Method for frequency acquisition, especially for initial frequency acquisition, of a mobile communications device

Abstract

A method is provided for frequency acquisition, particularly for initial frequency acquisition, pursuant to a known synchronization sequence for synchronizing a mobile communications device having a local oscillator with previously known transmit frequencies of a base station that transmit in a known channel raster with defined frequency points within a band, wherein the method includes the steps of determining the inband power of the known synchronization sequence via a sensor by scanning a frequency band, performing coarse determination of the local power maximum of the inband power and, thus, of the received carrier frequency over the scanned frequency band, producing a presumed channel frequency at which the base station is transmitting on the basis of knowledge of the channel raster of the transmit frequencies of the base station, performing fine determination of the received carrier frequency by comparison with the known synchronization sequence, and correcting the frequency deviation of the local oscillator from the transmitted carrier frequency.

Description

[0001] The invention is related to a method for frequency acquisition, in particular for initial frequency acquisition, by means of a known synchronization sequence for synchronizing a mobile communications device having a local oscillator with the previously known transmit frequencies of a base station that transmit in a known channel raster with defined frequency points.

[0002] In mobile communications devices, a standard for mobile communication makes available a synchronization sequence known throughout the network. In addition, the transmit frequency of the base station is known with a sufficiently high accuracy specified by the standardization bodies. This accuracy is currently 0.05 ppm in the UMTS Standard. The channel raster in which the base station transmits is known in this case a priori to the mobile station. In the case of a UMTS mobile radio network, said raster comprises channels at a spacing of 200 kHz at defined frequency points. By contrast, it is not known which channel is currently being used by the base station, nor which frequency drift the oscillator of the mobile receiving station has. The manufacturing tolerances and temperature spreads, as well as other possible external influences result in a frequency inaccuracy of the local oscillator which is typically +/−25 ppm. The frequency inaccuracy is even mostly still above this value when use is made of cost-effective local oscillators (LO). Such a large inaccuracy is disadvantageous, since this leads to a complicated and lengthy synchronization of the mobile part with the base station. This problem has so far been solved by calibrating the mobile parts of the communications device during fabrication, and thus lowering the frequency drift of the local oscillator to approximately +/−3 ppm. However, this is very time-consuming and cost-intensive.

[0003] Moreover, it has been proposed to make use for the purpose of initial autocalibration when first switching on the mobile part, given a frequency inaccuracy of, however, +/−25 ppm, of an algorithm in the case of which a carrier, modulated with the aid of an SCH channel (synchronization channel), of the base station is scanned by means of an SCH correlator accompanied by variation of the receive frequency. In this case, use has so far been made as SCH correlator only of a single so-called matched filter, which is matched to the synchronization frequency of the primary SCH channel, which is denoted from now on as SCH channel, for short. After initial autocalibration has been performed, a frequency accuracy of +/−3 ppm can be assumed in the synchronization operations following thereupon. Since the correlation bandwidth for the codes used is small according to specification—of the order of magnitude of 10-20 kHz—by comparison with the scanning range, whereas the scanned frequency band can comprise the complete UMTS bandwidth of 60 Mhz, this method is very time-consuming and costly. Step sizes smaller than 14 kHz must be used in this case for correlation. This step size can be obtained from simulation results. The initial frequency acquisition requires 6000 steps for a step width used of 10 kHz, for example.

[0004] It is therefore the object of the invention to make available a method by means of which a frequency acquisition in the mobile part of a mobile communications device can be accelerated.

[0005] The object is achieved by means of a method in accordance with patent claim 1. Developments of the invention are the subject matter of the subclaims.

[0006] The basic idea of the invention consists in determining the inband power by means of a sensor in parallel with the conventional SCH correlation which is used, as previously, as an integral constituent of the newly developed method. The presumed channel frequency at which the base station is transmitting is deduced from the determination of the local power maximum over the entire scanned frequency band. The channel spacing, known a priori, of the possible discrete channel frequencies is used for this purpose in a supplementary way. The channel spacing between the frequency points is 200 kHz in the case of a UMTS mobile radio network. Moreover, the transmit frequency of the base station is known with a sufficiently high accuracy, specified by the standardization bodies, of +/−0.05 ppm. Given a maximum frequency of 2.17 GHz, the result is a maximum uncertainty of approximately 20 Hz with reference to the signal transmitted by the base station. This high accuracy is achieved by GPS, or DVF77 signals. It is thereby possible for the channel frequencies emitted by the base station to be used as differential signals for tuning the local oscillator of the mobile station. This is attended by the advantage that the synchronization is performed substantially faster, and this leads to a greatly reduced power consumption in the mobile part. This means that the standby times of the mobile part can be significantly increased.

