 |
Description  |
|
|
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates generally to digital communication systems
and, more specifically, to a method and apparatus for adjusting
transmitter power in such systems both to minimize interference among
transmitters operating simultaneously and to maximize the quality of
individual communications.
2. Description of the Related Art
In a cellular telephone or personal communication system (PCS), a large
number of "mobile stations" communicate through cell sites or "base
stations.". The transmitted signal experiences multipath fading as the
mobile station moves in relation to features in the environment that
reflect the signal. Controlling mobile station transmitter power to
overcome multipath fading is described in U.S. Pat. No. 5,056,109, titled
"METHOD AND APPARATUS FOR CONTROLLING TRANSMISSION POWER IN A CDMA MOBILE
CELLULAR TELEPHONE SYSTEM," issued on Oct. 8, 1991 to the assignee of the
present invention and incorporated herein by reference.
If the mobile station transmits an excessively powerful signal, it will
interfere with the transmitted signals of other mobile stations. If the
mobile station transmits an insufficiently powerful signal, the base
station will be unable to recover the transmitted information from the
received signal. In the above-referenced patent, the base station measures
the power of the signal received from a mobile station and transmits power
adjustment commands to the mobile station over a separate channel. The
commands instruct the mobile station to increase or decrease transmission
power to maintain the average received signal power at a predetermined
level. The base station must periodically adjust the transmission power of
the mobile station to maintain an acceptable balance between interference
and signal quality as the mobile station moves.
The base station processor may monitor error rates in the received signal
to select an optimal power level at which to maintain the average received
signal. The base station processor detects errors as disclosed in
copending U.S. Patent application Ser. No. 08/079,196, titled "METHOD AND
APPARATUS FOR DETERMINING DATA RATE OF TRANSMITTED VARIABLE RATE DATA IN A
COMMUNICATIONS RECEIVER," and assigned to the assignee of the present
invention. In the exemplary CDMA cellular telephone system described in
the above-referenced U.S. patent and copending application, the mobile
station transmits "frames" comprising "symbols," which represent digitized
voice or other data. Further details on the exemplary CDMA cellular
telephone system are described in U.S. Pat. No. 5,103,459, titled "SYSTEM
AND METHOD FOR GENERATING SIGNAL WAVEFORMS IN A CDMA CELLULAR TELEPHONE
SYSTEM," issued Apr. 17, 1992 to the assignee of the present invention and
incorporated herein by reference.
The mobile station encodes frames at one of four rates; the rate is
selected according to the needs of the user. The maximum rate, which is
generally preferred for high quality voice transmissions or rapid data
transmissions, is called "full rate." Rates of one half, one fourth, and
one eighth of the full rate are called "half rate," "quarter rate," and
"eighth rate," respectively. Each symbol of a frame to be encoded at half
rate, quarter rate, and eighth rate is repeated two, four, and eight
times, respectively, to fill the frame. The frame is then transmitted to
the base station at a constant rate, regardless of the rate at which the
symbols are encoded.
The base station has no advance notice of the data rate at which a received
frame is encoded and the rate may be different from that of the previous
received frame. The base station decodes each received frame at each of
the four rates and produces a set of error metrics corresponding to each
rate. The error metrics provide an indication of the quality of the
received frame and may include a cyclic redundancy check (CRC) result, a
Yamamoto Quality Metric, and a re-encoded symbol comparison result. The
generation and use of these error metrics are well known in the art with
details on the Yamamoto Quality Metric provided in the article "Viterbi
Decoding Algorithm for Convolutional Codes with Repeat Request", Hirosuke
Yamamoto et al., IEEE Transactions on Information Theory, Vol. IT-26, No.
5, September 1980. The set of error metrics for the decoding of each frame
at each rate thus includes one or more of the CRC result, the Yamamoto
Quality Metric, and the re-encoded symbol comparison result. The base
station processor analyzes the sets of error metrics using a novel
decision algorithm and determines the most probable rate at which the
received frame was encoded. The base station then uses the rate decision
to select the corresponding decoded data from the multiple data rate
decodings to recover the transmitted frame information.
The base station processor also produces an "erasure" indication if the
quality of the frame data is too poor for the processor to determine the
rate. Similarly, the processor produces a "full rate likely" indication if
bit errors exist in the data but the rate is probably full rate. If an
erasure occurs, the base station may simply discard the frame or may
replace it with interpolated data.
It would be desirable to monitor the error rate of the received frames and
to periodically adjust the transmission power level to maintain the error
rate at an acceptable value. These problems and deficiencies are clearly
felt in the art and are solved by the present invention in the manner
described below.
