HK1056015B - Method and apparatus for compensating local oscillator frequency error - Google Patents

Description

Method and apparatus for compensating for local oscillator frequency error

Background

I. Field of the invention

The present invention relates to electronic circuits. More particularly, the present invention relates to a novel and improved method and apparatus for compensating Local Oscillator (LO) frequency error by discriminating the LO frequency over time.

Description of the related Art

Accurate frequency sources are extremely important for the operation of many electronic systems and devices. The frequency source is used as a timer within the electronic device and also as a Local Oscillator (LO) to tune the electronic device to the desired communication channel.

Many types of precision frequency sources are available. The particular type of frequency source implemented in a particular application is determined by the design constraints of the particular application. Atomic clocks exhibit extreme levels of frequency accuracy, however, their size, cost, and lack of tuning range greatly limit their practical application in electronic systems. Similarly, the piezoelectric effect of quartz crystals can be used to design accurate frequency sources. The small size of the frequency sources based on the relative accuracy of quartz crystals makes them popular for most consumer-based electronic devices.

The application determines the type and frequency accuracy required for the frequency source. Receivers for Global Positioning System (GPS) applications require LOs with a high level of frequency accuracy in order to quickly acquire and maintain synchronization with signals provided on the GPS carrier frequency transmitted from the satellites. The overview of GPS helps to account for the need for LO frequency accuracy in GPS receivers.

GPS is generally used for positioning. GPS utilizes geometric principles to accomplish the determination of position. A set of GPS satellites orbits the earth. The receiver can determine its exact location by knowing the position of the satellites and calculating the distance from the receiver to each of the plurality of satellites.

A GPS receiver calculates the distance from a satellite to the receiver by determining the time it takes for a signal transmitted by the satellite to reach the receiver. Once the receiver determines its distance from the satellite, it knows that it is located on a locus equidistant from the satellite. The satellite acts as a point source and the locus of points equidistant from a point is a sphere. When the receiver determines its distance from the second satellite, it knows that its position is located somewhere on the second sphere. However, knowing the distance from two satellites greatly simplifies its possible location. This is because the receiver is located somewhere on the intersection of the two spheres. The intersection of the two spherical surfaces is a circle. Thus, the receiver knows that its position is on the intersection circle. Determining the distance of the receiver from the third satellite results in a third sphere. The third spherical surface intersects the first two spherical surfaces and also intersects the elements defining the intersection of the first and second spherical surfaces. The three spherical intersections result in two different points at which the receiver may be located. Once the two points resulting from the intersection of the three spheres are determined, the receiver can estimate which of the two points is the correct location or the receiver can determine its distance from the fourth satellite.

Once the distances from the three satellites are determined, the receiver can estimate which of the two points is its correct location. This is possible because one of the two points is not a possible location. The correct one of the two points may be near the earth's surface, while the incorrect point may be very far from the earth's surface or deep within the earth's surface. If the distance from the fourth satellite is determined, the exact location of the receiver is known. The exact position can be known with the fourth satellite because the intersection of four spheres can only produce one point.

A major problem in GPS implementations is the accurate determination of the distance from the satellite to the receiver. The distance from the satellite to the receiver is calculated by measuring the time of arrival at the receiver of the signal transmitted from the satellite. Each satellite transmits two carrier frequencies, each modulated with a unique pseudorandom code. One of the carrier frequencies operates at 1575.42MHz and the other carrier frequency operates at 1227.60 MHz. The receiver demodulates the received signal to extract the pseudo-random code. The locally generated pseudo random code is synchronized with the demodulated pseudo random code. The delay between the two pseudo-random codes represents the time of arrival of the transmitted signal. The distance from the satellite is then determined by multiplying the time of arrival by the speed of light.