[0007] It is preferred to apply the method according to the invention upon the first startup and/or each further xth startup of the mobile communications device. Aging effects and temperature deviations of the components that are used to generate frequency are thereby automatically recalibrated. Since this takes place automatically during callup, no additional outlay is caused. By contrast, calibration via the temperature during fabrication would be time-consuming and cost-intensive. Moreover, the communications device can be matched adaptively to the user's surroundings in the case of the method described. Again, this method

[0008] leads to an increase in the product service life owing to the recalibration rendered possible. The determination of the dummy x is a function of a plurality of factors, for example of the temperature fluctuations in the surroundings of the oscillator, and the age of the latter. The value can be 1, such that a recalibration is performed with each synchronization, or else, for example, 2, 10, etc. The larger the value of x, the less is the power consumption, and longer standby times of the mobile part can be achieved.

[0009] It is preferred, furthermore, when the accurately determined frequency of the local oscillator is adopted after the synchronization as reference by storing the determined hardware calibration parameters in a memory of the mobile communications device. It is possible as a result to undertake a first calibration of the local oscillator in the RF section of the mobile station, that is to say the algorithm in this case automatically adopts the settings for calibrating the local oscillator which would otherwise have to be undertaken during the production process.

[0010] In particular, this method can be advantageously applied in the case of a UMTS mobile radio network, the determination of the inband power being performed by means of a power sensor, and the fine determination of the received carrier frequency being performed by means of an SCH correlator.

[0011] A possible development of the invention provides for synchronizing the mobile communications device with a first base station and to carry out the callup to this base station, but that a switch is made to a second base station during the call. This is advantageous whenever it emerges during a call that another base station is being received more strongly than the base station with which synchronization was originally carried out. Consequently, by comparison with the known method a higher level of reliability and of frequency accuracy is achieved than in the case of

[0012] a conventional method in which only an SCH correlator is used. The reason for this is that the results are confirmed independently of one another simultaneously by the SCH correlator and the power sensor. This means in the case of a UMTS mobile radio network that each of the possible channels of this network can be used for synchronization. For example, it is possible to perform synchronization with channel 2 of a base station of a first operating company; the actual callup takes place, however, subsequently on channel 5 of another base station of a second operating company. This is possible because the UMTS base stations are transmitting at approximately the same frequency, the difference between them being only +/−1 ppm.

[0013] The invention is described in more detail below with the aid of a preferred exemplary embodiment whose explanation is served by the figures set forth below, in which:

[0014] FIG. 1 shows a schematic of the cycle of an exemplary embodiment according to the invention, and

[0015] FIG. 2 shows the frequency bands to be scanned for the individual steps during the method according to FIG. 1,

[0016] FIG. 3 shows a measuring instrument of a power sensor over a frequency profile, and

[0017] FIG. 4 shows the measurement result of FIG. 3, combined with the measurement result of a pSCH correlator over a frequency band.

[0018] Specified in FIG. 1 are the individual groups of information that flow together in an evaluator 1 and from which the frequency acquisition can be undertaken.

[0019] Firstly, the a-priori knowledge 2 about the channel raster of the base station flows into the evaluation; furthermore, the result, measured with the aid of a power sensor 3, of the inband power over a

[0020] frequency band, as illustrated in FIG. 3. Also flowing in is the information that is obtained via a primary synchronization channel (pSCH) 4, as illustrated in FIG. 4. Signal packets 5 are transmitted slotwise in the pSCH 4 at a temporal spacing Ts1. The signal packets 5 illustrated in FIG. 1 have a double-peak signal 6 from a first base station and a single-peak signal 7 from a second base station. The received signal packets 5 are subjected to slotwise accumulation 9 via a matched filter 8. After the data thus obtained has been processed via an indicator 10 and a peak detector 11, the information obtained therefrom is passed on to the evaluator 1. The latter then performs the accurate determination of the received frequency of the local oscillator in the mobile station 16 of the mobile communications device.

[0021] The cycle of an accurate determination of the frequency drift of the local oscillator, and thus of the calibration of the latter with the aid of a UMTS mobile radio network is shown in FIG. 2. The frequency band covered by UMTS is 60 MHz wide and has enumerated channels. The center carrier frequency of the base station is approximately 2.1 GHz. The inband power, which has a bandwidth of approximately 3.84 MHz, is measured in a first iteration step by means of the power sensor 3 for the purpose of coarse determination of the transmit frequency of the base station. A coarse search is thereby placed ahead of the actual correlation. In conjunction with the a-priori knowledge of the channel raster, the results obtained fix the received channel frequencies, which stem from different base stations, as early as in the preselection from +/−200 kHz, since this corresponds exactly to the channel raster. This means that the received channels are determined in the 200 kHz raster by the power sensor 3. The maximum determined by the power sensor 3 is at approximately 10 MHz, this frequency already being reduced about the center carrier frequency of the base station. The step width can be 1 MHz because of the a-priori knowledge of the channel raster, it being possible to determine the

[0022] channel frequency up to +/−200 kHz, that is to say the power sensor 3 determines the received channels in the 200 kHz raster on which the base stations are transmitting. It is also possible to use a step width of over 1 MHz; however, it is preferred to use the lesser step width of 1 MHz when several base stations are transmitting.