SUMMARY OF THE INVENTION
The present invention comprises a method and apparatus for adjusting the
power level of a remote transmitter to provide a substantially constant
error rate in the received data. The present invention may be used in the
base station of a cellular telephone system to maximize the number of
mobile stations that may transmit simultaneously with minimal interference
by enhancing control over the power of the signal that each mobile station
transmits.
In the CDMA cellular telephone system described in the above-referenced
U.S. patent, the mobile station transmits a signal comprising frames of
digitized voice or other information to the base station at an initial
power level or setpoint. As described in the above-referenced copending
application, the information is encoded into either full rate, half rate,
quarter rate, or eighth rate data frames. The base station receives the
signal and decodes each frame at each of these rates. A corresponding set
of error metrics is produced for each rate that provides an indication of
the quality of the received information if the frame is decoded at that
rate. The base station processor then analyzes the sets of error metrics
using a decision algorithm and either provides an indication of the most
probable rate at which the information was encoded or provides an
"erasure" indication, i.e., an indication that the rate could not be
determined with the desired probability of correctness.
In the present invention, the base station processor counts the number of
consecutive frames encoded at a rate such as full rate and the number of
frames that are erasures. A count-of a predetermined number of consecutive
full rate indications, i.e., without an intervening less than full rate
indication, erasure indication or full rate likely indication, is
indicative of a high quality full rate transmission and is called a "full
rate run." If the processor detects a full rate run and then detects an
additional full rate frame, it should decrease the signal power to a level
at which a small but acceptable number of erasure or full rate likely
indications occur between the full rate frames. For example, one error
indication in 100 full rate frames, where each frame consists of 576
symbols and is transmitted at a rate of 28,800 symbols per second, is
inaudible in a transmission consisting of ordinary speech.
A count of a predetermined number of consecutive erasure indications, i.e.,
without an intervening other rate indication, is indicative of a poor
quality transmission and is called an "erasure run." If the processor
detects an erasure run, it should increase the signal power. The increased
signal power may overcome multipath fading, thereby reducing the erasure
rate.
A predetermined consecutive number of half rate, quarter rate, or eighth
rate indications is called a "variable rate run." As a further enhancement
in controlling transmitter power the processor may, while in a variable
rate run, also reduce the signal power if it detects a half rate, quarter
rate, or eighth rate indication. In addition while in the variable rate
run, the processor may increase the signal power if it detects an erasure
indication.
Although the present invention may be used to adjust the power level of
transmissions consisting of any type of data, it is optimized for
transmissions consisting of voice information. In communications systems
such as the cellular telephone system described in the above-referenced
copending application and U.S. patent, voice transmissions are encoded at
a variable rate; the complexity of the speech determines the rate.
However, continuous speech is generally encoded at full rate. Speech
occurring after a period of relative inactivity may be encoded at lower
rates, transitioning to full rate as the speech increases in complexity.
The algorithm thus expects to detect variable rate runs alternating with
full rate runs as the speaker pauses between words or syllables.
Therefore, the processor may also increase the signal power if it detects
an erasure indication or a full rate likely indication following a full
rate run. The increment by which the processor increases the power upon
detecting an erasure or full rate likely indication following a full rate
run need not be the same as the increment by which the processor increases
the power upon detecting an erasure run.
The foregoing, together with other features and advantages of the present
invention, will become more apparent when referring to the following
specification, claims, and accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
The features, objects, and advantages of the present invention will become
more apparent from the detailed description set forth below when taken in
conjunction with the drawings in which like reference characters identify
correspondingly throughout and wherein:
FIG. 1 is a block diagram showing the present invention in the base station
receiver of a cellular telephone system;
FIG. 2 is a generalized flow diagram of an exemplary power control setpoint
algorithm; and
FIGS. 3a-3c illustrate a detailed flow diagram of an exemplary power
control setpoint algorithm for a determined rate decision pattern.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
In a CDMA cellular communication system where system user capacity is a
function of the total system power, any reduction of mobile station power
facilitates an increase in system capacity. The present invention provides
a method and system for closely and dynamically controlling the mobile
station transmitter power as a function of the communication link. Through
dynamic control over mobile station transmitter power greater system
capacity may be achieved.
In FIG. 1, the present invention is used in a base station receiver of a
CDMA cellular telephone system. This receiver is described in the
above-referenced U.S. Patent and is now described only briefly. A mobile
station (not shown) transmits a communication signal, typically a CDMA
signal of a spreading bandwidth for example of 1.25 MHz at one frequency
band, to the base station radio receiver (not shown).
