EP2073409A1

EP2073409A1 - A network following method and a radio apparatus for in-vehicle use

Info

Publication number: EP2073409A1
Application number: EP07150400A
Authority: EP
Inventors: Michael Görtler; Jörg LEBENDER
Original assignee: Denso Ten Ltd
Current assignee: Denso Ten Ltd
Priority date: 2007-12-21
Filing date: 2007-12-21
Publication date: 2009-06-24

Abstract

A network following method of switching a radio tuner (2) between different frequencies over which a same programme is broadcast, the method comprising: tuning the radio tuner (2) to a first frequency (NDR) of a plurality of alternative frequencies on which a same programme is broadcast, so as to receive a first frequency signal of the programme; periodically monitoring quality indicators of the first frequency signal; periodically monitoring quality indicators of the signal of each of the other alternative frequencies (NDR AF1; NDR AF2) and storing the monitored results; and switching the radio tuner (2) from the first frequency (NDR) to a second frequency (NDR AF2) of the alternative frequencies in accordance with a quality value AQ determined from the monitored quality indicators for the first frequency signal (NDR) and for the other frequency signals (NDR AF1; NDR AF2); wherein the quality value, AQ, of each frequency signal is determined by applying a synergistic function of a field strength value, V(FS), and a distortion index, I_D, of the signal, where the distortion index I_D decreases as distortion D increases.

Description

The invention relates to a network following method and to a radio receiving apparatus for in-vehicle use.
In-vehicle reception of broadcast programmes using a radio receiving apparatus is well-known. Such reception suffers from difficulties which do not tend to occur for a stationary receiver because of the movement of the vehicle.
In particular, for receiving a radio signal at a stationary site a directional antenna such as a Yagi or dipole antenna may be used. This antenna is aimed in the direction of the transmitter site in such a fashion that the received signal originates mainly along a direct path with delayed and reflected signals arriving from lateral paths largely excluded.
In a vehicle, however, an omni-directional antenna needs to be used because the reception angle of the radio waves can adopt any value within 360 degrees. As a result of this, the vehicle receiver must be able to cope with all of the signals arriving at the receiving antenna, independent of the driving direction and the speed of the vehicle.
The signal received at the vehicle receiver suffers from multipath propagation and from Doppler shift, which is caused by the movement of the receiver itself or by moving reflecting obstacles. This is shown in Figure 1.
Here, two main propagation paths where the waves are reflected on discrete obstacles such as tall buildings, hills or mountain ridges can be seen. As the average length of the two paths is different, the signals will arrive at different time instants in the vicinity of the car. Additionally the signals are exposed to local dispersion due to further reflections in the immediate surrounding of the receiving antenna.
These influences lead to a sum signal at the antenna which is continuously varying in amplitude, phase and frequency. In consequence, this fading signal has a strong impact on the quality of the audio signal after demodulation in the receiver. In particular, for an FM signal the following may occur:

a)Heavy distortion due to missing parts of the FM signal spectrum;
b)Fluctuating noise occurring each time the signal level falls below the limiting threshold of the receiver;
c)Periodical noise bursts (paling fence effect) in case the fading signal is dominated by two main paths with Doppler shift.

