Telecommunications – Transmitter and receiver at same station – With frequency stabilization
Reexamination Certificate
2000-10-11
2003-09-30
Nguyen, Lee (Department: 2682)
Telecommunications
Transmitter and receiver at same station
With frequency stabilization
C455S192100, C375S326000, C375S344000
Reexamination Certificate
active
06628926
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a method of automatic frequency correction of a received radio signal in a digital mobile communications system.
2. Description of the Related Art
In a digital mobile communications system, such as a GSM (Global System For Mobile Communication) system, data is transferred, referring to
FIG. 1
, between a transmitter
10
and receiver
20
as a radio signal over a physical channel which may use frequency and/or time division multiplexing to create a sequence of radio frequency channels and time slots. Typically, every frequency band is divided into many time division multiple access (TDMA) frames, with eight users per frame, each user being allocated time to send a single burst of information. The precise frequencies used vary by country. The physical channel over which the signal is sent is unknown to receiver
20
.
FIG. 2
illustrates a representative structure of receiver
20
in which the burst
30
is received by antenna
35
, the signal is amplified by amplifier
40
, and passes through filters and mixers
50
. The signal is then converted from an analog to digital by A/D converter
60
.
In a GSM system, the data to be transferred, whether modulated by EDGE (Enhanced Data Rates for Global System For Mobile Communication Evolution), GMSK (Gaussian Minimum Shift Keying), or some other scheme, is formed into a burst containing a sequence of 156.25 complex symbols (each symbol having a real and imaginary part). 156.25 symbols is the number of symbols that, by definition, fit into a single timeslot for transmission by transmitter
10
.
A normal burst
100
in a GSM system, as shown in
FIG. 3
, is comprised of six components: a first “tail bits” field comprising three symbols
105
, a first set of 58 symbols of encrypted data
110
, a training sequence (or “pilot symbols”)
120
of 26 symbols in length and known as a mid-amble because it comes between two data fields, a second set of 58 symbols of encrypted data
130
, and a second “tail bits” field
135
comprising three symbols (not shown), and a guard period
137
which is empty and extends for a period equivalent to 8.25 symbols. The size of these fields shown in
FIG. 3
are not drawn to scale. In the GMSK modulation scheme, a symbol is equivalent to a bit so there are 148 bits in a burst. In the EDGE/8PSK modulation scheme, a symbol corresponds to three bits so there are 444 bits in a burst. See GSM 05.02 §5.2
Bursts
(
Digital Cellular Telecommunications System
(
Phase
2+);
Multiplexing and Multiple Access on the Radio Path; Version
8.1.0
Release
1999). The training symbols are known to the receiver in advance of receiving the burst; the data symbols are not known to the receiver in advance.
In a digital mobile communications system, the frequency of a burst as it is received as a signal by receiver
20
frequently varies from the burst that is sent by a certain amount of “frequency offset.”
FIG. 4
illustrates the effect of a frequency offset. A signal
140
having a real part (represented by the x-axis) and an imaginary part (represented by the y-axis) is transmitted from point I to be received at point II. As is shown with specific regard to point A-E, while the real part of the signal remains constant, the phase of the signal changes relatively radically with time (represented by the z-axis).
FIG. 5
illustrates the performance of a GSM system employing 8PSK modulation without any frequency correction which results in a high bit-error rate. On the x-axis is the signal-to-noise ratio in dB. The y-axis represents the bit-error rate. The curve labeled “0 Hz” shows the bit-error rate performance if the system is not disturbed by any frequency offset, as is the case in an ideal world. Clearly, in such a case, there is no need for frequency correction and the 0 Hz curve shows the best results that can be obtained in this system. The other four curves in
FIG. 5
show the bit-error rate of the same system (without any frequency correction) if the system is subject to frequency offsets. Clearly, the bit-error rate becomes high for a frequency offset of 100 Hz or higher. For example, a frequency difference of 100 Hz in a GSM system causes the phase of a signal to change by almost 20 degrees during a burst.
Where the transmitted burst is a constant signal, that is, the transmitter
10
does not intentionally vary the phase of the transmitted signal, the receiver
20
observes a time-varying phase during the receipt of the signal at the receiver because of an undesired frequency offset, which must be accounted for and corrected at the receiver end of the system. This undesired frequency offset has two causes. One cause is the oscillators in the transmitter
10
and receiver
20
that usually do not have exactly the same frequency, although ideally these frequencies should be the same. Another cause of the frequency offset is movement of the receiver
20
. As a receiver
20
moves toward the transmitter, the Doppler-effect will cause a frequency change. A speed of 100 km/h, for example, causes a frequency offset of about 100 Hz and a phase change of almost 20 degrees. Both causes of frequency offset can occur simultaneously and the effects can be cumulative.
The frequency offset presents a more significant problem in a communication system, particularly in a system, device or product which has a high sensitivity to frequency errors, such as the modulation scheme in the EDGE-GSM system where a phase change to the transmitted signal is intentionally introduced as a means of inserting desired information into the signal. The receiver
20
must be able to detect from the phase of the received signal the information bits that were transmitted. However, receiver
20
cannot distinguish between phase changes due to the transmitted information and phase changes introduced by the frequency offset. If receiver
20
ignores the presence of a frequency offset and treats the received signal as if there were no frequency offset, the transmitted message will be incorrectly detected at the receiver
20
, which would analyze and interpret the phase changes introduced by the frequency offset as if they were intentionally inserted by the transmitter
10
.
There must therefore be a method of accounting and compensating for the signal distortion caused by the frequency offset, especially in EDGE systems, where a cumulative phase rotation of as much as 40 degrees in our example can seriously affect performance. Because there is no way of measuring the frequency offset, the prior art describes a method of correcting the frequency offset by estimating the frequency offset at digital signal processor
70
(FIG.
2
), as described below, based on the training symbols
120
from the received burst, which training symbols the receiver
20
knows before the burst is received. The unknown frequency offset &phgr; of the received signal is modeled as a phase rotation e
j&phgr;k
introduced into the signal. Once the frequency offset
&phgr;
is determined, the offset
&phgr;
can be corrected and compensated for by back-rotating the signal over an angle
&phgr;
using a prior art digital signal processor structure represented by FIG.
6
.
The prior art estimator is based on the model of the received signal shown in
FIG. 7. A
burst x(k) is sent through a channel having impulse response (or “tap”) h(k) and received as signal z(k). The frequency offset
&phgr;
is a function of the channel impulse response h(k) which the signal is transmitted, but because h(k) is unknown by receiver
20
h(k) is not provided and must be estimated. In addition to the modeling of phase rotation e
j&phgr;k
, the model of received signal z(k) also accounts for additive white noise w(k), which represents impairments caused by fading, multipath propagation, channel dispersion, additive noise, etc., that are introduced into the burst x(k) by things such as physical objects in the path of propagating radio waves. The frequency offset
&phgr;
is thus modeled mathematically as:
z
(
k
)
Hedqvist Christer
Stigmer Par
van de Beek Jaap
Cohen & Pontani, Lieberman & Pavane
Nguyen Lee
Nokia Networks Oy
LandOfFree
Method for automatic frequency control 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 automatic frequency control, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for automatic frequency control will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3100565