GB2137052A - Improvements in or Relating to the Control of Mobile Radio Communication Systems - Google Patents

Abstract

A radio-telephone communications system with a plurality of selectable base stations TR for relaying speech and data radio signals to mobile units is controlled by channel monitors 3 and a master controller 4. The monitors 3 are caused to store digital data describing the quality of audio signals provided by the associated base stations TR in terms of peak and mean values of the amplitude probability density histogram and the master controller 4 interrogates the monitors 3 to select the base stations TR having the best- quality received incoming signal. <IMAGE>

Description

SPECIFICATION Improvements in or Relating to the Control of Mobile Radio Communication Systems The present invention relates to mobile radio communications systems and, more particularly, to the control of multiple base stations in radiotelephone systems.

Various problems arise with the allocation of carrier frequencies to different base stations in an attempt to increase the area covered by mobile radio transmission with switchable channels.

Conversely, by operating on one common carrier frequency different base stations will exhibit different reception qualities and interference can occur. There is a need for improved control which will automatically select the best-received signal.

In accordance with the present invention, a mobile radio communications system includes a plurality of base stations-receivers or transmitter/receivers-and employs a control method which monitors audio signals from all the base stations simuitaneously, provides digital data describing the quality of the audio signals and enables the comparative assessment of such data for the selection of the line or channel with the best quality signal. The digital data describing the signals is derived by a technique which ascertains the amplitude probability density of each signal. The description data is broadly equivalent to the signal to noise ratio of the corresponding reception.

The control method preferably utilises a master controller which controls the operation of channel monitors which receive the individual audio signals, conveniently e.g. via telephone lines or UHF microwave links, from the base stations. The master controller can cause each channel monitor to perform a sequence of sampling operations designed to provide 8-bit or 1 6-bit digital "words" in a memory describing its associated audio signal in terms of statistical parameters.

The master controller can then vote or select the best quality signal and can switch channels to provide optimum results subject to certain constraints.

In contrast to analogue or carrier-based prior art selection systems control in accordance with the present invention provides a number of advantages. Different types of base station and transmissions can be used in the overall system and no modification is necessary to the base station since the control is effective at the audio output stage. Because the control is effected at a central location the geographical location of the base stations does not affect the assessing operation and the degradation of signals over different lengths of telephone line will be also assessed.

Control means in accordance with the invention comprises channel monitors for receiving audio signals from a number of stations, each channel monitor having a channel processor which converts its audio signal into digital signals and performs operations on the digital signals to ascertain the the shape of the amplitude probability density histogram of the audio signal.

Preferably the shape is ascertained by stored "words" representing the peak and mid-quartile mean values of the signal's histogram. A master controller provides digital control signals for all the channel monitors and serves to compare the stored words to assess the best-received signal.

The channel monitors and the master controller intercommunicates via a digital and an audio bus. Each channel monitor has further devices to identify the nature of the audio signal i.e. whether speech, noise, data or an optional "receiver muted" signal tone. The channel monitors and master controller preferably employ microprocessors programmed to perform, inter alia, the aforesaid functions. This enables the control means and the communications system to be readily adapted to said particular needs.

The invention may be understood more readily, and various other aspects and features of the invention may become apparent, from consideration of the following description.

An embodiment of the invention will now be described, by way of example only, with reference to the accompanying drawings wherein: Figure 1 is a block diagram of a typical mobile radio communications systems utilising control means constructed in accordance with the invention.

Figure 2 is block diagram illustrating part of the control means made in accordance with the invention; Figure 3 is a more detailed representation of the master controller of control means; Figure 4 is a more detailed representation of one of the channel monitors of the control means; and Figures 5 to 8 are curves depicting various relationships pertaining to the invention.

As shown in Figure 1 a number of transceivers TR, which may be in fixed locations, receive data and/or speech information (FM or AM) by radio transmissions. The transceivers TR would normally operate at similar frequencies and are notionally associated with respective channels 1, 2 and 3. Three such transceivers TR are shown merely by way of example. The demodulated or detected information is conveyed as audio signals to a central control means 5 which serves to select the best-received signal to relay to one or more users or other equipment via output lines 20.As illustrated, the audio signals are fed to the control means 5 via cables 11, 12, 1 3 but other methods such as further radio links can replace the cables 11, 1 2, 1 3. As is known, the transceivers TR may provide recognisable muting tones when quiescent i.e. when not receiving radio transmissions. The control means 5 comprises individual channel monitors 3 and a master controller 4. The monitors 3 and the controller 4 inter-communicate via a common audio bus 6 and a common digital bus 7.