One way of mitigating the above problems is to employ at least two independent antennae. Switching can then be performed such that the antenna which delivers the better signal is connected to the receiver input, thus providing more stable reception conditions. Another way of mitigating the problem is to use antenna phase diversity. Here, the signals of two independent antennae are fed to two tuners. The IF signals of both tuners are then processed so that they can be added and thus simultaneously contribute to a better reception signal.
These two methods are unable to help significantly in cases where the signal level becomes too low.
Therefore in Europe the Radio Data System (RDS) has been developed. This system provides a means for performing 'network following'. In the RDS system, detailed information about other frequencies of the same programme or transmitter chain are included in an RDS data channel of the received programme signal. This information enables a receiver to check for alternative frequencies on which the same programme (radio station broadcast) is transmitted, and to change the frequency to which it is tuned if its received signal level becomes too low. In other words, the receiver is able to use the RDS system to 'follow' a particular network (radio programme/channel broadcast) across different frequencies.
In the RDS system, a data link layer consists of blocks of data. Each block typically has 26 bits, consisting of 16 bits of data and 10 check bits. An RDS Group consists of 4 blocks, as shown in Figure 2. In the Group, block A includes a program identification (PI) code, which is the unique identifier of a radio station. The PI code is repeated 11.4 times per second and uniquely identifies a programme (radio station channel) and the country in which the programme is broadcast. In addition, use of the code enables selection of a programme independently of frequency, and further enables a receiver to perform automatic changes (i.e. without user input) of the received frequency so as to change to an alternative frequency on which the same or related programmes are broadcast. Also transmitted in the RDS data is alternative frequency information. The alternative frequency information includes the other frequencies on which the programme currently tuned to is also being broadcast. The PI code and the alternative frequency information enable the radio receiver to perform the above-described network following.
In a conventional in-vehicle radio receiver, a field-strength or level detector is provided. This detector detects the signal (field) strength of the channel tuned to. The field strength is directly related to signal quality because the signal-to-noise ratio mainly depends on it. The detector needs to be calibrated to compensate for the tolerances of the analogue components in the receiver hardware, so as to ensure that its output is an accurate reflection of the signal strength.
In the conventional receiver, when a particular frequency is tuned to and the detected field-strength of the received signal on that frequency falls below a pre-set threshold, stored alternative frequency information is searched and the alternative frequency having the best field-strength is tuned to. In other words, switching is performed between a first frequency on which a particular programme is broadcast to a second frequency on which the same programme is broadcast.
The stored alternative frequency information is maintained in the background by periodically tuning the tuner for a very brief time period to an alternative frequency. The time period for which the tuner is tuned to an alternative frequency is kept short enough so as not to noticeably disturb the sound heard by a listener. Each occasion on which the tuner is diverted from the tuned (listened to) frequency it is tuned to a different alternative frequency in a cyclic fashion, so that data (field strength) for all of the alternative frequencies is obtained.
It is also known to measure the multipath and noise of the frequency tuned to, i.e. of the frequency being listened to. The multipath can be measured using a multipath detector which evaluates amplitude fluctuations of the received signal. An FM signal is broadcast with a constant amplitude level, and hence fluctuations indicate a deterioration of signal quality. At multipath conditions large level fluctuations can be measured. The noise (ultra sonic noise) can be measured by considering the amplitude of the high frequency content of the multiplex signal.
Thus, by measuring the multipath and the noise of the received frequency, in addition to its field strength, a change of frequency tuned-to can be initiated if the detected multipath value or the detected noise value of the received frequency signal reaches a threshold. The alternative frequency selected to be switched to is the alternative frequency having the best field strength.
This method of network following suffers from substandard performance in certain situations. In particular, it can occur that the selected alternative frequency (i.e. the new frequency) has a large value of multipath and/or noise, despite having a high field strength. In this situation, the sound output by the receiver actually deteriorates after the change.
In addition, the switch may be performed to an alternative frequency but not to exactly the right frequency. For example, the frequency switched to could be 100 kHz beside the 'correct' alternative frequency. In this case the sound quality is also significantly deteriorated.
Further, a significant problem can occur in weak signal areas. Namely, due to the low signal strength a search is performed every few seconds to find a better alternative frequency. This leads to a large amount of mutes, and the resulting alternating change between a noisy sound and mute is very irritating to the listener.
It is therefore desirable to provide a network following method, and a radio receiver apparatus which can perform network following, that can mitigate the above problems.
According to a first aspect of the invention, there is provided a network following method of switching a radio tuner between different frequencies over which a same programme is broadcast, the method comprising: tuning the radio tuner to a first frequency of a plurality of alternative frequencies on which a same programme is broadcast, so as to receive a first frequency signal of the programme; periodically monitoring quality indicators of the first frequency signal; periodically monitoring quality indicators of the signal of each of the other alternative frequencies and storing the monitored results; and switching the radio tuner from the first frequency to a second frequency of the alternative frequencies in accordance with a quality value determined from the monitored quality indicators for the first frequency signal and for the other frequency signals; wherein the quality value, AQ, of each frequency signal is determined by applying a synergistic function of a field strength value, V(FS), and a distortion index, I_D, of the signal, where the distortion index I_D decreases as distortion D increases.
