One bit digital quadrature vector modulator

Details One bit digital quadrature vector modulator One bit digital quadrature vector modulator

1998-08-17

2002-01-15

Ghayour, Mohammad H. (Department: 2734)

Pulse or digital communications

Pulse code modulation

Differential

C375S261000

Reexamination Certificate

active

06339621

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates to electronic signal processing and, in particular, to a digital quadrature vector modulator using a single bit delta-sigma modulation and a method for generating signals using the same.
BACKGROUND OF THE INVENTION
The invention applies to the fields of electrical engineering, electronics, communications engineering and signal processing. The quadrature modulation technique is applicable to virtually all quadrature modulation schemes, which include for example, Quadrature Amplitude Modulation (QAM), Quadrature Phase Shift Keying (QPSK), Quadrature Quadrature Amplitude Modulation (Q
3
AM), Orthogonal Frequency Division Modulation (OFDM) and many others schemes.
Analog Quadrature Vector Modulation (AQVM)
Quadrature Vector Modulation using analog techniques is currently in wide use in communications and in many other fields which require signal processing. Commercial AQVMs, such as that illustrated in
FIG. 1
, are available from suppliers such as Mini-Circuits Division of Scientific Components, Hewlett-Packard Co., Watkins Johnson Co., Analog Devices, Inc. and many others. These devices are useful for general signal processing applications and are often used to implement single side-band (SSB) radio frequency signal modulation. Typical input spectrum
20
and output spectrum
22
for SSB AQVMs are shown in FIG.
2
. The multiplier functions in AQVMs are often realized using diode mixers, field effect transistor (FET) based mixers, four quadrant multipliers such as Gilbert cells, or other analog multiplication techniques. The quadrature components for the input signal may be created using hybrid baluns if the bandwidth of the input signal is limited. In communications systems the input signals are often digital baseband signals, the quadrature components being generated digitally and subsequently converted to baseband analog signals before they are input to the AQVM.
In general a signal
24
, f
B,
to be input to an AVQM is split (by e.g. a splitter
2
shown in
FIG. 1
) into in-phase and quadrature components, I
B
and Q
B
respectively. The quadrature input signals may be generated using analog techniques, but, as noted above, are usually generated digitally and converted into analog prior to input to the vector modulator. As shown in
FIG. 1
, an analog modulating signal, often generated by a local oscillator (LO)
4
is also split by a splitter
6
into quadrature components, I
LO
and Q
LO.
The quadrature components and the in-phase components are multiplied by multipliers
8
and summed by a summer
10
to produce I
B
*I
LO
+Q
B
*Q
LO
. The resulting output creates upper and lower sideband products
28
and
30
centred about the local oscillator frequency
26
as shown in FIG.
2
. The lower sideband
28
is composed of in-phase I and Q products while the upper sideband
30
is composed of products of opposite phase. If the amplitude and phase matching of the analog modulators is very good, the upper sideband (USB) signal amplitude will be very small compared to the in-phase lower sideband (LSB) signal amplitude. The ratio
32
of the amplitude of the in-phase sideband to the out-of-phase sideband products is often referred to as the image rejection, image suppression or sideband suppression. The cancelled sideband is often referred to as the image.
In a similar fashion an upper sideband may be obtained by interchanging the I
LO
and Q
LO
products to obtain I
B
*Q
LO
+Q
B
*I
LO
.
It is desirable in many communications applications to conserve bandwidth by using only a single sideband of the modulation products. However, because the frequency offset of the image (i.e. the unwanted sideband) from the desired sideband signal, f
LO
−(f
LO
−f
B
), is often very small compared to the local oscillator frequency, f
LO
, it may be very difficult or impossible to remove the image frequency by means of a filter. For this reason the image rejection performance (i.e. the unwanted sideband suppression) of the AQVM is of critical importance to the overall performance of the modulation technique and can be a fundamental limiting factor in the use of the technique. The image rejection of AQVMs varies depending upon the application, but is typically between 15 and 45 dB. Rejection above 30 dB often requires special manual tuning and often varies with temperature and frequency. For these reasons AQVMs can be costly to implement in large volume applications.
The Shannon-Hartley theorem (
Modern Quadrature Amplitude Modulation
, Webb and Hanzo, IEEE Press 1995, p. 39) states that the capacity and maximum transmission rates of a communications channel are limited by the available carrier to noise ratio. The image level from digital quadrature modulators presents noise like interference to the modulation scheme and hence poses a fundamental limit to the level of modulation and transmission rates which can be achieved using these analog modulators.
Digital Quadrature Vector Modulation (DQVM)
As taught by the Nyquist Sampling Theorem, it has long been understood that communications signals may be fully represented by their digital equivalents (
Certain Factors Affecting Telegraph Speed,
H. Nyquist, Bell System Tech Journal, April 1928, pp. 617). It is also widely known that quadrature vector modulation can be accomplished using digital techniques. Digital signals, however suffer from noise caused by the quantization process. The signal-to-noise ratio (S/N) of a signal quantized to n bits and having an equal probability of existing at each of the 2
n
levels, is given by the relationship 20 log(n
2
−1) dB and increases by 6 dB per bit (
Introduction to Communication Systems,
F. G. Stremmler, Addison Wesley 1977 pp. 455).
Conventional digital modulators can be implemented in many forms. For example commercially available dedicated hardware multipliers, for instance, from Analog Devices Inc., can be used. Dedicated signal processing components such as the Texas Instruments TMS320 family of digital signal processor integrated circuits may also be used. It is also possible to implement a digital modulator in software using general purpose computers such as the Intel x86 family of processors. Conventional Digital Signal Processors (DSP) can achieve excellent image rejection due to the level of phase and amplitude matching which can be maintained in the digital process.
Conventional DSP techniques quantize both the baseband and the modulating signals to a sufficient number of bits to keep the noise to acceptable levels for the system application. Modern communications systems often require 10 bit quantization or higher.
Conventional DSP requires a number of multi-bit digital multiplications to be executed to complete the modulation function. These multiplications must be executed in real time. The circuits required to execute such multiplications utilize a large number of digital gates that consume a relatively large amount of power, and are limited in their maximum clock rates due to the size and complexity of the digital computations required. Because of the size of the integrated circuits involved, multi-bit multiplication circuits are relatively expensive and operate at slower clock speed than single bit digital circuits using the same technology. The use of multi-bit digital quadrature vector modulation is therefore often limited in its applicability to high speed communications systems because of cost, complexity, size, power requirements, and performance limitations of the circuits required by conventional techniques.
Delta Sigma (&Dgr;&Sgr;) Modulation
Over the past twenty years a number of authors have described &Dgr;&Sgr; modulators for use as Digital-to-Analog Converters (DACs) . See for example
Oversampling Delta-Sigma Data Converters,
Candy and Temes, IEEE Press (1992). Delta Sigma modulation is a method of achieving high signal-to-noise ratios over limited bandwidths using single bit signals modified by feedback. &Dgr;&Sgr; modulators require high sampling rates relative to the applied signal. To date &

Affiliated with

Also associated with

Rating

One bit digital quadrature vector modulator does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with One bit digital quadrature vector modulator, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and One bit digital quadrature vector modulator will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2859823