Figure 2 depicts in schematic form one of the channel monitors 3 in conjunction with the controller 4. Figure 3 is a more detailed circuit representation of the controller 4 of Figure 2 and Figure 4 is a more detailed circuit representation of one of the monitors 3 of Figure 3. As shown in Figure 2, one monitor 3 is provided for each channel and each monitor comprises a channel processor 21. As shown in Figure 2, the channel audio input, e.g. from the associated cable 11, 12, 1 3 passes via a 600 ohm balanced transformer TK to an amplifier 10 and a "Vogad" 1 5. The amplifier 10 and Vogad 15 provide a predetermined amplitude signal for further processing. The amplifiers 10 and Vogads 1 5 for all the channels are designed to provide the same predetermined amplitude signal for all the channels.This audio signal is then assessed as to whether it is the mute tone or information. A narrow band pass filter 16 selects the mute tone from the output of the Vogad 15 and feeds a detector 1 7 which provides a signal e.g. to inhibit selection of the the channel processor 21. The channel processor 21 itself can take the form of an integrated circuit, more particularly, an EPROM microprocomputer device of the type MC68705 R3 (Motorola). Further filters 1 8, 1 9 select the expected speech and data information from the full bandwidth of the channels and signals are generated which are indicative of the presence of the data or speech. The audio-information signal, whether speech of data, is fed to an analogue/digital converter 21' which is part of the channel processor 21 and the digital signals are processed as described hereinafter.The analogue signal fed to the converter 21 ' of the processor 21 is also fed to a data channel composed of a data filter 22 and amplifier 23 and a switch 24 and to a speech channel composed of an amplifier 25 and a switch 26. In the presence of data or speech the appropriate switch 24, 26 can be gated by control signals from the processor 21 under control of the controller 4 to allow the data to be fed to the audio bus 6 if the channel in question is the best-selected one.

An additional detector 27 senses data information and provides an alert signal to the processor 21.

The master controller 4 is composed of a control processor 28 and a control signal generator 29 which provide digital control signals to all the channel processors 21 via the digital bus 7. A microcomputer device MC 6802 (Motorola) may serve as the processor 28 and the generator 29. As mentioned previously the data or speech from the best-selected channel ts passed to the audio bus 6 and thence passes via amplifiers 30 and 600 ohm balanced transformers T2K to the output lines 20 as speech data or combined outputs.

The processor 28 and generator 29 provide digital signals to the individual channel processors 21 to cause the processors 21 to perform a sequence of operations under the controlof programmes. Each processor 21 is caused to ascertain the amplitude probability density of the signal feeding the analogue to digital converter 21' unless inhibited by the detection of the mute tone.

In essence, each signal is assessed according to a number of pre-determined amplitude levels, the information is stored in digital form in a memory of the processor 21 and values are calculated to represent the shape of the associated Amplitude Probability Density Histogram. The values are taken as an indication of the signal to noise ratio of the associated transceiver TR. This sequence occurs during a sampling mode during which the control processor 28 serves to cause the channel processors 21 to sample their signals and store 8bit words describing the signal. A typical sample period would be about 50-70 milli seconds wherein some 400 samples may be taken. The control processor 28 then interrogates each channel processor 21 in sequence to select the channel processor 21 with the stored values indicating the best signal to noise ratio.The control processor 28 then performs a control operation typically in 20 milli seconds and in accordance with one or more control algorithms stored as a programme and the best-selected signal can be routed to the lines 20. The analysis of the signals will now be described in more detail. The audio signals fed to the individual channel monitors 3 may have widely differing mean voltage levels and the "Vogads" 1 5 normalise the signals to a known amplitude level so that the noise content of the signal can be taken as its relative signals to noise ratio. The normalised signals are then analysed by the respective processors 21. Each signal is sampled at a sampling rate of 8 KHz by the converter 21' and the samples are represented as 8 bit words with a resolution of 1 9.5 mV per increment.As mentioned, 400 samples are taken and stored every sampling period and these samples are evaluated by a process known as "binning".