By calculating the quality value using a synergistic function, the effect on the quality value of the interaction between the field strength and the distortion is realised more accurately than merely summing the field strength and distortion. In other words, according to the synergistic function, a change in the quality value AQ due to a change in the distortion D (hence in the distortion index I_D also) at a fixed field strength value V(FS) depends on the field strength value V(FS).
In this way, an accurate quality value of the received first frequency signal and of the other frequency signals, which reflects the sound perceived by a listener taking account of the field strength and distortions, is used to determine whether to switch frequency among the alternative frequencies.
According to one embodiment, the synergistic function is a multiplication of the field strength value V(FS) and the distortion index I_D of the signal. By using such a synergistic function, the quality value of a high strength signal which suffers from strong distortion can be appropriately lowered. Using this relationship as the synergistic function thus enables the effect of strong distortion to be accurately reflected in the quality value even for a high field strength signal.
According to a preferred embodiment, the field strength value V(FS) of each signal is a field strength quality value Q(FS) determined by: detecting the field strength of the received signal; and performing a non-linear conversion on the detected field strength to convert it into the field strength quality value Q(FS).
In this way, the actual correlation between the field strength of the signal and the audible quality of the output sound as perceived by a listener is taken into account.
The non-linear conversion may be performed using a segmented interpolation (e.g. graph or algorithm). Alternatively, the non-linear conversion could be performed using a non-linear conversion formula, or a plurality of different conversion formulae for different field strengths. As a further alternative, the non-linear conversion could be performed by looking-up the field strength and its respective field strength quality value Q(FS) in a field strength conversion table.
According to a further preferred embodiment of the invention, the distortion value D of the signal is calculated using: D = b*M + c*N, where M is a value of the multipath, N is a value of the noise, and b and c are calibration values previously determined for the radio apparatus of the radio tuner.
According to a second aspect of the invention, there is provided a radio apparatus for in-vehicle use, the apparatus comprising: a tuner for tuning the radio apparatus to a first frequency of a plurality of alternative frequencies on which a same programme is broadcast, and for periodically sampling each of the other alternative frequencies; detecting means for detecting signal quality indicators of the signal of the first frequency and of the signals of the other alternative frequencies; signal quality determining means for determining a quality value AQ from the detected signal quality indicators for the signal of the first frequency and for the signals of each of the other frequencies; and switching means for switching the tuner from the first frequency to a second frequency of the alternative frequencies in accordance with the quality values AQ of the signals of the alternative frequencies; wherein the signal quality determining means is operable to determine the quality value AQ of the signal of each frequency by applying a synergistic function of a field strength value, V(FS), derived from the field strength detected by the detecting means, and a distortion index I_D of distortion D due to noise and multipath detected by the detecting means, wherein the distortion index I_D decreases as the distortion D increases.
According to one embodiment, the synergistic function applied by the signal quality determining means is a multiplication of the field strength value V(FS) and the distortion index I_D of the signal.
According to a preferred embodiment, the signal quality determining means is operable to determine the distortion D of each signal by calculating: D = b*M + c*N, where M is a value of the multipath detected by the detecting means, N is a value of the noise detected by the detecting means, and b and c are calibration values previously determined for the radio tuner.
According to a further preferred embodiment, the signal quality determining means is operable to calculate a field strength quality value Q(FS) as the field strength value V(FS) of each signal, by performing a non-linear conversion on the field strength detected by the detecting means so as to convert the field strength to the field strength quality value.
In one configuration, the signal quality determining means is operable to calculate the field strength quality value Q(FS) using a segmented interpolation (e.g. algorithm or graph) stored in memory. In an alternative configuration, the signal quality determining means could be operable to calculate the field strength quality value Q(FS) by performing a look-up operation in a field strength conversion table stored in memory, wherein the field strength conversion table stores field strengths and their respective quality values.
According to another embodiment, the switching means is configured to switch the tuner from the first frequency to the second frequency if the quality value AQ of the first frequency reaches a predetermined threshold.
According to an alternative embodiment, the switching means is configured to switch the tuner from the first frequency to the second frequency if the quality value AQ of the first frequency diminishes to a predetermined threshold and the quality value AQ of the second frequency is higher than that of the first frequency.
According to a further alternative embodiment, the switching means is configured to switch the tuner from the first frequency to the second frequency if the quality value AQ of the second frequency becomes higher than that of the first frequency.
According to another preferred embodiment, the apparatus further comprises an alternative frequency table stored in memory, wherein the alternative frequency table includes each alternative frequency and its respective quality value AQ.
Reference is now made, by way of example only, to the accompanying drawings, in which:

