GB2168878A - Radiocommunication system - Google Patents

Abstract

A radiocommunication system (CMR) includes several base stations (e.g. BS1 and BS2), a concentration station (CS), a digital telecommunication exchange (MTX) and mobile stations such as MS. The mobile stations (MS) are connected to the base stations (Bs1/2) via radio links. Each base station comprises an antenna connected to an electronic unit (ERU) including a transmitter, a receiver and a local control unit to control the operation of the transmitter and receiver. The electronic units (ERU) located in a predetermined area are all connected to the concentration station (CS) via analog telecommunication lines. The purpose of the concentration station is to convert analog (voice and binary data) signals coming from the electronic units into digital signals which are then multiplexed and transmitted on the digital telecommunication line to the digital telecommunication exchange (MTX) and vice versa. <IMAGE>

Description

SPECIFICATION Radiocommunication system This invention relates to a radiocommunication system for transmitting voice and binary data signals of the kind including at least one mobile station, a telecommunication exchange, at least one base station coupled to the mobile station via a radiocommunication link and to the telecommunication exchange via a telecommunication link, the base station being able to transform the signals into digital signals and transmit them on the telecommunication link A radiocommunication system of this kind is already known from European patent application No.

0.113.662-A2. If in such a system the binary data signals are transmitted by the mobile station as minimum-shift-keying (MSK) signals, also called fast frequency-shift-keying (FFSK) signals, i.e. by using two frequencies differing by a value equal to half the baud rate of the binary data signals, it may be difficult in the telecommunication exchange to distinguish between these two frequency signals. This is particularly true if these frequencies are shifted for one or other reason towards each other, i.e. in such a way that their difference or deviation is modified, and when the baud rate and hence the difference of the nominal frequency values is small.When the frequency recognition has to be performed within a single bit period of the binary data signal, additional difficulties arise when the duration of this period is of the same order of magnitude as that of a single cycle of the lowest frequency. In a practical embodiment, the frequencies are 1200 Hertz and 1800 Hertz and the baud rate of the binary data signals is equal to 1200.

It should be noted that a possible way of making a distinction between the two frequencies is to use filters but these are expensive and do not allow the frequencies to be recognised within a single bit period of the binary data signals.

An object of the present invention is to provide a radiocommunication system of the above kind but which permits, in the telecommunication exchange, an easy and fast recovery of the binary data signals generated by the mobile station and which makes use of a minimum of additional equipment to that normally used in this exchange.

According to the invention in its broadest aspect, a radiocommunication system of the kind referred to is characterised in that the mobile station is able to transmit the binary data signals on the radiocommunication link as frequency-shift-keying modulation signals using first and second frequencies, that the base station includes sampling means to sample the modulation signals and the voice signals and to transmit the thus-obtained sample signals on the telecommunication link, and that the telecommunication exchange includes a demodulator having noise-filtering means for the sample signals, oversampling means to derive additional sample signals from the sample signals, and zero crossing detection means to detect a change of polarity of the sample signals and the additional sample signals and to recover the binary data signals.

The sampling means are used for sampling both the binary data signals and the voice signals, and operate at a frequency chosen so as to permit the recovery of the voice signals from the sample voice signals. Such a simple recovery is not always possible for the binary data signals because the necessary noise-filtering means may introduce frequency shifts of such a value that the recognition of the frequencies by a zero crossing method becomes inaccurate in certain circumstances. However, by deriving additional sample signals from the already sampled binary data signals, such a frequency recognition is made possible in an easy way and with a minimum of additional equipment.

An embodiment of the invention will now be described by way of example with reference to the accompanying drawings, in which: Figure 1 illustrates a radiocommunication system CMR according to the invention, Figure 2 is a block diagram of a mobile radio trunk module MRT of Figure 1, Figure 3 shows signals appearing at various locations of the demodulator DD of Figure 2, Figure 4 is a block diagram of this demodulator DD, Figures 5a to 5d show portions of signals generated by the modulator MD of Figure 2.