Basically for each 19.5 mV increment of input signal possible within the range -0.62 to +0.62 volts, an 8 bit wide memory location (or "bin") is reserved, giving 64 bins in total. For each occurrence of a voltage corresponding to a particular bin, the contents of that bin are incremented by one. Voltages outside the range -0.62v to +0.62v do not cause an increment and hence their occurrence is ignored. After the 400 samples have been taken the process is complete and the result is a stored histogram of Input Voltage against number of occurrences of that voltage. Since it is unlikely that any one bin will contain the whole 400 samples (i.e. a D.C. level or a signal of amplitude less than 19.5 mV, including Noise) it is possible to limit the contents of each bin to 255 (a convenient 8 bit definable number).

In the event of a bin reading 255 the next bin (having an address higher by one) is incremented to assure that the total number of increments received by the 64 bins is preserved. A sample space graphically depicted in Figure 5 is thus established.

A general statistical theorem, known as the Central Limited theorem, states that the probability distribution of any quantity which arises as the sum of the effects of a large number of separate contributions is the Gaussian or Normal distribution. Electrical Noise is one such quantity and the processors 21 produce such a distribution when presented with electrical Noise.

It can be expressed by the function:

where a=Standard Deviation and (r2=x2=Root means square value of the original waveform So er2=Vnoise (Rms) volts.

For a standardised Normal Variable the maximum value of F(x) can be described by:

Therefore for any normal distribution Peakoc 1/cr or

Peakoc y-Vnoise Hence the "Peak" as calculated can be used as a measure of the noise content of a waveform.

Figure 6 depicts various histogram waveforms.

From the stored histograms in the processors 21, two values known as the "Peak" and "Quartile mean" can be derived and stored for each stored histogram in order to obtain comparative measure of the noise content of the corresponding input signal. The peak value is calculated using a "moving window" technique designed to smooth the curve of the histogram, thus reducing the likelihood of a spurious result from influencing the selection procedure. A 4-bin wide "window" is moved across the total 64 bins in increments of one bin. For each increment the sum of the 4 bins is then taken as the "peak" value. To evaluate the quartile mean, the contents of all 64 bins is divided by 2. This gives a measure of the total area beneath the curve between the voltage limits.

The master control processor 28 reads and stores the values for peak and quartile means for each channel processor 21. If the value falls in a certain region of Peak/Quartile mean relationship the processor 28 assumes the signal is of a type associated with that region.

Figure 7 illustrated how different signal types, with varying quantities of noise have unique ranges of values for the peak and quartile mean associated with them. It is this feature which enables the type and comparative noise content of a signal to be determined. The "undefined" areas are included to ensure that a signal which inhabits a shared region of PeaWQuartile means relationship is not allowable to feature in the signal type decision process. This process is a democratic one i.e. for each channel it is decided whether the signal is speech tone, noise only or an undefined signal and the type is taken to be the one represented by the most channels.Once the signal type has been defined the control programme can select which channel is presenting the lowest noise by comparing peak values in the case of speech or quartile means values in the case of tone. Syllables themselves induce a low peak. However, when the syllable has finished the "gap" induces a high peak (with a fast attack) which then slowly decays due to the action of the Vogad 1 5. The arrival of the next syllable rapidly forces the peak back down to a low level. Comparison of clean and noisy Histogram shows that the "word" peaks are not significantly different, whereas the value of gap peaks are much greater for clean than for noisy speech. The control processor 28 software analyses the gap peaks of all the channels and selects the one which persistently portrays the highest value and consequently the iowest noise.

Thus, with speech the selection is effectively made by comparing the residual channel noise occuring during inter-word or inter-syllabic gaps.

The twin peaks of the histogram of a tone (see Figure 6b) is a characteristic of the probability analysis of a waveform of a repetitive nature.

However, the Vogad 1 5 is adjusted such that the peaks occur outside the voltage cut-off points, leaving only the centre "trough" for analysis. This has the effect that for a good quality signal the majority of occurrences arise with the two peaks and consequently less occurrences are found in the analysed region, thus reducing the quartile mean.