Fig. 1 shows multipath signal propagation from a transmitter to a radio receiver in a vehicle;
Fig. 2 shows blocks of data in an RDS group in the RDS data link layer;
Fig. 3 shows a radio receiver apparatus according to an embodiment of the invention;
Fig. 4 is a measurement graph showing received signal strength (field strength) on the horizontal axis and output signal and noise values on the vertical axis, for a particular type of radio receiver;
Fig. 5 is a segmented interpolation graph reflecting a non-linear relationship between field strength and audio quality of the output audio sound;
Fig. 6 is a diagram showing frequency switching according to an embodiment of the invention;
Figure 7 shows schematically the detection of offset by an offset detector.

Fig. 3 shows a radio receiver apparatus according to an embodiment of the invention. The apparatus includes an antenna 1, a tuner 2, a digital signal processor (DSP) 3 including detectors 4, a microcontroller (CPU) 5, memory 6, an operating portion 7, a display portion 8, a CD input 9, audio amplifier circuitry 10 and speakers 11. The apparatus also has a connector for connection to the power supply of the vehicle (i.e. to the vehicle battery).
Through the antenna 1 incoming radio waves can be picked up and input into the apparatus. The tuner 2 can be tuned to different frequencies, so that different radio programmes can be received. The DSP 3 performs signal processing in the digital domain on the signal output by the tuner 2. The detectors 4 enable indicators of the received signal quality to be detected. The CPU 5 is in overall control of the apparatus, for example it controls the frequency to which the tuner is tuned. Memory 6 associated with the CPU 5 stores program code for operating the apparatus, and also stores operating information such as alternative frequency information, which may be dynamically updated during use.
The operating portion 7 is provided to enable a user to input operation instructions, such as an instruction as to the radio station to be tuned to or the sound volume to be output. The operating portion could be, for example, a combination of buttons and knobs on the front panel of the apparatus. The display 8 enables information to be displayed to the user, such as whether the input signal is from the radio or a CD, and which radio station is being listened to. The display 8 could be an LCD display. The CD input 9 enables a CD to be input into the apparatus and played. The speakers 11 output the desired audio sound, after processing by the audio amplifier circuitry 10, to the listeners (users).
The detectors 4 include a fieldstrength detector for detecting the signal strength of the received frequency signal. The signal strength is directly related to signal quality because the signal-to-noise ratio mainly depends on it. This detector is calibrated to compensate for tolerances of the analogue components of the receiver hardware.
The detectors 4 also include a multipath detector for detecting the multipath of the signal. This detector evaluates amplitude fluctuations of the received signal. It does not require any alignment.
An Ultra Sonic Noise (USN) detector is also included in the detectors 4. This detector detects the amplitude of the high frequency content of the multiplex (MPX) signal. This is measured in the frequency range of around 80-150kHz. The USN detector gives information on noise due to an adjacent channel.
In use, a user sets the apparatus to a particular radio station using the operating portion 7 and the CPU 5 sets the tuner 2 to a frequency on which that station is broadcast. The quality of the received signal is then monitored by the detectors 4 and the CPU 5. In other words, certain quality indicators of the signal are detected periodically by the detectors 4 and analysed by the CPU 5.
These quality indicators include the fieldstrength, multipath and noise detected by the fieldstrength detector, the multipath detector and the noise detector, respectively. A quality value AQ is evaluated for the received signal of the current (tuned-to) frequency using these parameters. This (audio) quality value AQ is calculated using a synergistic function. According to the present embodiment, the synergistic function is realised by the equation (1): $AQ = Q (FS) * ({k * I}_{D})$

where Q(FS) is a field strength quality value related to the field strength detected by the field strength detector, I_D is a distortion index and k is a calibration parameter. The distortion index I_D is an index which increases in value as distortion D in the signal decreases in value. The distortion D depends on the multipath and noise present in the signal, and may be calculated by summing these effects using an equation (2): $D = b * M + c * N$

