Pulse or digital communications – Receivers – Angle modulation
Reexamination Certificate
1998-07-14
2001-08-28
Le, Amanda T. (Department: 2634)
Pulse or digital communications
Receivers
Angle modulation
C375S340000, C329S304000, C329S306000
Reexamination Certificate
active
06282248
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates generally to the demodulation of a digital signal. More particularly, it relates to an adaptive technique for demodulating a modulated digital signal having a widely variable baud rate.
2. Background of Related Art
High speed data transmission systems, e.g. modems, operate in general by modulating a high frequency carrier corresponding to a desired channel with a low frequency digital signal at a fixed baud rate. The modulated data signal is transmitted to a receiver, which demodulates the received high frequency modulated signal to recover digital symbols at the desired baud rate.
In such data transmission systems, the baud rate of the transmitting and receiving devices are generally fixed at a discrete level, and generally include a modulator and/or demodulator which operates at a fixed baud rate. Any fine adjustments which might be made in the baud rate of the receiver are typically made in the sampling rate of an analog-to-digital (A/D) converter sampling the incoming analog signal. However, these conventional baud rate adjustments are limited to just a few hundreds or thousands of parts per million of the baud rate, and are not able to adjust through a wide range of baud rates without requiring additional and/or different filtering for each different baud rate. The need for additional and/or different filtering for each received baud rate is cumbersome and expensive to implement. Moreover, different receivers must be developed and manufactured for each expected baud rate.
Many signal processing systems are hybrid analog/digital systems whereby front end signal processing is performed using analog circuit functions, and remaining signal processing is performed using digital computation after the intermediate analog signal has been converted to a digital number stream. Quadrature Amplitude Modulation (QAM) modem receivers are examples of hybrid systems typically implemented in this way.
A quadrature demodulator is often used within a QAM modem receiver to frequency-translate a bandpass spectral region of an intermediate frequency analog signal into two lowpass analog signals spectrally centered at or near 0 Hertz (Hz). Typically, the analog signals being demodulated have been band-limited to a specific intermediate frequency (IF) spectral region (fixed center frequency, fixed bandwidth equal the constant bandwidth channel span) using dedicated analog circuitry. The two outputs produced by the quadrature demodulator are baseband signals, having only low frequency spectral content, and exhibit a mathematical quadrature relationship with respect to each other.
A quadrature demodulator/analog-to-digital converter subsystem is used to transform a passband analog input signal into two digital number streams. The number streams produced by this subsystem approximate the two baseband analog signals produced by the quadrature demodulator defined above. Such quadrature demodulator/analog-to-digital converter subsystems are typically designed to handle only a specific channel bandwidth and to produce baseband streams of specific, fixed output rate. Such quadrature digital number streams are often used as input to digital signal processing (DSP) algorithms for communication, process control, estimation, and other signal transformation purposes, in addition to their application to QAM modem receivers.
It is generally recognized that there are system advantages to replacing the analog quadrature demodulation function described above with an equivalent digital quadrature demodulation set of operations. With this approach, A/D conversion is performed upon the single, fixed-bandwidth, bandpass IF analog aggregate channel input signal to the replaced analog quadrature demodulator rather than upon the replaced demodulator's two low pass output signals. In order for the resulting set of uniformly-spaced amplitude samples to retain all information contained within the original waveform (i.e., the original continuous time analog signal, using only this set of discrete samples, can be re-synthesized), the sampling rate (in Hz) used for this A/D conversion must generally satisfy three design constraints:
(1) the sampling rate f
s
must be greater than the bandwidth of the bandpass analog signal being processed;
(2) the resulting sampled waveform is high pass in spectral content (no DC component); and
(3) the resulting sampling rate f
s
sampling-produced spectral translates of the negative and positive components of the original bandpass analog signal do not exhibit any spectral image overlap.
Having satisfied these three sampling rate restrictions, the system designer is usually free to perform any additional required linear filtering operations upon the digital signal either (a) by passband digital filtering the real number stream produced by the A/D converter before digital quadrature demodulation, or (b) by mathematically-equivalent lowpass digital filtering a complex number stream (mathematical notation for filtering two real number streams) after digital quadrature demodulation.
Whereas both the passband or baseband signal processing approaches have near-equivalent implementation complexity when the signal to be processed is of fixed bandwidth and fixed location within the fixed bandwidth IF channel, the baseband signal processing approach proves more efficient for implementing systems which must isolate and demodulate arbitrary baud rate transmit signals located at arbitrary spectral positions within a fixed bandwidth, multi-channel aggregate analog signal. For such applications, an analog bandpass filter precedes the A/D converter and is used to remove out-of-aggregate-band energy from the analog signal to be passed to the A/D converter.
The preferred A/D sampling rate for a more conventional quadrature demodulator/A/D converter subsystem is an integral multiple of the subsystem output sample rate. Satisfying this constraint reduces complexity of the rate decimator algorithm that is otherwise generally considered necessary to handle more arbitrary input to output sample rate relationships.
There exists a need for a variable bandwidth channel tuner/variable baud rate QAM demodulator subsystem that operates at a fixed sampling rate and which can produce high spectral fidelity digital number streams at arbitrary output rates. Furthermore, there exists a need for a variable bandwidth channel tuner/variable baud rate QAM demodulator subsystem having fixed and efficient computational complexity, independent of the specific input sampling rate/output rate ratio. There also exists a need for a variable bandwidth channel tuner/variable baud rate QAM demodulator subsystem which enables a QAM modem receiver to feature a fixed sampling rate, yet handle signals over wide, continuously variable baud rate and spectral location spans. There is a further need for a receiver which is efficient in circuit usage and which can receive any of a wide range of baud rates, e.g., varying by a factor of 30 or more.
SUMMARY OF THE INVENTION
It is an aspect of the invention to provide a method of demodulating a signal that can be used to demodulate incoming signals having varying baud rates from a possibly wider bandwidth aggregate channel line signal. Such a method includes the step of sampling the signal at a first frequency. The sampled signal is then demodulated to create baseband samples at the first frequency. The baseband samples are then interpolated to produce samples at a second frequency. The second frequency is preferably determined to be a whole number multiple of twice a target baud rate. For example, the second frequency can be a power of two multiple of twice the target baud rate. The interpolated samples are then decimated to produce samples at a third frequency, which is preferably twice the target baud rate. The decimated samples are then equalized to output symbols at a fourth frequency which corresponds to the target baud rate. Based on this method, a demodulator can receive possibly wider band signals containing mu
Farrow Cecil William
Raghunath Kalavai Janardhan
Udovic Daniel
Agere Systems Guardian Corp.
Bollman William H.
Le Amanda T.
LandOfFree
Variable baud rate demodulator does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Variable baud rate demodulator, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Variable baud rate demodulator will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2445226