Pulse or digital communications – Spread spectrum – Direct sequence
Reexamination Certificate
2000-07-07
2004-04-27
Tran, Khai (Department: 2631)
Pulse or digital communications
Spread spectrum
Direct sequence
C375S130000
Reexamination Certificate
active
06728301
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates generally to the field of direct sequence spread spectrum (DSSS) communications and, more particularly, to a system and method for controlling a voltage controlled oscillator in the down-conversion of a frequency in a code division multiple access (CDMA) wireless receiver.
2. Description of the Related Art
In DSSS communications, such as CDMA systems, pseudorandom noise (PN) sequences are used to generate spread spectrum signals by increasing the bandwidth (i.e., spreading) of a baseband signal. A forward link waveform transmitted by the base station may be comprised of a pilot waveform and a data waveform. Both of the waveforms are received with the same relative phase and amplitude distortions introduced by the channel. The pilot waveform is an unmodulated PN sequence which aids in the demodulation process, as is well-known in the art as “pilot-aided demodulation.” Conventional pilot-aided demodulation methods typically include the steps of (i) demodulating the pilot waveform, (ii) estimating the relative phase and amplitude of the pilot waveform, (iii) correcting the phase of the data waveform using the estimated phase of the pilot waveform, and (iv) adjusting the weight of data symbols used in maximal ratio combining in a RAKE receiver based on the estimated amplitude of the pilot waveform. Steps (iii) and (iv) above are performed as a “dot product” as is known in the art. In some conventional methods, a controller having a central processing unit (CPU) and and/or a digital signal processor (DSP) performs each step described, including the dot product function.
FIG. 1
illustrates a conventional IS-95 forward link base station transmitter multiplexing section
10
(prior art). A pilot channel
12
is generated that has no data. That is, the data is predetermined to be all “0” bits. The pilot channel is modulated, or covered with a Walsh code from Walsh code generator
14
at 1.2288 Mcps (megachips per second). Sixty four orthogonal Walsh codes, each of 64 bits, are used in the IS-95A and 95B systems. The IS-2000 standard uses 192 Walsh-like codes. Each channel is modulated with a unique Walsh code. Walsh code H
0
is used to modulate the pilot channel.
Also depicted is a traffic or paging channel, which shall be referred to herein as an information channel. Data is input at one of a plurality of data rates from 9.6 kbps (kilobits per second)to 1.2 kbps. The data is encoded at encoder
16
, one bit per two code symbols, so that the output of the encoder
16
varies from 19.2 ksps (kilosymbols per second) to 2.4 ksps. Symbol repetition device
18
repeats the codes from 1 to 8 times to create a 19.2 ksps signal. Additional data rates are used in other CDMA standards. Alternately stated, either 1, 2, 4, or 8 modulation symbols are created per code symbol. Then, the information channel is scrambled with a long code at the same 19.2 ksps rate. The information channel is covered with a different Wash code from that used to cover the pilot channel, code HT for example.
After being modulated with Walsh codes, each channel is spread with a common short code, or PN sequence. Each channel is split into I and Q channels, and spread with I and Q channel PN sequences. A 90 degree phase shift is introduced by multiplying the I channels with a sin function, while the Q channel is being multiplied with a corresponding cosine function. Then, the I and Q channels are summed into a QPSK channel. In the IS-95A standard, the same baseband symbols are assigned to an I and Q quadrature. The combination of all the QPSK channels, including pilot, synchronization, paging, and traffic channels can be considered a sample stream. As a final step, the sample stream is frequency up-converted at mixer
20
with a local oscillator signal, having a local oscillator frequency, and transmitted.
FIG. 2
is a conventional IS-95 CDMA receiver (prior art). At the mobile station receiver
50
the transmitted signals are accepted as analog information, down-converted in frequency to baseband, converted into a digital sample stream at A/D
52
, split into I and Q channels, multiplied respectively by sin and cosine functions to remove the
90
phase shift. Conventionally, a multi-finger RAKE receiver is used to resolve multipath variations in the sample stream, so that degradation due to fading can be minimized. Three demodulation fingers, demodulation finger
1
(
54
), demodulation finger
2
(
56
), and demodulation finger
3
(
58
) all receive the same I and Q sample stream. Each demodulation finger is assigned one of the sample stream multipath variations. PN codes and Walsh codes are generated with a delay consistent with the multipath delay of the sample stream to be demodulated. The sample stream from the multipaths is coherently combined in combiner
60
based on a maximal ratio combining (MRC) principle.
To effectively demodulate phase modulated signals, it is important to remove phase errors between communicating transmitters and receivers. The voltage controlled oscillator devices have inherent frequency accuracy tolerance, and tolerances due to temperature variations, which do not permit a precision open loop LO signal to be used. Further, multipath and Doppler shifts due to vehicle speed can introduce frequency errors, or frequency offsets from the transmitted frequency, into the signal as received. For example, temperature can introduce errors of +/−2 parts per million (PPM), aging can introduce +/−1 PPM per year, the process of soldering (application of IR energy) can introduce errors of +/−1 PPM, initial frequency tolerance can introduce +/−1 PPM, power supply variations can introduce +/−0.3 PPM, load variations can introduce +/−0.2 PPM, and Doppler can introduce up to +/−0.15 PPM errors into an output frequency of 2 gigahertz. Over a 3 year product life, and carrier frequency of 2 gigahertz, a +/−15 kilohertz error can result.
Thus, the LO frequency driving mixer
51
must be tracked with respect to the actual received frequency. To accomplish this frequency tracking, phase changes are measured in the demodulated pilot symbols, as these symbols have a predetermined value.
Conventionally, a frequency discriminator multiplies a present pilot symbol by previous pilot sample to convert the frequency error, the difference between the carrier signal and the LO, into a phase error. To simplify the operation, only the negative part of the imaginary part of the product is typically used for the phase error. The phase data can be used to aid in bit decisions at the demodulation finger, or used as the input to an automatic frequency control (AFC) loop to actually null out the error.
In an AFC loop the problem is further complicated by the fact that the carrier signal is received in a plurality of delays associated with multipath. Each version, or carrier signal delay may have a slightly different center frequency, or frequency offset from the other carrier signal delays. Therefore, a correction based upon one carrier signal delay is not optimum for the others. To address this problem, the phase errors of each carrier signal delay (each multipath) are combined in a maximal-ratio combining process that weights stronger versions of carrier signal delays over more weakly received versions.
FIG. 3
is a schematic block diagram illustration of a conventional automatic frequency control system
70
(prior art). Such a system would be used to track frequency drift in the voltage controlled oscillator
72
associated with the local oscillator. It should be noted that one AFC discriminator is provided for each demodulator finger. The correction to the oscillator
72
is based on some weighted sum of locked fingers. Consequently, the error measurement E(t) is a function of an equally weighted sum of all locked demodulator fingers as follows:
E
(
t
)={function of sum of
e
i
(
t
)}
Where i=1,2,3 . . . n (demodulator finger).
The input
Brady III Wade James
Neerings Ronald O.
Telecky , Jr. Frederick J.
Texas Instruments Incorporated
Tran Khai
LandOfFree
System and method for automatic frequency control in spread... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method for automatic frequency control in spread..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for automatic frequency control in spread... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3240809