where M is the multipath determined by the multipath detector, N is the noise determined by the Ultra Sonic Noise detector, and b and c are calibration parameters. Thus, the multipath and the noise have a linear influence on the distortion.
Calculation of the audible quality value AQ from the values detected by the field strength, multipath and noise detectors of the detectors 4 results in a single value (quality value AQ) which accurately represents the actual audio quality experienced by a listener, i.e. by the user in the vehicle.
By calculating the quality of the audio signal in this way, the combined synergistic effect of each of the key individual factors affecting the audio quality, namely of the field strength, multipath and noise, is accurately reflected. In particular, for a given field strength, a change in the quality value due to an increase in distortion index (decrease in distortion) is accurately reflected, because the change depends on the field strength value and not just on the change in the distortion index. Hence, a judgement as to the actual audio quality of the current (tuned-to, presently received) frequency as perceived by a listener can be made.
By evaluating signal quality in this way, a major improvement is made over just considering the field strength, or considering each of the field strength, the multipath and the noise in isolation.
As already mentioned, the values of M and N are values output by the multipath and USN detectors. The distortion D is calculated from these values according to the equation (2), and the obtained value is converted into the distortion index I_D and inputted into equation (1) to enable the quality value AQ to be calculated.
As stated above, the parameters k, b and c are calibration parameters. The values of the calibration parameters k, b and c depend on the antenna system and on the tuner hardware. The values of these parameters are determined in advance for the radio receiver and are then set. In other words, pre-set values of the parameters are loaded into the memory 6 associated with the CPU 5 before sale of the receiver to an end user.
In this regard, these parameters are usually determined for the type of radio receiver; hence, a given model of radio receiver/antenna will have particular values of the calibration parameters k, b and c.
The field strength quality value Q(FS) is related to the actual field strength measured by the field strength detector. It has been known that field strength directly relates to signal quality because of the dependence of the signal-to-noise ratio S/N on it.
The present inventors have considered, however, that the relationship between the field strength and the audio quality heard by a user is not a linear relationship as has previously been assumed. In other words, depending upon the absolute value of the field strength, the effect of an increase or decrease in the field strength varies.
Accordingly, in calculating the quality value AQ, the field strength quality value Q(FS) is preferably used instead of simply using the field strength FS. In this way, a much improved correlation between the field strength and the sound quality perceived by a user is realised.
Figure 4 is a graph showing the received signal strength (field strength) on the horizontal axis and the output signal and noise values on the vertical axis, for a particular type of radio receiver.
In Fig. 4, the line A represents the output signal, the lines L and R represent the left and right output signals (channels) when the output signal is output in stereo, and the line NO represents the noise. The output signal is only output in stereo if the signal strength of the received signal is above a minimum level.
It can be seen from Fig. 4 that when the received signal has a very low field strength, for example in the range of about 0-20 dBµV, the signal-to-noise ratio is very poor but nevertheless increases significantly over this range. Hence, in this field strength range, each extra dBµV of signal strength will result in a marked improvement in the S/N ratio.
At about 30 dBµV stereo is opened and the radio apparatus outputs the output signal as left and right stereo signals. In the region from about 20-30 dBµV, the S/N ratio is still improved for each extra dBµV of the received signal, but the effect is less marked than in the lower field strength region of 0-20 dBµV.
In the region from about 35 to 44 dBµV, during which the audio channel separation is continuously increased from mono to maximum stereo, there is a much less significant improvement in the S/N ratio for each extra dBµV of the received signal.
Further, after about 44 dBµV at which the stereo is maximum open, there is no longer an increase in the S/N ratio for additional increases in the field strength.
Hence, simply to use the detected field strength as an indicator of quality results in an inaccurate reflection of the actual audio quality perceived by a listener. Advantageously, according to a preferred embodiment of the invention, a field strength quality value Q(FS) is used instead of the field strength FS.
The field strength quality value Q(FS) is determined by taking account of the non-linear relationship between the actual detected field strength and the signal-to-noise ratio of the output audio signal. The correlation between the field strength and the signal-to-noise ratio of the output signal can be realised in various ways.
One way of performing the correlation is to generate a table (for the radio receiver type to be used) in which actual detected field strength values are converted into field strength quality values Q(FS) in accordance with the relationship between each field strength value and the S/N ratio. In other words, by generating a measurement graph such as that shown in Fig. 4, a field strength quality value can be assigned to each field strength in accordance with the signal-to-noise ratio at each field strength. The values can then be written into a look-up table and stored in the memory 6 associated with the CPU 5.
An alternative way of performing the correlation is to use the graph of Fig. 4 to produce a segmented interpolation graph of the type shown in Fig. 5. Here, the horizontal axis shows the field strength in dBµV and the vertical axis shows the field strength quality value Q(FS) (as a number, no unit).
On the graph, a first segment is provided for a field strength in the range of 0 to 20 dBµV, a second segment is provided for a field strength in the range from 20 to 30 dBµV, and a third segment is provided for a field strength from 30 to 44 dBµV. Above 44 dBµV, additional increases in the field strength do not additionally increase the field strength quality value Q(FS).
It can be seen that the gradient of the first segment is steeper than that of the second segment, which is in turn steeper than that of the third segment. In other words, the non-linear relationship between the field strength of the received sound and the S/N ratio of the output sound is reflected by the different segments of the graph. The interpolation data of the graph is stored in the memory 6 associated with the CPU 4, for example as segmented interpolation algorithm(s). It is then used to determine a field strength quality value Q(FS) for a particular field strength by interpolation.