The radiocommunication system or cellular mobile radio system CMR represented in Figure 1 includes several base stations such as BS1 and BS2, a concentration station CS, a digital telecommunication exchange MTX and mobile stations such as MS, e.g. radio telephones in cars.

The mobile stations MS are connected to the base stations BS1/2 via radio links. Each base station BS1/2 comprises an antenna connected to an electronic unit ERU including a transmitter, a receiver and a local control unit to control the operation of the transmitter and receiver. The electronic units ERU of the base stations BS1/2 located in a predetermined area, e.g. within a radius of about 15 kilometer, are all connected to the concentration station CS via analog telecommunication lines.

The concentration station CS includes analog-to-digital/digital-to-analog converters (not shown) which are connected on the one hand to the associated base stations BS1/2 via normal 2 x 600 ohms telephone lines, and on the other hand to a multiplexer/demultiplexer (not shown) also included in CS. The multiplexer of CS is itself connected to the digital telecommunication exchange MTX and more particularly to a mobile radio trunk module MRT thereof via a 32-channel digital telecommunication line (trunk). The purpose of the concentraton station CS is to convert analog (voice and binary data) signals coming from the electronic units ERU on the analog telecommunication line into digital signals which are then multi plexed and transmitted on the digital telecommunication line to the digital telecommunication exchange MTX and vice versa.MTX is similar to the digital telecommunication exchange S1240 described in "Electrical Communication" Volume 56, No. 2/3, 1981. It only differs from the latter in that it is provided with mobile radio trunk modules MRT and that its software is able to process radio communications. These mobile radio-trunk modules MRT are connected to the digital switching network DSN via terminal control elements TCE, both DSN and TCE being also described inthe last mentioned review.

The analog voice and binary data signals are transmitted from the mobile stations MS to the base stations BS1/2 on the radio link by frequency modulation. More particularly, the binary data signals are translated into two sinusoidal signals F1 and F2 representing the binary values 1 and 0 respectively and these two signals F1 and F2 are transmitted on the radioi link as minimum-shift-keying (MSK) modulation signals also called fast frequency-shift-keying (FFSK) modulation signals. The baud rate of the binary data signals is 1200 and the values of the frequencies of the above signals F1 and F2 are F1=1200 Hertz and F2=1800 Hertz respectively. A binary value 1 is thus equivalent to 1 cycle of F1 and a binary value 0 is equivalent to 1.5 cycle of F2.

After demodulation in the receiver of the electronic unit ERU the two signals F1 and F2 are transmitted in an analog way to the concentration station CS where they are sampled. These sample signals are then transmitted to the digital telecommunication exchange MTX and more particularly to the mobile radio trunk module MRTthereof via the digital telecommunication line. Each sample signal is coded into 16 bits which are transmitted as one channel of a time division multiplex signal TDM comprising 32 of such channels. Moreover, the TDM signal has-a frame rate of 125 microseconds.

As will be described in more detail below, the binary data signals are recovered from the incoming sample signals in the mobile radio trunk module MRT prior to being transmitted to the rest of the digital telecommunication exchange MTX.

In the other direction, binary data signals generated by MTX are translated by MRT into sample signals according to the two signals F1 and F2. These sample signals are then sent to the concentration station CS on the digital telecommunication line by a TDM signal similar to the one mentioned above. In this concentration station CS these sample signals are transformed into analog signals with corresponding frequencies F1 and F2 prior to being transmitted further to the base stations BS1/2. From the base station BS1/2 these analog signals F1 and F2 are sent to the mobile stations MS via the radio link by the frequency modulation technique mentioned above.

A simplified block diagram of the part of the mobile radio trunk module MRT which handles the above sample signals and the binary data signals is represented in Figure 2. MRT has six terminals COUT, CIN and TIN, TOUT, TCIN, TCOUT coupled to the concentration station CS and to the associated terminal control element TCE respectively and includes a demodulator DD, a modulator MD, a control unit CU and two electronic switches S1 and S2. The control unit CU controls S1 and S2 and is connected to MD, to DD via terminal DOUT thereof and to DSN via input terminal TCIN and output terminal TCOUT of the TCE.The terminal COUT is connected to the demodulator DD via terminal DIN and the modulator MD is connected to a terminal D of the change-over switch S2 which also has a common terminal C connected to the terminal CIN and a terminal S to which the terminal TOUT is connected. The switch S1 has a common terminal C to which the terminals COUT and DIN are connected, and a terminal S connected to the terminal TIN.