[0023] The fine adjustment of the frequency of the local oscillator is performed via the pSCH correlator, which is specified according to the prescribed standard. The results of this correlator are used for the purpose of correcting frequency deviations of the local oscillator that exist because of manufacturing tolerances and temperature fluctuations.

[0024] On the basis of the channel raster thus found, it is possible to use a second iteration step to achieve the fine tuning of the oscillator drift by means of the pSCH oscillator with a substantially lower outlay. During the second iteration step, the received signal is compared with the pSCH sequence already known and uniform throughout the network. The pSCH correlator has a very narrow correlation bandwidth, that is to say a clear scanning result which is, however, of very high power is obtained here only in conjunction with very low frequency detuning. The correlation bandwidth is 10-20 kHz, in particular approximately 16 kHz, and this can be obtained from simulation results. Other values can also result for the correlation bandwidth if the synchronization codes change during the standardization process.

[0025] Since the channel frequency is already determined up to +/−200 kHz, the exact frequency at which the base station is transmitting can be determined in few steps in a third iteration step with a step width below the correlation bandwidth, for example 10 kHz here.

[0026] The third iteration step is, however, unnecessary when the a-priori knowledge about the channel raster is used, and the local oscillator (LO) of the mobile station has already been so well calibrated that it has a remaining inaccuracy of only +/−3 ppm; this corresponds to 6 kHz.

[0027] FIG. 4 shows a diagram in which, firstly, the inband power 12 illustrated in FIG. 3 and its maximum 13 are shown. Secondly, the output 14 of the pSCH correlator in the baseband is plotted below the inband power 12 of the power sensor 3. Here, as well, a correction was undertaken by the carrier frequency of the base station of approximately 2.1 GHz. The output 14 of the pSCH correlator has a correlation peak 15 at 10 MHz. The actual natural frequency of the local oscillator in the mobile station can be deduced from this result, since the channel raster with which the base station is transmitting is known a priori to the mobile station. In this case, the carrier frequency of the base station is precisely calibrated with an accuracy of 0.05 ppm. The deviation of the local oscillator of the mobile station 16 from the known reference frequency is determined by comparing the position found for the measured correlation peak 15 with the known channel frequency of the base station. The data for autocalibration are thereby obtained. The manufacturing tolerances in the natural frequency of the local oscillator of +/−25 ppm can be compensated to the extent that it is possible to assume a maximum deviation of +/−3 ppm for future synchronization operations. Future synchronization processes are thereby substantially accelerated.

[0028] If a first case of synchronization is involved, that is to say the device is being started up for the first time at the customer's, there is consequently a maximum deviation of 50 ppm from the reference frequency. This means that the pSCH correlation can be performed, for example, in 50 steps of step width 1 ppm, that is to say 2 kHz. After the first calibration has been performed, the maximum permissible deviation is then only 3 ppm. The pSCH correlation then requires only six steps given an identical step width of 2 kHz. This means that after initial synchronization has been performed

[0029] the automatic frequency acquisition described operates almost 10 times more quickly.

[0030] During the initial calibration of the local oscillator in the RF section of the mobile part 16, the algorithm can undertake settings for calibrating the local oscillator that would otherwise have had to be undertaken during the production process. Instead of undertaking the calibration during the production process, this is done automatically during the first startup of the device at the customer's. After the synchronization, the accurately determined frequency of the local oscillator is taken over as reference by storing the hardware calibration parameters.

Claims

1. A method for frequency acquisition, especially for initial frequency acquisition, by means of a known synchronization sequence for synchronizing a mobile communications device having a local oscillator with the previously known transmit frequencies of a base station that transmit in a known channel raster with defined frequency points within a band, having the following steps: a) determining the inband power of the known synchronization sequence by means of a sensor by scanning a frequency band, b) coarse determination of the local power maximum of the inband power, and thus of the received carrier frequency over the scanned frequency band, c) producing the presumed channel frequency at which the base station is transmitting on the basis of knowledge of the channel raster of the transmit frequencies of the base station, d) fine determination of the received carrier frequency by comparison with the known synchronization sequence, and e) correction of the frequency deviation of the local oscillator from the transmitted carrier frequency.

2. The method as claimed in claim 1, characterized in that it is applied upon the first startup and/or each further startup of the mobile communications device.

3. The method as claimed in one of the preceding claims, characterized in that the accurately determined frequency of the local oscillator is adopted as reference by being stored in a memory of the mobile communications device.

4. The method as claimed in one of the preceding claims, characterized in that