All transmitting satellites are time synchronized. However, mobile receivers are only weakly synchronized to satellites. The weak time synchronization of the receiver with the satellites introduces errors in the positioning. As described above, different times of arrival correspond to different distances. The locus of points equidistant from a point is a sphere of radius equal to that distance. However, if only the time of arrival is known to fall within a time range, i.e. the measured time plus or minus some error, then only the distance is known to fall within a range of values. Knowing that the distance falls within a range of values, the locus of points equidistant from the source is the spherical shell. The thickness of the spherical shell is equal to the error of the distance measurement. Three spherical shells (each corresponding to a position estimated from an additional satellite) produce two volumes, one of which represents the position of the receiver. Recall that in the case of discrete distances, the intersection of three spheres results in two points, rather than two volumes.

The time synchronization problem may be partially solved by including a distance measurement from a fourth satellite. First, the time error is assigned to an assumed value, even zero. The distances from the three satellites are then determined. As previously described, the intersection of the three spheres defined by the three range measurements results in two different points, one of which is the position of the receiver. The distance from the fourth satellite defines a fourth sphere. Ideally, the fourth sphere meets only one point with the other three spheres, without timing error. However, the four spheres do not intersect when there is timing error. There is no timing error between satellites. Thus, the timing error from the receiver to one satellite is the same as the timing error from the receiver to any satellite in the satellite group. The timing error may be determined by adjusting the value of the hypothesized timing error. The timing error is determined when the four spheres intersect at a single point.

The solution of timing error is only one of the problems that must be dealt with when using GPS to achieve position location. A receiver for GPS positioning must be implemented in a small physical size at a relatively low cost. When implementing GPS receivers in consumer oriented devices, size and cost constraints become more important. New requirements for wireless telephones include the ability to determine the location of a caller. In the case of emergency calls, such as 911 calls in the united states, the particular location of the radiotelephone is important. Although there are still physical design constraints, the receiver must search for and acquire satellite signals quickly.

The receiver design must balance cost, received signal sensitivity, and search time. The receiver design cannot simultaneously maximize all parameters. Significant improvements in receiver sensitivity or search time result in increased receiver cost.

A major component of the complexity associated with searching for and acquiring satellite signals is the frequency error caused by the receiver Local Oscillator (LO). The LO is used in the receiver to downconvert the received signal to a baseband signal. The baseband signal is then processed. In the case of signals received from GPS satellites, the baseband signal is correlated with all possible pseudo-random codes to determine which satellite originated the signal and the time of arrival of the signal. LO frequency errors greatly complicate the search and acquisition process. Any LO-induced frequency error creates additional search space that must be covered. Furthermore, the LO frequency error represents a separate dimension above which the arrival time must be searched. In this way the search space is increased in proportion to the frequency error, since the search for the time of arrival has to be conducted over all possible frequency errors.

Many parameters cause real or perceived LO frequency errors. Both the circuit operating temperature and the temperature gradient on the circuit board affect the LO frequency. Furthermore, the frequency stability for the LO frequency reference directly affects the LO frequency stability. An additional component to the frequency error is the doppler shift effect caused by the receiver velocity. Even in case the receiver LO is very accurate, there may be a perceived frequency error due to doppler shift effects. The frequency shift can cause a significant increase and decrease in the satellite transmission frequency. Although both the satellite and the receiving LO may be quite stable, the signal at the receiver may be subject to frequency shifts. Doppler shift within the receiver due to receiver motion is uncorrectable and only causes frequency errors that already exist in the receiver.

What is needed is a way to reduce LO frequency error to reduce the search space that must be covered in baseband signal processing. The reduction in search space results in lower search complexity, thereby resulting in higher receiver sensitivity and reduced search and acquisition times.

Summary of The Invention

The present invention is a novel and improved method and apparatus for reducing the frequency error of a Local Oscillator (LO) by identifying the LO over a number of operating conditions and compensating the LO based on the operating conditions.

When operating in the phone mode, an external frequency source with little frequency uncertainty is provided to the receiver. The receiver uses an external frequency source as a frequency reference. The receiver estimates the frequency error of the LO using an external frequency source as a frequency reference. Simultaneously with the frequency estimation, the receiver monitors different predetermined parameters (which are known to have an impact on LO accuracy and frequency stability). Operating temperature and temperature gradients along the board are examples of parameters that affect LO accuracy. The monitored parameter values and LO frequency are stored in a memory location. In addition, the frequency error may be stored in a table. This provides a series of tables for identifying the LO.

