of the PCM channel in the data transmission scheme. By enabling larger distances between the constellation levels, one is allowed to communicate with a minimum of error, or stated differently, in a more robust manner. It will be noted that if the levels associated with the constellation are close together, the levels are more difficult to distinguish from one another, resulting in detection errors.
While the use of the equivalence classes is known, when transmitting information utilizing such equivalence classes if there is 180-degree phase reversal in the channel (i.e., the channel is “flipped” or negated), the transmitted data are corrupted and no useful information is obtained.
Phase reversals, while rare in the United States, are intentionally inserted into the communications channel for billing purposes in Europe and other locations. What in effect is done is that the channel is phase reversed by switching the transmission line pairs periodically, e.g., once a second. Each of the phase reversals is then counted at a central office or end user location. The counted phase reversals are then used as a “meter” for billing purposes.
It will be appreciated that in a PCM modulation system if the channel is phase reversed once every second, the loss of data is catastrophic. This is because information relating to the identity of an equivalence class is lost. One approach to solving the problem of the 180-degree phase reversal in the channel is to utilize overly large equivalence classes such that if the phase reversal occurs during the transmission of the equivalence class, the data maps back into the same class such as that there is no ambiguity.
The problem with such an approach is that such a scheme uses twice as many points to protect against channel reversal. This type of solution also decreases the transmit rate significantly. In one embodiment such an approach would decrease the transmit rate by 1 bit per symbol, or 8000 bits/sec. for a 56 kbit/sec modem.
SUMMARY OF THE INVENTION
In order to immunize the PCM modulation system from phase reversals, in the subject system an encoding scheme is provided for the equivalence classes in which equivalence class pairs are provided with “sign bits” which are opposite in value. Thus for each equivalence class there is a sign bit and there is a magnitude bit or set of bits.
It is a finding of this invention that the corruption of the data due to phase reversals of the communication channel is due to misidentification of equivalence classes at the receiver because the equivalence class label is corrupted. By relabelling the equivalence classes in the manner specified herein, and by using a differential encoder at the transmitter and a differential decoder at the receiver, any errors in the equivalence class label due to the channel inversion nonetheless result in a correct identification of the equivalence class at the receiver.
As to the equivalence class pairs, making the sign bit of one of the pair
0
and the other of the pair
1
, and by utilization of differential encoding and decoding, if there is a phase reversal in the channel, then after differential decoding, the sign bit of the decoded equivalence class matches the sign bit of the originally intended and encoded equivalence class. Further, the magnitude bits are unaffected by the phase reversal, because a phase reversal does not alter magnitude. Thus the identity of the equivalence class is not lost.
More specifically, in a PCM modem system in which equivalence classes are used to communicate information from a transmitter to a receiver, a method is provided to solve the problem of garbled data transmission due to 180° phase reversals in the communications channel. This is accomplished by remapping the equivalence classes into a form that can be differentially encoded and decoded such that the identity of equivalence classes is unaffected by a phase reversal in the channel. It is the loss of identity of an equivalence class which results in a corrupted transmission. In one embodiment, a differential encoder/decoder pair is utilized with the relabeled equivalence classes to permit identification of the equivalence classes unaffected by the phase reversal of the channel. This is because the sign bits assigned to the equivalence classes which may be affected by the phase reversal in the channel, when received, match the original sign bits upon differential decoding. Thus the information transmitted is received correctly whether the phase reversal is absent or present.
Note that equivalence classes are paired in terms of two sets of points where by negating the members of one set one obtains the members of the other set. Because of the subject equivalence class labeling system and because of the existence of the pairs, if the channel does reverse phase, there is no effect upon the classes. If there is a phase reversal, a class is mapped to its pair-mate. This does no harm because with the differential mapping the constellation points represent the same data.
In the illustrated embodiment, the mapping format used provides equivalence class labels in a binary-coded form involving a sign bit and magnitude bits. The pairs of original transmitted equivalence classes are equal in every bit position except one, the sign bit, where they are opposite.
Otherwise stated, the original equivalence classes are mapped to a number expressed in binary notation having a sign bit and magnitude bits. After mapping, by differentially encoding the sign bit of each equivalence class, the received equivalence classes can be decoded correctly regardless of a positive or negative phase of the channel. The success of the subject method is because if the signs of the differentially constructed equivalence classes are all “flipped” or sign-inverted, the received equivalence classes will nonetheless match the originally transmitted equivalence classes in identity which avoids corrupted or unintelligible data.
It will be appreciated that the magnitude bits are not affected by phase reversal of the channel. Magnitude bits are in reality voltage amplitudes whose magnitude is independent of the sign. Thus at the receive side, detecting amplitude, e.g., absolute value, yields the same value whether inverted or not. For example, the magnitude +10 and the magnitude of −10 are the same.
On the other hand, if the value of +10 is in equivalence class A, and the value −10 is in equivalence class B, then a phase reversal, while not changing the amplitude, puts a member of one equivalence class into another, its “phase-pair mate”. In this case the equivalence classes are confused as to their identity. As a result, received data is detected as a different, unintended pattern of data.
REFERENCES:
patent: 4713817 (1987-12-01), Wei
patent: 5040191 (1991-08-01), Forney, Jr. et al.
patent: 5387885 (1995-02-01), Chi
patent: 5818879 (1998-10-01), Eyuboglu et al.
patent: 6173015 (2001-01-01), Eyub
Kim Dae-Young
Maurer Patrick
Mehrabanzad Sepehr
Chin Stephen
Jiang Lenny
Motorola Inc.
LandOfFree
Method and apparatus for achieving 180° phase invariant... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for achieving 180° phase invariant..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for achieving 180° phase invariant... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2852572