When voice signals are transmitted from MTX to the concentration station CS, the change-over switch S2 is brought into position S by the control unit CU. The voice signals applied to the terminal TOUT are then transmitted to the terminal CIN via the terminals S and C of the change-over switch S2. When binary data signals are transmitted from MTX to the concentration station CS, the change-over switch S2 is brought into position D by the control unit CU. The binary data signals applied to the terminal TCOUT are then transmitted to the control unit CU which sends them, at appropriate moments, to the terminal CIN via the modulator MD and the terminals D and C of the change-over switch S2.Finally, when voice signals or binary data signals are transmitted from the concentration station CS to the terminals COUT, they are applied to the control unit CU via the terminal DIN, the demodulator DD and the terminal DOUT.

According to the type of signals, voice or binary data, the control unit CU closes or opens the switch S1 which is in fact a "mute switch" preventing the binary data signals from being transmitted to MTX.

The operations of the demodulator DD and of the modulator MD is described in more detail hereinafter Reference is hereby made to Figure 3 wherein line a shows the sampling rate of frame clock pulses of the time divisioin multiplex TDM signal. The TDM signal has frames T1 of 125 microseconds subdivided into 32 channels of 16 bits each, as already mentioned. The binary data signals are represented in a digital way in line e and in an analog way in line f. Lines 3e and 3f show an example of alternative binary values 1 and 0 of which the corrsponding signals F1 and F2 are schematically represented in an analog way and in a digital way (sample signals) in the lines b and c respectively. It is to be noted that these successive signals F1 and F2 have a continuous phase relationship. Since the TDM signal has a frequency of 8 kilo-Hertz (T1 = 125 microseconds) the duration T2 of a binary value is equal to 8/1.2 = 6.666 frames of the TDM signal or 6.666 x 125 = 833 microseconds. Hence, the length ofthree consecutive binary values (3 x T2) is exactly equal to that of 20 frames of the TDM signal.

The extraction of the binary data signals (lines e and f) from the TDM signal (line c) is performed by only two demodulators such as DD (Figure 2) handling the even and odd numbered channels respec tively. Thus, the operation of each demodulator DD mainly consists in collecting the sample signals of the TDM signal during a channel and in translating these sample signals into binary data signals during the next following channel while the other demodulator is then collecting the sample signals of the TDM signal. The operation of the demodulator which is able to decode the FFSK signal after one cycle of the signals F1 or F2 is described hereinafter by making reference to the block diagram of the demodulator DD shown in Figure 4, and to Figure 3.As already mentioned, DD has an input terminal DIN connected to the terminal COUT of the mobile radio trunk module MRT (Figure 2) and an output terminal DOUT connected to the control unit CU or MRT (Figure 2). The sample signals applied to DIN by CS successively pass through a noise-filtering circuit FC and an oversampling circuit OS. At the output of the oversampling circuit OS, the sample signals are simultaneously applied to a zero crossing detection circuit ZC, a level detection circuit LV and a baud rate synchronisation circuit BD. The output of the zero crossing detection circuit ZC is connected to the input of a decision circuit DC and the output of the level detection circuit LV is connected to another input of the baud rate synchronisation circuit BD.The signals at the respctive outputs of the decision circuit DC and the baud rate synchronisation circuit BD are combined in a gating circuit GC whose output is connected to the terminal DOUT.