The LO may be converted to a higher accuracy GPS mode in which the LO output frequency is controlled to achieve a lower frequency error. In the higher accuracy GPS mode, the receiver no longer uses an external frequency source. The receiver continues to monitor predetermined parameters for discrimination of the LO. The receiver then uses the current results of the monitored parameters and compares them to the values in the previously stored table. An estimate of the LO error is then made based on a comparison of the current parameter and the stored parameter measurements. The LO is then compensated based on the previous discrimination to correct the estimated error.

In an alternative embodiment, the frequency error is reported to the receiver, which may simplify the signal acquisition process. In another embodiment, the LO operates in a higher accuracy mode, in which the output of the LO is compensated for frequency error and the frequency error is also reported to the receiver. In yet another embodiment, the phone mode that discriminates frequency errors operates simultaneously with the GPS mode.

Brief Description of 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 of a receiver;

FIG. 2 is a block diagram of a local oscillator;

FIG. 3 is a schematic diagram illustrating a search space;

fig. 4 is a block diagram of a receiver implementing LO discrimination;

fig. 5 is a block diagram of an alternative embodiment of a receiver implementing LO discrimination;

6A-6B are flow diagrams of a LO discrimination process; and

fig. 7 is a flow chart of the LO compensation process.

Detailed description of the preferred embodiments

Fig. 1 is a block diagram of a generic receiver 100. The antenna 102 serves as an interface between the broadcast signal and the receiver 100. The antenna is tuned to optimally receive the signals transmitted in the L-band, wherein the receiver is configured as a GPS receiver. In the case of a GPS receiver, the broadcast signal source is a set of GPS satellites orbiting the earth. The signal received by the antenna 102 is coupled to the downconverter 110. A downconverter 110 downconverts the RF signal received by the antenna 102 to a further processed baseband signal. The main components of downconverter 110 are mixer 112 and Local Oscillator (LO) 114. The down-converter 110 may also include filters and amplifiers (not shown) to maximize the quality of the final baseband signal. The received signal within the downconverter 110 is coupled from the antenna 102 to a mixer 112. Any filtering and amplification of the signals within the down-converter 110 is not shown in order to simplify the block diagram to its functional components. The mixer 112 serves to efficiently multiplex the received signal with the LO 114 signal. The final signal output from the mixer 112 is centered at two primary frequencies. One frequency component of the output of mixer 112 is centered on the sum of the center frequency of the received signal and the operating frequency of LO 114. The second frequency component output by mixer 112 is centered on the difference between the received signal center frequency and the LO 114 operating frequency. When the received signal is quadrature modulated, two mixers 112 and 113 are used in the down-converter 110. The accepted signal is used as input to two mixers 112 and 113. The second input on the first mixer 112 is the LO 110 signal. The second input on the second mixer 113 is the LO 114 signal shifted by ninety degrees in a phase shifter (not shown). The final output of the first mixer 112 is labeled incident phase output (I) and the final output of the second mixer 113 is labeled quadrature phase output (Q).

The outputs I and Q from the downconverter 110 are coupled to filters 122 and 124, respectively, which are used to remove unwanted frequency components from the mixer 112 and 113 and to pre-process the downconverted signal prior to subsequent signal processing.