A further way of converting the detected field strength to a field strength quality value Q(FS) would be to formulate a non-linear equation reflecting the varying influence of the field strength on the signal-to-noise ratio (and hence on the overall quality), or to employ a separate linear equation for each field strength range (e.g. to employ different linear equations corresponding to the gradients of the segments shown in the graph of Fig. 5).
By calculating Q(FS) in one of these ways, and using the calculated Q(FS) value in the above equation (1), an audible sound quality value AQ which more accurately reflects the sound quality heard by the listener is realised. This is because the Q(FS) value more closely reflects the actual sound quality than the detected field strength value.
As outlined above, the values of the calibration parameters k, b and c are determined for a particular radio receiver/antenna combination and are loaded into the radio receiver before sale to an end user.
One mechanism for determining the values of these calibration parameters is through testing using a sample listener panel. The members of the panel are people who have a very good sense of hearing, for example those with 'perfect pitch'. In the testing, sound (e.g. music) is output through the radio receiver apparatus and is listened to by the members of the panel, for example whilst driving on a predetermined test route. At different points along the route, different amounts of noise and multipath will be present in the received signal and are measured by the respective detectors. The field strength is also measured. In addition, the quality of the output sound is rated by the members of the panel. By using the measured noise, multipath and field strength values, and the perceived quality values given by the panel, the optimum values of the parameters can be determined.
By using the equation (1) in which the audible sound quality AQ depends upon the field strength, multipath and noise according to a relationship determined using the quality perceived by a listener panel, it is ensured that the determined audible sound quality AQ accurately reflects the sound being heard in the vehicle.
In use, the fieldstrength detector, the multipath detector and the noise detector of the detectors 4 periodically monitor the fieldstrength, multipath and noise values of the received signal, i.e. of the frequency signal currently tuned to. The CPU 5 then calculates the value AQ from these detected values. For example, the CPU 5 takes the detected fieldstrength and looks up the corresponding field strength quality value Q(FS) in a look-up table stored in the memory 6. The CPU then retrieves the values of the calibration parameters from the memory 6 and calculates the audible sound quality AQ using the values of k, b, c, M, N and Q(FS).
In addition, the CPU 5 controls the tuner 2 so as to periodically tune the tuner 2 to the other alternative frequencies on which the programme tuned to is also being broadcast. Thus, the CPU 5 causes the tuner 2 to be tuned for a short amount of time to one of the alternative frequencies on which the same programme is being broadcast. In this regard, the alternative frequency is tuned to for a length of time short enough so as not to be noticeable to a listener, but long enough to sample the quality indicator values (e.g. field strength, noise and multipath) of the alternative frequency. This length of time should be less than 10 milliseconds. The CPU 5 controls the tuner 2 so as to tune to each of the alternative frequencies cyclically. In this way, sound quality information on each alternative frequency can be gathered in the background whilst the tuner is essentially, as far as the listener is concerned, tuned to one specific frequency (i.e. present, received frequency).
If the AQ value of the present frequency signal falls below a preset threshold, then the CPU 5 can consider whether or not to instruct the tuner 2 to switch to one of the other alternative frequencies on which the same programme is broadcast. In making the determination, the CPU 5 considers which of the alternative frequencies that has been sampled in the background has the best AQ value. This can be achieved by storing the AQ value of each alternative frequency in the memory 6 when that alternative frequency is sampled in the background. Thus, the CPU 5 can search the latest AQ values for the alternative frequencies stored in the memory 6. The CPU 5 can in this way select the alternative frequency having the best AQ value.
After selecting the alternative frequency having the best AQ value, the CPU 5 performs a comparison of the AQ value of this alternative frequency against the AQ value of the presently received frequency signal. The CPU 5 instructs the tuner 2 to switch the received frequency only if the AQ value of the alternative frequency is higher than the AQ value of the presently received frequency. In this way, switching to an alternative frequency having a lower quality than the presently received signal quality is avoided. Also, alternating switching between mute and noisy signals in a weak signal area is avoided, because the CPU 5 ensures that the tuner 2 remains tuned to the same frequency signal for as long as no alternative frequency having a better audible quality value AQ is available.
In addition, the CPU 5 may initiate a switching of the frequency that the tuner 2 is tuned to if the audible quality value AQ of one of the alternative frequencies sampled in the background is higher than the AQ value of the presently received frequency signal. The CPU 5 may cause the tuner to perform such switching even if the AQ value of the presently received signal is above the threshold. In this way, it can be ensured that the signal having the best quality is always tuned to. In other words, the optimum frequency is always selected from the available alternative frequencies. The radio receiver is able to switch up to twenty times per minute between different alternative frequencies.
In a preferred embodiment, the alternative frequencies and their respective quality values AQ are stored in a dynamically updated table in the memory 6. Each time the tuner 2 is briefly tuned to an alternative frequency to sample the quality parameters of that frequency, the table is updated. The alternative frequencies in the table are continuously re-ordered in order of their audible quality values AQ. Hence, the alternative frequency having the best AQ value is always the first row in the table. Other values my be stored in the table also, such as the multipath, noise, field strength and field strength quality values.
In addition to performing network following as above, the radio receiver also uses further measures to maintain satisfactory audio quality under various reception conditions. These further measures include mono stereo blend, high cut, soft mute and bandwidth control.