The control means can provide a variety of visual and/or signal indications to a user which are not shown in the drawings and the control processor 28 and/or the channel processors 21 can be programmed to cope with a variety of user demands of which the following are just typical examples.

1. Bias can be provided as between channels so that for equal equivalent signal to noise ratio on two channels one is given preference.

2. Hysterisis can be provided as between channels so that channel change over will not occur unless the difference between the equivalent signal to the noise ratios is greater than some pre-determined threshold.This effect can bs accomplished by artificially enhancing the parameters describing the currently selected channel. A certain percentage of the original peak and quartile the mean values associated with the selected channel are either added in the case of peak or subtracted in the case of mean, to make the channel currently selected hold an advantage in the "best value" comparison procedure. The percentage can be varied over the range 0.6dB.

3. Data information may be locked so that channel switching will not occur if certain data is present.

4. Emergency or coded data information may be given preference at all times.

5. Confidential channels can be established and maintained as such.

6. Where a selected base station is to used also as a transmitter in the case of "talk-through" automatic keying can be provided.

7. Afacility may be provided whereby a mute tone signal is not required, because the system is able to detect the absence of speech or tone modulation. If no modulation is detected, on a given channel then no further selection is made on that channel. This may be accomplished by using the peak values of the Amplitude probability Histogram, to form a second histogram, in which the relative frequency of occurrence of a given peak determines the height of a given ordinate.

Pure noise signals may have low peaks (i.e. a muted receiver) or high peaks (i.e. line noise) but successive peaks of a given noise level tend to be closely located around an average value. Only modulated signals give large variations in successive peak values. Thus the unmodulated signal may be identified by setting a threshold level for peak frequency as represented in Figure 8.

Claims (14)

1. A method of controlling a mobile radio communication system by simultaneously monitoring signals provided by base stations, describing the quality of the signals by digital data and assessing the data to select the base station providing the best quality signal.

2. A method according to claim 1, wherein the digital data is derived by assessment of parameters associated with the amplitude probability density of the signals.

3. A method according to claim 1, wherein the signals are sampled repetetively to provide 8-bit or 16-bit digital words describing the signals in terms of statistical parameters.

4. A method according to claim 1 , wherein the digital data describing the signals represents the shape of the amplitude probability density histogram of each of the signals.

5. A method according to claim 4, wherein the digital data describing the signals represents the peak and mid quartile mean values.

6. A method according to any one of claims 1 to 5, wherein the monitored signals are audio signals.

-

7. A method according to any one of claims 1 to 5 and further comprising assessing whether the signals provided by the base stations represent speech, tone, noise or data information.

8. A method of controlling a mobile radio communications system substantially as described with reference to any one or more of the Figures of the accompanying drawings.

9. A control means for a mobile radio communication system comprising base stations in the form of receivers or transmitters/receivers; said control means comprising channel monitors for receiving signals from the stations and for converting the signals into digital data describing the quality of the signals, and a master controller for interrogating the channel monitors to assess the best-received signal and enable the selection of the appropriate base station.

1 0. Contrdl means according to claim 9, wherein the channel monitors each employs a channel processor which stores digital words representing the peak and mid-quartile values of an amplitude probability histogram of the associated signal under control of the master controller.

11. Control means according to claim 9 or 10, wherein the channel monitors employ devices serving to normalise the amplitudes of all the signals for their conversion.

12. Control means according to any one of claims 9 to 11, wherein the channel monitors receive audio signals from the base stations, the channel monitors and the master controller intercommunicate via a common digital bus and a common audio bus and the master controller connects the audio bus to outputs to interconnect the selected base station to said outputs.

1 3. Control means according to claim 12, wherein each channel monitor has devices to identify the nature of the signals provided by the base stations and to gate the signals to the audio bus.

14. Control means according to any one of claims 9 to 13, wherein the digital data is derived as described herein with reference to Figure 5.

1 5. Control means according to any one of claims 9 to 14, wherein the channel monitors and/or the master controller utilise replaceable pre-programmed or programmable integrated circuit devices.

1 6. A radio communications system or control means therefor substantially as described with reference to, and as illustrated in any one or more of Figures 1 to 4 of the accompanying drawings.