The filtered I and Q signals are coupled to a correlator bank 130. The correlator 130 processes the I and Q signals using digital signal processing techniques. The correlator digitizes the I and Q signals in an analog-to-digital converter (ADC) for digital signal processing. Correlator 130 is used to determine the phase offset of the received satellite signals when receiver 100 is configured for GPS positioning. There is no previous information about its location when the receiver 100 is first turned on. The receiver 100 determines its initial position by searching all possible pseudorandom code sequences transmitted by each satellite. In addition, the receiver 100 must search all possible phases of all possible pseudorandom codes. The search is performed by a plurality of correlators operating in parallel with minimizing the search time required by receiver 100. Each correlator operates on a single pseudo-random sequence. The correlator attempts to determine the phase offset between the internally generated pseudorandom code and the code received from the satellite. Pseudo-random codes that do not correspond to satellite signals have no correlation due to the random nature of the codes. Furthermore, the correct pseudorandom code has no correlation with the received signal unless the phases of the two code signals are adjusted. Thus, when adjusting the phase of both signals, the correlator 130 only provides an indication of correlation in the correlators that have the same pseudo-random code as the received signal.

The results of the correlator are coupled to the peak detect 140 processor. Many correlators operate simultaneously and provide results to the peak detection processor in parallel and simultaneously. The peak detection 140 processor determines the maximum possible pseudorandom code and phase offset of the received signal.

GPS utilizes orthogonal codes for each satellite. This allows all satellites to transmit at the same frequency at the same time. Such that the receiver receives information from multiple sources simultaneously. The correlators 130 operate independently of each other and can determine the phase of the received pseudorandom code in the presence of other orthogonal codes. Thus, the peak detection 140 processor 140 is simultaneously provided with correlation codes that identify a plurality of pseudorandom codes as well as those code phase offsets. Since each satellite is assigned a pseudorandom code, the identification of the pseudorandom code identifies the particular satellite as its source. In addition, the phase offset determination of the code determines the time of arrival of that signal. Processor 150 analyzes the information in the peak detection 140 processor to calculate the position of receiver 100. The simultaneous determination of the pseudorandom code and the code phase offset allows the processor 150 to make an estimate of the receiver position when updating the peak detection 140 processor.

However, if the LO 114 frequency in downconverter 110 is not accurate, the search process is complicated. Fig. 2 shows a block diagram of a typical phase-locked loop (PLL) synthesized LO 200. The reference oscillator 202 serves as a frequency reference for the PLL. The reference oscillator 202 may be a fixed frequency oscillator or a steady state Voltage Controlled Oscillator (VCO) with a small tuning range. The wireless telephone may utilize a voltage controlled temperature compensated crystal oscillator (VCTCXO) as the reference oscillator 202. If a VCO is used as the reference oscillator 202, a reference adjustment control line 204 is provided.

The output of the reference oscillator 202 is coupled to a reference divider 210. The frequency of the reference oscillator 202 is scaled down using a reference divider 210. This is important because the output frequency of the PLL is proportional to the frequency input to the phase detector 220. The output of reference divider 210 is provided as one input to phase detector 220.

VCO 240 generates output 244 of the PLL. The VCO must be able to tune over the desired frequency range of the PLL. The voltage applied to the VCO control line determines the operating frequency. The output 244 of the PLL may be used as an input to the mixer in the down-converter. The output 244 of the PLL is also coupled to an input of an output frequency divider 250. The output frequency divider 250 scales the frequency output 244 so that the frequency input to the phase detector 220 (the scaled output of the reference oscillator 204) is multiplied by the scaling factor of the output frequency divider to obtain the desired output frequency. The output of output divider 250 is provided as a second input to phase detector 220.

Phase detector 220 compares the output of reference divider 210 with the output of output divider 250 and generates an error signal as an output. The error signal output from phase detector 220 is coupled to loop filter 230. The band of loop filter 230 limits the error signal from phase detector 220. The output of the loop filter 230 is used as a control voltage on the VCO 240. Thus, it can be seen that the PLL output 244 derives its frequency accuracy from the frequency accuracy of the reference oscillator 202.

Errors in LO frequency accuracy complicate the search process. Fig. 3 shows the total search space 300 that each correlator must cover. Each correlator in the GPS receiver must search for the possibility of all code phases. The code phase search space 310 is shown in fig. 3 as a vertical search space. Each bin in the code phase search space 310 represents a minimum discernable phase difference. The length of the short pseudo-random code for GPS is 1023 bits long. If the pseudo-random nature of the code results in negligible correlation for all code phase offsets greater than zero, then the code phase search space 310 must cover all possible code phases. Thus, at least 1023 lattices are required in the code phase search space 310 to uniquely identify the phase of the pseudorandom code.