Mono stereo blend makes use of a gliding mono/stereo separation depending on signal level and multipath distortion. The CPU 5 controls mono stereo blend. As can be seen in the graph of Fig. 3, below a field strength of the received signal of about 35 dBµV the CPU 5 controls the output to be mono. Between about 35 and 50 dBµV stereo is opened until it reaches maximum stereo. In this way, distortions at low field strength are made less noticeable by outputting the audio in mono.
High cut is the reduction of higher audio frequencies in case of increasing noise and audio distortion, because the most annoying audio distortions for a listener are those in the high frequency band. To perform high cut, the CPU 5 causes the received signal to be passed through an adaptive low pass filter to reduce this part of the audio frequency spectrum. The cut off frequency and the suppression rate can be set by parameters which are evaluated during test drives.
Soft mute is an adaptive reduction of the audio volume at low field strength values where the signal-to-noise ratio is bad. Control of soft mute is performed by the CPU. The reduction of the audio level makes distortion and noise less annoying due to the properties of the human ear. Start and slope of soft mute is set by parameters and can be evaluated during test drives.
Bandwidth control becomes active if the IF filter is not able to suppress adjacent channel interference. In this case, spectrum overlapping occurs between the tuned-to channel and the adjacent channels. This happens mainly in regions with a channel grid of 100 kHz. The selectivity of the IF filter needs to be adaptive, so that the bandwidth of the channel filter can be reduced under the control of the CPU 5 if necessary. This bandwidth reduction results in a suppression of the adjacent channel signals, while keeping the distortion of the desired signal low.
If the CPU 5 causes the tuner 2 to switch from a presently-received frequency which has a weak signal to a new frequency among the alternative frequencies which has a strong signal, at the time of switching it will cause high cut and mono stereo blend to be applied so as to make the change smooth. In other words, the CPU 5 will use these measures to ensure that an abrupt change is not heard by a listener. The CPU 5 may also cause bandwidth control to be applied when performing such a switch.
In addition, in weak signal areas in which no alternative frequency is available having a satisfactory AQ value, the CPU 5 will optimise the sound output by using high cut, stereo blend and bandwidth control.
Figure 6 shows an example of frequency switching (network following) according to an embodiment of the invention. The lower portion of Figure 6 shows a vehicle travelling along a road and receiving the signals NDR, NDR AF1 and NDR AF2. In other words, the radio receiver of the vehicle can tune to any of NDR, NDR AF1 and NDR AF2 to receive the same programme. The upper part of Figure 6 shows the relative field strength values of the frequency signals NDR, NDR AF1 and NDR AF2 at different points along the road.
As can be seen from the lower part of the figure, at the start of the journey reception of the signal NDR is good. At a short distance further along, the field strength of the signal NDR AF1 becomes better than that of the signal NDR. At this point, a prior art receiver would cause a switching to be performed to the signal NDR AF1 because of its better field strength value. This is in spite of the signal NDR AF1 suffering bad distortion due to multipath and hence having a lower overall signal quality than the signal NDR.
According to the embodiment of the invention, however, in which the audible quality values AQ of NDR and NDR AF1 are compared, it is determined by the CPU 5 that the overall quality of NDR is better than that of NDR AF1. Hence, no switching is performed.
Further along the road, the signal NDR becomes disturbed. This leads to a reduction in the audible quality value AQ of the presently received signal NDR. This is noted by the CPU 5 which can determine from the continuously updated alternative frequencies table stored in the memory 6 that the audible quality value of the frequency signal NDR AF2 is better. Hence, at this point the CPU 5 causes the tuner 2 to switch from the signal NDR to the signal NDR AF2.
At this point, in the prior art receiver no switch is made from NDR AF1 to NDR AF2 because NDR AF1 has a better field strength (as can be seen from the top part of figure 6). Thus, the prior art receiver remains tuned to NDR AF1 despite its overall poorer signal quality due to the impact of multipath. The prior art receiver only switches to NDR AF2 further along the road, after the signal strength of NDR AF2 has become higher than that of NDR AF1.
Hence, it can be seen that by using the synergistic function of equation (1) to evaluate an overall signal quality value (namely the quality value AQ) for each alternative frequency, a radio apparatus embodying the invention is able to optimise the performance of frequency switching or network following.
According to a further embodiment, an offset detector is provided in the detectors 4 also. The offset detector measures the offset between channel frequency and tuned frequency. As the deviation between channel and tuned frequency is expected to be small in normal circumstances, a large offset (less than the channel grid of 100 kHz) indicates disturbance, such as interference from an adjacent channel. By using the offset detector, the CPU 5 can distinguish whether a received RDS PI code is assigned to the currently tuned frequency or to a neighbouring channel.
Figure 7 shows schematically the detection of offset by the offset detector. Here, the in-vehicle radio receiver is tuned to 97 MHz, on which a particular programme is being broadcast. This same programme is also being broadcast on an alternative frequency of 98.0 MHz, and the radio receiver obtains this information from the RDS data included in the RDS data link layer of the tuned-to signal at 97.0 MHz. The CPU 5 causes the tuner 2 to tune to the alternative frequency for a short AF check, i.e. for a brief period of time (< 10 ms) to evaluate the quality parameters of the alternative frequency. When the tuner 2 switches, however, it detects the unrelated signal at 98.1 MHz being broadcast by a different transmitter. At this stage, erroneous information about the signal strength and other parameters could be obtained, but the offset detector detects the offset and hence no network following is performed.
In addition, a pilot detector may be provided in the detectors 4. The pilot detector indicates the presence of a 'pilot tone'. If the pilot tone exceeds a certain threshold the detector output flag is 'set', if it stays below the threshold then the flag is 'reset'.
Still further, a pause detector may also be provided in the detectors 3. The pause detector indicates whether the received audio signal stays below a certain level threshold. If so, 'pause' is output for as long as this level condition is kept. The level threshold and the minimum pause time (the minimum time the audio has to be below the threshold until the detector signals a pause on its output) are adjustable. The pause detector can be used to determine the appropriate time instant to start an AF-update that shortly interrupts the audio. By performing the AF update while a pause is detected the AF update will almost not be audible.