In order to aid in understanding of the present invention, a brief
discussion of the mobile station data encoding for transmission is
provided. In the exemplary embodiment user data provided at various data
rates is encoded and formatted for transmission in data frames typically
20 milliseconds in length. The user data along with frame overhead data
are preferably forward error correction encoded the effective data rates
for this example are 9.6 kbps (full rate), 4.8 kbps (half rate), 2.4 kbps
(quarter rate) and 1.2 kbps (half rate). It should be noted that a
constant symbol rate for the frames is preferred but is not necessary.
In this example rate 1/3 convolutional encoding is used to produce three
symbols for each user data or frame overhead bits. For a full rate frame,
corresponding to a 9.6 kbps data rate, a total of 192 user data and frame
overhead bits are encoded to produce 576 symbols for the frame. For a half
rate data frame, corresponding to a 4.8 kbps data rate, a total of 96 user
data and frame overhead bits are encoded to produce 288 symbols for the
frame. Similarly for quarter rate and eighth rate data frames,
respectively corresponding to 2.4 and 1.2 kbps data rates, a total of 48
and 24 user data and frame overhead bits are encoded to produce 144 and 72
symbols for the respective rate frame. It should be noted that groups of
symbols are converted into a respective orthogonal function sequence or
code of a set of orthogonal function codes according to the value of the
symbol set. In the exemplary embodiment six symbols for a binary value
that is used to select one of sixty-four Walsh function sequences each
sixty-four chips in length. Further details on this modulation scheme is
disclosed in the above mentioned U.S. Pat. No. 5,103,459.
At the base station the signal is received at antenna 100 and provided to
receiver 102 for frequency downconversion and filtering. Analog-to-digital
(A/D) converter 104 receives the analog spread spectrum signal from
receiver 102 and converts it to a digital signal. A pseudorandom noise
(PN) correlator 106 receives the digital signal and a PN code provided by
a PN generator 108. PN correlator 106 performs a correlation process and
provides an output to a Fast Hadamard Transform digital processor or
filter 110.
In a preferred embodiment of a multipath diversity receiver PN generator
108 generates a plurality of a same PN codes with timing offsets dependent
upon the particular path of the signal. PN correlator 106 correlates each
of the PN codes with a respective path signal to produce a respective
orthogonal function symbol data. Filter 110 converts the orthogonal
function symbol data into soft decision symbol data for each multipath
signal. The multipath symbol data is then combined and provided as soft
decision symbol data for decoding by user data decoder 112.
Filter 110 as part of the conversion process determines from each
orthogonal function symbol from each multipath signal an energy value.
Keeping in mind that each orthogonal function symbol is converted into a
group of data symbols, the energy values from the different paths are
combined to produce a corresponding symbol energy value. Filter 110 in
addition to providing soft decision data to decoder 112, also provides the
symbol energy value to power averager circuit 114.
Decoder 112, which typically includes a Viterbi decoder, receives the
filter soft decision symbol data output and produces user data and decoder
error metrics which are provided to rate determination processor 116.
Processor 116 may send the user data to a digital-to-analog converter or
other output circuitry (not shown). Decoder 112 is described in further
detail in the above-referenced copending U.S. Patent application and is
only briefly described herein.
Upon reception at the base station, decoder 112 decodes each frame at each
possible rate and provides a corresponding set of error metrics
representative of the quality of the symbols as decoded at each rate.
Error metrics for decodings at each rate include, for example, a symbol
error result based upon a re-encoding of the decoded bits to produce
re-encoded symbols that are and then compared with the received symbols
and a Yamamoto Quality metric. In addition, for full rate and half rate
frames a CRC check result is performed on CRC bits in the frame overhead
bits.
After decoder 112 has decoded each frame, processor 116 executes the rate
determination algorithm described in the above-referenced copending U.S.
Patent application to determine the most likely rate at which the frame
was encoded. The algorithm uses the error metrics provided by decoder 112
to estimate or decide the rate at which the frame of data was transmitted.
Once processor 116 determines the rate for the frame of data, the data is
interpreted by control bits included in the frame as either control or
user data with the user data output for further use. From the error
metrics processor 116 determines whether the received data frame contained
data that was transmitted at either full rate, half rate, quarter rate or
eighth rate and generates a corresponding rate indication. This rate
indication is provided to outer loop power control processor 118, whose
function is described in further detail later herein.