As can be seen in fig. 3, an increase in the frequency search space 320 increases the overall search space 300 proportionally. Since the frequency error of any code phase error is independent of each other, the frequency search space 320 represents an additional search dimension. Each bin of the frequency search space 320 represents a minimum discernable frequency interval. The size of the minimum discernable frequency interval is a function of the number of samples and the total integration time. The minimum discernable frequency interval decreases with increasing total integration time. In addition, a sufficient number of samples are required to achieve the desired discernable frequency spacing. An increase in LO bias results in an increase in the frequency search space 320.

The receiver correlates each sample defined in the overall search space 300 with each other. The results are continuously accumulated to further improve the signal-to-noise ratio (SNR) of the received signal. The LO bias causes the accumulated result to appear in multiple bins corresponding to the frequency bias. The "smearing" of the signal is shown in figure 3 in a number of shaded frequency bins. While an LO that does not show a bias enables the accumulated results to appear in only a single frequency bin. The signal identification is greatly improved by increasing the SNR.

Fig. 4 shows a block diagram of an LO stability circuit in a GPS enabled wireless phone 400. The radiotelephone 400 incorporates a telephone transceiver 410 that allows communication through a radiotelephone system. The radiotelephone 400 also incorporates a GPS receiver 420 to assist in position location. In the embodiment shown in fig. 4, the radiotelephone 400 operates in either the telephone mode or the GPS mode, both modes not operating simultaneously. However, if the radiotelephone 400 has sufficient processing power, the phone and GPS modes may operate simultaneously.

Radio Frequency (RF) signals are coupled to the radiotelephone using antenna 402. The RF signals coupled through antenna 402 include transmit and receive signals of a telephone transceiver 410 and receive signals of a GPS receiver 420. In the embodiment shown in fig. 4, GPS receiver 420 and telephone transceiver 410 share the same LO 450. As discussed above, inaccuracies in LO450 result in a larger search space for GPS receiver 420. Thus, the embodiment shown in fig. 4 utilizes information received by telephone transceiver 410 to discriminate LO450 so that the frequency error of LO450 can be minimized when GPS receiver 420 searches.

To discriminate the internal LO450, the radiotelephone 400 is provided with an external signal having high frequency stability. In a wireless system, such as a Code Division Multiple Access (CDMA) system specified in the 95-B mobile STATION COMPATIBILITY FOR digital subscriber line, Telecommunication Industry Association (TIA)/Electronic Industry Association (EIA), signals are continuously broadcast by a BASE STATION. The signal continuously broadcast by the base station includes a pilot channel and a synchronization channel. Both signals have high frequency stability and either can be used as an external reference needed to discriminate LO 450.

A radiotelephone 400 designed to operate in a CDMA system (such as that specified by TIA/EIA 95-B) incorporates a searcher in the receiver to continuously search for the presence of pilot signals. In the wireless telephone 400, a receiver in the telephone transceiver 410 receives a pilot signal transmitted by a base station (not shown).

The presence of the pilot signal can be utilized by the radiotelephone 400 to improve signal acquisition in the GPS mode. The receiver uses the frequency-stabilized pilot signal as an external frequency reference to determine the frequency error in LO 450. The frequency error determined by the receiver is reported to the oscillator discrimination circuit 430. In addition, sensors 440, 442 are distributed in the radiotelephone 400 to monitor factors that cause frequency error in the LO 450. Factors that may be monitored by sensors 440, 442 include, but are not limited to, temperature gradient, operation of the RF Power Amplifier (PA), RF PA duty cycle, battery voltage, accumulated supply time, humidity, or any other variable that determines the frequency error that causes LO 450. Sensor 440 couples a signal to oscillator discrimination circuit 430. A plurality of digital values corresponding to the readings of the sensor 440 are averaged and the average value is stored in an array of the memory 434. If the sensor 440 outputs an analog value, the oscillator discrimination circuit 430 digitizes the reading before averaging and storing the average in the memory 434. If the sensor 440 outputs a digital value, the oscillator discrimination circuit 430 does not need to further condition the signal, but merely saves the averaged digital sensor 440 reading. The averaging is performed by a processor 432 forming part of the oscillator discrimination circuit 430.