Claims

A network following method of switching a radio tuner between different frequencies over which a same programme is broadcast, the method comprising:
tuning the radio tuner to a first frequency of a plurality of alternative frequencies on which a same programme is broadcast, so as to receive a first frequency signal of the programme;

periodically monitoring quality indicators of the first frequency signal;

periodically monitoring quality indicators of the signal of each of the other alternative frequencies; and

switching the radio tuner from the first frequency to a second frequency of the alternative frequencies in accordance with a quality value AQ determined from the monitored quality indicators for the first frequency signal and for the other frequency signals; wherein

the quality value, AQ, of each frequency signal is determined by applying a synergistic function of a field strength value, V(FS), and a distortion index, I_D, of the signal, where the distortion index I_D decreases as distortion D increases.
A method according to claim 1, wherein the synergistic function is a multiplication of the field strength value V(FS) and the distortion index I_D of the signal.
A method according to claim 1 or 2, wherein the field strength value V(FS) of each signal is a field strength quality value Q(FS) determined by:
detecting the field strength of the received signal; and

performing a non-linear conversion on the detected field strength to convert it into the field strength quality value Q(FS).
A method according to claim 3, wherein the non-linear conversion is performed based on a non-linear relationship between the field strength of the received signal and the signal-to-noise ratio of the audio output signal.
A method according to claim 4, wherein the non-linear conversion is performed using a segmented interpolation, said segmented interpolation including a first segment for a field strength range in which a high frequency cut is performed on the audio signal, and a second segment in which the high frequency cut is not performed on the audio signal.
A method according to claim 5, wherein the segmented interpolation includes a third segment for a field strength range in which the channel separation of the audio output signal is continuously increased from Mono to full Stereo, in accordance with the field strength.
A method according to any preceding claim, wherein the distortion D of the received signal is calculated from: $D = b * M + c * N,$

where M is a value of the multipath, N is a value of the noise, and b and c are calibration values previously determined for the radio apparatus.
A method according to any preceding claim, wherein the radio tuner is switched from the first frequency to the second frequency if the quality value AQ of the first frequency diminishes to a predetermined threshold.
A method according to any preceding claim, wherein the radio tuner is switched from the first frequency to the second frequency if the quality value AQ of the second frequency becomes higher than that of the first frequency.
A method according to any preceding claim, wherein the alternative frequencies and their current quality values AQ are stored in a table which is continuously updated, said alternative frequencies being continuously ordered and re-ordered in the table according to their respective quality values AQ.
A method according to any preceding claim, the method further comprising, when a said alternative frequency is tuned to, determining the offset value of the requested alternative frequency and the actual frequency tuned to.
A radio apparatus for in-vehicle use, the apparatus comprising:
a tuner for tuning the radio apparatus to a first frequency of a plurality of alternative frequencies on which a same programme is broadcast, and for periodically sampling each of the other alternative frequencies;

detecting means for detecting signal quality indicators of the signal of the first frequency and of the signals of the other alternative frequencies;

signal quality determining means for determining a quality value AQ from the detected signal quality indicators for the signal of the first frequency and for the signals of each of the other frequencies; and

switching means for switching the tuner from the first frequency to a second frequency of the alternative frequencies in accordance with the quality values AQ of the signals of the alternative frequencies; wherein

the signal quality determining means is operable to determine the quality value AQ of the signal of each frequency by applying a synergistic function of a field strength value, V(FS), derived from the field strength detected by the detecting means, and a distortion index I_D of distortion D due to noise and multipath detected by the detecting means, wherein the distortion index I_D decreases as the distortion D increases.
A radio apparatus according to claim 13, wherein the synergistic function applied by the signal quality determining means is a multiplication of the field strength value V(FS) and the distortion index I_D of the signal.
A radio apparatus according to claim 12 or 13, wherein the signal quality determining means is operable to determine the distortion D of each signal by calculating: $D = b * M + c * N,$

where M is a value of the multipath detected by the detecting means, N is a value of the noise detected by the detecting means, and b and c are calibration values previously determined for the radio tuner.
A radio apparatus according to any of claims 12 to 14, wherein the signal quality determining means is operable to calculate a field strength quality value Q(FS) as the field strength value V(FS) of each signal, by performing a non-linear conversion on the field strength detected by the detecting means so as to convert the field strength to the field strength quality value.
A radio apparatus according to claim 15, wherein the signal quality determining means is operable to calculate the field strength quality value Q(FS) using a segmented interpolation algorithm stored in memory, said segmented interpolation algorithm including segments for different field strength ranges in which the radio apparatus is operable to output the audio signal with different effects.
A radio apparatus according to any of claims 12 to 16, wherein the switching means is configured to switch the tuner from the first frequency to the second frequency if the quality value AQ of the first frequency diminishes to a predetermined threshold.
A radio apparatus according to any of claims 12 to 17, wherein the switching means is configured to switch the tuner from the first frequency to the second frequency if the quality value AQ of the second frequency becomes higher than that of the first frequency.
A radio apparatus according to any of claims 12 to 18, the apparatus further comprising an alternative frequency table stored in memory, wherein the alternative frequency table includes each alternative frequency and its respective quality value AQ, said alternative frequencies being continuously ordered and re-ordered in the table according to their respective quality values AQ.