Telecommunications – Receiver or analog modulated signal frequency converter – Modulation distortion or cross talk correction or elimination
Reexamination Certificate
1998-10-27
2001-02-20
Nguyen, Lee (Department: 2746)
Telecommunications
Receiver or analog modulated signal frequency converter
Modulation distortion or cross talk correction or elimination
C455S296000, C375S346000
Reexamination Certificate
active
06192227
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to noise detection in communication systems and more particularly relates to a method for detecting nonlinear distortion utilizing moment invariants.
BACKGROUND OF THE INVENTION
Few people today would deny that the world is currently undergoing a communications revolution of some sorts for many years. All of the modern societies of the world today, already place heavy reliance on their communications infrastructure to support their economies, to contribute to the standard of living, to provide communications for the military, etc. Few would argue that society today would not be as advanced if it were not for the benefits derived from modern, efficient, quick and inexpensive communications that is available in many part of the world today. The current and past trend in most high technology areas is better, faster, and cheaper as time goes on and the area of communications is no exception.
Most communications systems, however, are not noise free systems. They function to transmit information from one place to another but with noise being added along the way. Today's modern digital transmission systems are less prone to external noise than the older analog systems but bit errors still occur nevertheless. Noise is especially a problem for the majority of analog lines connecting central offices to people's homes. The majority of home users and small businesses currently connect to the Internet and other on-line services via dial up modems.
The modem signals generated are corrupted by various types of noise that is present on the telephone lines that carry and transmit the signal to the modem on the other end of the line. A block diagram illustrating a typical channel between two modems communicating over a telephone line is shown in FIG.
1
. The path, generally referenced
10
, of the telephone call begins at one of the modems
12
which is connected by the local loop
14
to the end or central office
16
. From the central office, the call proceeds to a toll office
20
via one of many toll-connecting trunks
18
. The toll office
20
is connected to an intermediate switching office
24
via one or more very high bandwidth intertoll trunks
22
. The toll offices
20
may be connected via one or more intermediate switching offices
24
. The path the signal takes from the other modem is similar but in the opposite direction.
A block diagram illustrating a model of a typical telephone channel with the various noise and distortion sources highlighted is shown in FIG.
2
. The channel, generally referenced
30
, in one direction comprises modem A
32
connected to local loop A
34
which in turn is connected to hybrid A
36
. The hybrid performs 2-wire to 4-wire conversion. The transmit signal path comprises summer
38
followed by various noise and distortion sources. The resulting signal input to the hybrid B
54
and transmitted to modem B
58
via local loop B
56
.
Gaussian noise, linear and nonlinear distortion encountered along the local loops A and B. Both near and far end echoes from the channel are encountered at the hybrids. Attenuator
62
simulates the far end echo added to the receive path A via summer
64
. Similarly, attenuator
60
simulates the far end echo added to the receive path B via summer
38
.
Blocks
40
,
68
represent round trip delay (RTD), amplitude distortion (AD) and envelope delay distortion (EDD). Blocks
42
,
70
represent intermodulation distortion (IMD). Blocks
44
,
72
represent phase jitter while blocks
46
m
74
represent frequency offset errors. In addition, attenuation of the signal is represented by attenuators
48
,
76
. Noise generators
82
,
80
represent noise that is added to the channel via summers
50
,
78
respectively.
Further, both the pulse code modulation (PCM) and adaptive differential pulse code modulation (ADPCM) blocks
52
,
66
introduce both quantization noise and nonlinear noise. Note that in practice, the different types of distortion and noise described above in the typical communications channel require the applications of different algorithms and processing in order to recover the original data from the distorted noisy waveform.
The detection of nonlinear distortion in the channel is important as different signal processing is performed for this type of distortion. Techniques for detecting nonlinear distortion already exist in the prior art. Some of these techniques will now be presented.
In a linear system, the frequencies input into the system are the frequencies that are output of the system. In a nonlinear system frequencies may be output that were not present at the input. Based on this idea, one prior art method includes transmitting a rake of frequency tones wherein the rake of tones contains one or more amplitude nulls at potential intermodulation frequencies. The presence of a frequency tone at one or more of the amplitude nulls implies that nonlinear distortion potentially exists.
A second prior art method for detecting nonlinear distortion includes transmitting flat spectrum noise, i.e., white noise, having a bandstop pit that contains a low noise level. If the noise level of the pit relative to the higher non bandstop noise level is increased, it implies that nonlinear distortion potentially exists.
A third prior art method for detecting nonlinear noise is to alter the transmitted power level of the signal by a known factor. If the received level of the signal is relatively unchanged by the same factor as the transmitted signal, it implies that nonlinear distortion potentially exists. This method assumes that there is no offset, i.e., the DC offset has been filtered out.
A major disadvantage of the above described methods of detecting nonlinear distortion is that the nonlinear noise is usually not detected with sufficient probability of success and with a high probability of false alarms. This is due to the fact that other types of noise nearly always accompany the nonlinear noise. In many cases, the total power of the linear distortion due to echoes, noise, Gaussian noise and quantization noise together exceed the power of the nonlinear noise by a factor of 10 or more. In such a situation, the above described prior art methods are not effective.
SUMMARY OF THE INVENTION
The present invention is a method for detecting nonlinear distortion over communication channels. The method utilizes moment invariants in determining whether nonlinear distortion exists. The method is particularly applicable to detecting the presence of nonlinear distortion over telephone lines that are used for modem communications.
The detection of the presence of nonlinear distortion is important in order that appropriate processing operations be performed which are required in the presence of such distortion. The signal processing required in the presence of nonlinear distortion is different than that required in the presence of other types of distortion and noise. The particular processing operations that are performed in the presence of nonlinear distortion function to improve the reliability of data transfer especially at higher data rates.
The nonlinearity distortion is detected by calculating high order moment invariants of the error magnitude present at the input to the slicer stage in the receiver. The method of the invention is effective, in part, because the probability density function of the noise is different, i.e., has a different shape, for a channel that contains nonlinear distortion than for one that contains linear, i.e., Gaussian, distortion. In addition, the moment invariant is more sensitive to the size of the error and not necessarily the energy level of the noise.
There is provided in accordance with the present invention, in a communications system having a channel and a receiver coupled to the channel, a method of detecting the presence of nonlinear distortion, the method comprising the steps of generating an error signal, calculating a moment invariant utilizing the error signal and declaring the existence of nonlinear distortion
Akin Gump Strauss Hauer & Feld & LLP
Conexant Systems Inc.
Nguyen Lee
LandOfFree
Method for detecting nonlinear distortion using moment... 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 for detecting nonlinear distortion using moment..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for detecting nonlinear distortion using moment... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2610713