The oscillator discrimination circuit 430 also averages a plurality of frequency error readings determined and reported by the telephone transceiver 410. The average frequency error reading is also stored in an array in memory 434. The average frequency error is stored in the memory 434 location relative to the readings of the corresponding average sensor 440-442. In this form, the snapshot of the operating environment and the corresponding frequency error of LO450 are classified. As long as the radiotelephone 400 is operating in the telephone mode, the oscillator discrimination circuit 430 continues to accumulate new sensor 440 readings and corresponding frequency errors. When the radiotelephone is operating in GPS mode, the oscillator discrimination circuit 430 uses previously saved sensor 440 readings and frequency error information to assist in signal acquisition by the GPS receiver 420.

To assist in GPS signal acquisition, the oscillator discrimination circuit 430 reads the value of each sensor 440-442. The processor 432 then compares the current sensor 440-442 values to the previously stored array of values. The possible LO450 frequency error is determined from the previously stored values corresponding to the sensor 440-442 readings. If there are no readings of the accurate sensors 440-442 in the array, the processor 432 adds or extrapolates from the existing values. Whereby oscillator discrimination circuit 430 determines a possible LO450 frequency error. Oscillator discrimination circuit 430 then generates an error signal that is applied to LO control line 438 to compensate for the frequency error. In one embodiment, an oversampled high dynamic range delta-sigma modulator is used as a digital-to-analog converter to convert the error signal from a digital value to an analog value that is applied to the LO. The oscillator discrimination circuit 430 may optionally send the frequency error value to the GPS receiver 420 on the information bus 436. Knowing the frequency error allows the GPS receiver 420 to reduce the search space and acquire the signal with fewer computations. Oscillator discrimination circuit 430 may optionally provide a combination of both corrections. The oscillator discrimination circuit 430 may provide an indication of the frequency error to the GPS receiver at the beginning of entering GPS mode, and may then correct for any frequency offset by providing a signal on the LO control line while the radiotelephone 400 remains in GPS mode. Actively correcting the LO450 frequency offset minimizes signal smearing effects that occur when the LO450 frequency drifts over multiple frequency bins during successive correlation accumulations.

The signal provided on LO control line 438 is used on PLL synthesis LO 200 for frequency error compensation of LO450 as shown in fig. 2. Referring back to fig. 2, recall that the output frequency 244 is proportional to the output of the reference oscillator 202. Once the VCO gain of the reference oscillator 202 is known, the change in the output frequency 244 for a given reference adjustment 204 voltage change can be determined. In this way, the oscillator discrimination circuit 430 of fig. 4 may calculate a voltage to drive the reference adjustment 204 line of the PLL synthetic LO 200 to compensate for the determined frequency error.

Fig. 5 shows an alternative embodiment of a radiotelephone 500. The radiotelephone 500 of fig. 5 incorporates the telephone transceiver 410 and GPS receiver 420 described above. However, in the wireless telephone 500 of fig. 5, the telephone transceiver 410 uses either the first LO 550 or the second LO450, which is distinct from the GPS receiver 420. The terms first and second LO are used herein to distinguish an LO for the telephony transceiver 410 from an LO for the GPS receiver 420. The terms first LO and second LO are not used to describe multiple LOs used in a receiver for multiple frequency conversion needs. The operation of oscillator discrimination circuit 430 is somewhat different where two different LOs 450 and 550 are used. The telephone transceiver 410 may continuously receive the pilot signal and report the corresponding frequency error to the oscillator discrimination circuit 430. The frequency error of the first LO 550 of the telephony transceiver 410 is effectively used as a proxy for the frequency error of the second LO450 of the GPS receiver 420. When two LOs 450 and 550 are used, radiotelephone 500 also need not operate in different phone and GPS modes, provided that sufficient processing power is present in radiotelephone 500. But the discrimination of the frequency error operates independently and simultaneously with the correction of the frequency error of the second LO450 of the GPS receiver.