In the case where the error metrics provided by decoder 112 indicate to
processor 116 that the received frame was corrupted beyond that which the
error correction techniques employed by decoder 112 may correct, processor
116 does not decide the rate of the data for the frame. Processor 116 in
this case does not use or provide an output of the data for that frame,
with the frame being considered an erasure frame. Processor 116 for the
erasure frame, generates and provides an erasure indication to processor
118 indicative that could not determine the rate at which the frame was
encoded.
In the case where the error metrics provided by decoder 112 indicate to
processor 116 that the received frame is a corrupted full rate frame that
was corrected by decoder 112. Typically in this case the metrics indicate
only that an error exists in the CRC. From this information processor 116
determines that the most likely the rate of the data for the frame is that
of full rate, and identifies the frame as a full rate likely frame.
Processor 116 uses or outputs the data as if it were full rate data with a
conditional understanding that it may contain errors. Processor 116 for
the full rate likely frame generates and provides a full rate likely
indication to processor 118.
The rate decisions and detected frame errors may be used as an indication
of the power level at which the mobile station need transmit signals at to
maintain a quality communication link. In those cases where a number of
frames are received at a rate or rates in which the occurrence of frames
in error is low, the mobile station transmitter power may be reduced. This
transmitter power reduction may continue until the error rate begins to
rise to a level which may adversely affect the quality of the
communication link. Similarly the power may be increased where the errors
adversely affect the quality of the communication link.
Upon receiving the rate indications from processor 116, processor 118
executes a novel algorithm to control a power level setpoint. This
setpoint is used as discussed with reference to FIG. 1 in generating power
commands which control the power of the mobile station transmitter power.
As mentioned previously filter 110 provides the scaled symbol energy value
to power averager 114. Power averager 114 sums or averages the scaled
symbol energy values over a 1.25 millisecond interval, i.e. corresponding
to a group of six Walsh symbols or thirty-six data symbols, and provides a
received power level signal to comparator 120.
Processor 118, which includes appropriate internal counters, program memory
and data memory, computes under program control a power level setpoint
signal as described below and provides it to comparator 120. Processor 118
may be either located at the base station through which the mobile station
is in communication with or at a remote location such as the mobile
telephone switching office (not shown). In the situation where the mobile
station is communicating through multiple base stations, with power
control provided through the multiple base stations, from a control
standpoint the location of processor 118 at the MTSO is more convenient.
In those situations where processors 116 and 118 are located together the
function of these two processors may be combined into a single processor.
Comparator 120 compares the received power level signal and the power level
setpoint signal, and provides a deviation signal representative of the
deviation of the received power from the power level setpoint set by
processor 118. Power up/down command generator 122 receives the deviation
signal and generates either a power up command or a power down command,
which the base station transmits to the mobile station (not shown). Should
the signal from power averager circuit 114 fall below the threshold
established by the power level setpoint signal, the deviation signal
generated by comparator results in the generation of power up command.
Similarly, should the power averager circuit signal exceed the power level
setpoint signal, a power down command is generated. These power commands
are provided to transmitter 124 where inserted into the data being
transmitted to the mobile station. Transmitter spread spectrum modulates
and transmits the modulated data via antenna 100 to the mobile station.
Transmitter 124 typically transmits the CDMA signal in a different
frequency band than the mobile station transmission but of the same
spreading bandwidth, e.g. 1.25 MHz.
FIG. 2 illustrates a generalized flow diagram of this algorithm used to
dynamically adjust the power level setpoint, and thus indirectly modify
the mobile station transmitter power. The implementation of the algorithm
seeks to effect a reduction or increase in the mobile station transmitter
power as a function of the link quality with respect to various frame rate
data. In this implementation a pattern of rate decisions is used to modify
the power level setpoint. Although the exemplary embodiment is described
with reference to using the rate decision as an indicator of patterns,
other parameters may be used.
In FIG. 2, a group of one or more of frame rate decisions is provided for
inspection, step 150. This group may be comprised of a collection of
sequential frame rate decisions, or according to some other order, and/or
which may be dependent upon the frame rate. The group of rate decisions
are inspected to determine if their pattern is matched to predetermined
rate decision pattern P.sub.1, step 152. If there is a pattern match, a
modification in the power level setpoint is made, step 154. This
modification may be in the form of an increase or decrease in the power
level setpoint by an incremental value. This increase or decrease in the
power level setpoint ultimately results in a corresponding increase or
decrease in the mobile station transmitter power. In those cases where a
rate decision pattern match indicates a good communication link, the power
le | | |