Fig. 6A and 6B show block diagrams of the LO discrimination process. Referring to fig. 6A, the process begins at block 602. Block 602 may represent initiating an LO discrimination process by the control processor. Once the process is initiated, the routine proceeds to block 604 where an external frequency source is received. The external frequency source may be input to the receiver or received over the air as described in the receivers of fig. 4 and 5. An external frequency source is used as a frequency reference to calculate the frequency error of the LO in block 606. Where the CDMA pilot signal is used as an external frequency source, the CDMA receiver determines the frequency error of the LO. The routine proceeds to block 608 and stores the frequency error value determined in block 606. The program then proceeds to a decision block 610 to determine whether a predetermined number j of frequency error samples have been saved. The predetermined number j represents the number of averages of the frequency error samples. This number can be as low as and as high as can be tolerated to implement hardware and timing constraints in the device. If j samples have not been saved, the process returns to block 604 to capture additional samples. Once the predetermined number j of samples has been saved, the program proceeds to block 620 where the j frequency error samples are averaged. In an alternative embodiment, a moving average of the frequency error may be calculated. Moving averages have the advantage of being able to discriminate the LO frequency over very long periods of time. The disadvantage is that the moving average does not respond quickly to changes in the operating environment that cause LO frequency errors.

Once the samples are averaged, the routine proceeds to block 622 where the average frequency error is stored in memory. After the average frequency error is saved, the process proceeds to point 630. Point 630 does not represent the operation of the program. But merely serves to connect fig. 6A and 6B. The process continues in fig. 6B, and the process proceeds to block 640, where a reading of the sensor is received. At least one sensor reading is required and the upper limit of the sensor reading is limited only by the amount of hardware and processing power available in the implemented device. Each sensor reading is saved to memory at block 642. The program then proceeds to decision block 650 to determine whether a second predetermined number k of samples are saved from each sensor. If the second predetermined number k of sensor readings has not been captured and saved, the routine returns to block 640 to capture further samples. Once the second predetermined number k of sensor samples have been captured and saved, the program proceeds to block 660 where each sensor reading is averaged over the k previously saved values. The number of sensor readings that require averaging is selected by the designer as is the case with the number of frequency error samples that require averaging. The average sensor reading is saved in memory at block 662. At this point the LO discrimination process is complete and the procedure may end or the discrimination of the LO may continue through return point 603 as shown in fig. 6B.

Fig. 7 shows a block diagram of the LO compensation procedure performed once at least one LO discrimination procedure loop has taken place. The process begins at block 702. This start may represent the start of a GPS mode in a wireless telephone implementing a GPS receiver and a telephone transceiver. Alternatively, this start may represent the end of a loop of the LO discrimination procedure, wherein LO compensation continues to occur, just as in the LO discrimination process.

The program then proceeds to block 704 where the sensor values are read. These sensor readings represent the most recent sensor readings. The program then proceeds to decision block 710 where the sensor value is compared to the previously saved sensor reading. If the sensor reading matches a value already present in the discriminatory array, then the program proceeds to block 730 where the array is queried for a frequency error corresponding to the saved sensor value. If, however, the sensor value is not present in the LO discrimination array, then the routine proceeds to block 720 where the frequency error is calculated by adding or extrapolating the saved sensor reading to match the current sensor reading and thereby produce an estimated LO frequency error. From block 720 or 730 the process proceeds to block 740 where appropriate LO corrections are calculated based on the estimated LO frequency error. The modification of the LO is calculated by determining the frequency error from the LO discriminator array and calculating the LO control signal from a transfer function known to relate the LO control line signal to the output frequency. Where the LO control line is the voltage controlled signal of the VCO, the transfer function is determined by the VCO gain. Once the LO modification procedure is determined, the process proceeds to block 742. The routine applies LO corrections to the LO at block 742. Alternatively, or in addition to applying LO corrections, the program may report the data to the GPS receiver. The data may consist of the determined LO frequency error and any corrections applied to the LO. Using this information and the compensated LO, the GPS receiver can acquire the signal more quickly and efficiently.

The previous description of the preferred embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without the use of the inventive faculty. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (15)

1. An apparatus for compensating for Local Oscillator (LO) frequency error, comprising:

a radiotelephone receiver for receiving base station signals;

a GPS receiver including an LO;

circuitry for receiving the base station signal and calculating an LO frequency error from the base station signal;

at least one sensor that monitors at least one variable that causes LO frequency error; and

an oscillator discrimination circuit that receives the calculated LO frequency error and a reading from the at least one sensor;

wherein the oscillator discrimination circuit generates a compensation signal based on the calculated LO frequency error and the reading from the at least one sensor.

2. The apparatus of claim 1 wherein said oscillator discrimination circuit applies said compensation signal to said LO to correct said LO frequency error.

3. The apparatus of claim 1, wherein said oscillator discrimination circuit stores readings from said at least one sensor and a calculated LO frequency error in a memory device when operating in a phone mode, and wherein said oscillator discrimination circuit generates said compensation signal when operating in a GPS mode.

4. The apparatus of claim 3 wherein said oscillator discrimination circuit generates said compensation signal by comparing a most recent set of readings from said at least one sensor to readings stored in a memory device and generates said compensation signal based on a corresponding calculated LO frequency error stored in the memory device.

5. The apparatus of claim 3 wherein said oscillator discrimination circuit adds or extrapolates readings stored in a memory device to match a most recent set of readings from said at least one sensor in order to generate an estimated LO frequency error, and wherein said oscillator discrimination circuit generates said compensation signal based on the estimated LO frequency error.

6. The apparatus of claim 1 wherein said wireless telephone receiver is adapted to receive code division multiple access CDMA signals.

7. The apparatus of claim 6 wherein said base station signal is a CDMA pilot signal.

8. The apparatus of claim 7, wherein said oscillator discrimination circuit stores readings from said at least one sensor and a calculated LO frequency error in a memory device when operating in a phone mode, and wherein said oscillator discrimination circuit generates said compensation signal when operating in a GPS mode.

9. The apparatus of claim 8 wherein said oscillator discrimination circuit applies said compensation signal to said LO to correct said LO frequency error.

10. The apparatus of claim 8, wherein said at least one sensor comprises a temperature sensor.

11. A method of compensating for local oscillator, LO, frequency error in a mobile telephone system having a mobile telephone receiver and a GPS receiver, comprising:

monitoring readings and LO frequency error of at least one sensor in a mobile phone operating mode, including

Receiving a base station signal having a high frequency stability with said mobile telephone receiver;

generating the LO frequency error by comparing the base station signal to an LO frequency;

receiving a reading from the at least one sensor; and

compensating the estimated LO frequency error in the GPS mode of operation based on readings from the at least one sensor and the monitored LO frequency error in the mobile phone mode of operation.

12. The method of claim 11, further comprising:

storing readings from the at least one sensor and corresponding LO frequency errors.

13. The method of claim 12, wherein the compensating step comprises:

receiving a most recent set of readings from the at least one sensor;

an estimated LO frequency error is generated using a most recent set of readings from the at least one sensor and readings from the at least one sensor received in the first mode.

Generating an LO correction signal from the estimated LO frequency error; and

the LO correction signal is applied to the compensating LO.

14. The method of claim 11 wherein said base station signal is a code division multiple access CDMA pilot signal.

15. The method of claim 14, wherein said at least one sensor comprises a temperature sensor.