The noise-filtering circuit FC is a band pass filter for frequencies ranging from 600 to 1800 hertz and is used to reduce the noise level of the incoming sample signals at terminal DIN. This noise level may be very high particularly due to the radio transmission between the mobile stations MS and the base stations BS1/2. After filtering, the sample signals schematically represented in line c of Figure 3 are transmitted to the oversampling circuit OS which doubles the number of sample signals by generating an additional sample signal occurring in the middle of two consecutive sample signals and having a value equal to the average of the latter sample signals. The sample signals appearing at the output of the oversampling circuit OS are schematically represented in Figure 3 at line d. In this way the original 8 kilo Hertz sample signals are transformed into 16 kilo-Hertz sample signals.The necessity of this oversampling will be described later.

The sample signals (line d of Figure 3) at the output of OS are then supplied to the zero crossing detection circuit ZC which is adapted to detect changes of polarity of these sample signals and then to generate zero crossing output signals. Such polarity changes for instance occur in points Z1 and Z2 where the sample signals cross the zero or X axis. These zero crossing output signals are then transmitted to the decision circuit DC which by counting the number of samples betwen such two zero crossing signals is able to determinate the frequency F1 or F2 of the sample signals. The decision circuit DC then generates binary data signals such as the ones shown at line e in Figure 3 but delayed with respect to the corresponding sample signals.This delay (not shown in Figure 3 lines e and f) is obviously due to the fact that the decision circuit DC can only generate a binary data signal after the determination of the frequency F1 or F2, i.e. during the next following channel.

The sample signals at the output of the oversampling circuit OS are also supplied to the baud rate synchronisation circuit BD and to the level detector LV. The latter detector generates a signal indicating whether the level of the received sample signals is sufficient or not and supplies this indication signal to the baud rate synchronisation circuit BD. As a function of the value of this indication signal the baud rate synchronisation circuit BD generates a corresponding validation signal which is synchronised with the sample signals received from the oversampling circuit OS. This validation signal is applied to the gate circuit GC which, accordingly, blocks or allows the transmission of the output signals of DC to the termi nal DOUT.When receiving such output signals from DC the control unit CU searches for the occurrence of a predetermined sequence of binary values knowing that such a sequence always immediately precedes a predetermined series of valid binary data signals. When it detects such a sequence, CU opens the switch S1 (Figure 2) and sends the binary data signals to MTX via the terminal TCIN. After transmission of the predetermined series of binary data signals, CU again closes the switch S1 expecting then voice signals to be transmitted and resumes its search for the above-mentioned predetermined sequence of valid data signals.

The reason for the oversampling mentioned above is explained hereinafter. The decision circuit DC which is not shown in detail is adapted to count the number of sample signals between two successive zero crossings indicated by ZC, i.e. after one half cycle of the signal F1 or F2 while the validation of the frequency only occurs after a complete cycle to decrease the probability of an error. Indeed, with the above given values the following Tables 1 and 2 may be calculated. These tables give the frequencies in the range to which F1 (1200 Hertz) and F2 (1600 Hertz) belong, corresponding to an integer number of sample signals in the case of sampling and oversampling respectively.These frequencies are calculated with the formula sample signal frequency frequency = of samples TABLE 1 for a 8 kilo-Hertz sample signal number of frequency samples (Hertz) 4 2000 5 1600 6 1333 7 1143 TABLE 2 for a 16 kilo-Hertz sample signal number of frequency samples (Hertz) 8 2000 9 1778 10 1600 11 1455 12 1333 13 1231 14 1143 First Table 1 is considered. From this table it follows that with sample signals of 8 kilo-Hertz, the signal F1 (1200 Hertz) is represented by 6 or 7 samples, whilst the signal F2 (1800 Hertz) is represented by 4 or 5 samples. This uncertainty over the number of samples is due to the fact that T2 (Figure 3) is not a multiple of T1 so that the phases of the signals F1 and F2 have an undefined position with respect to the frame clock pulses of the sample signals.Moreover, these phases may also be altered by the noise filtering circuit FC since when the FFSK signals swap from F1 to F2 this circuit FC cannot instantaneously follow this change of phase. The original signal F2 is then delayed at the beginning of its cycle and, since FC tries to recover the correct frequency F2 (1800 Hertz), the signal is compressed at the end of the cycle.

As a result, the signal at the output of FC is distorted with respect to the original signal F2. A similar frequency distortion may also occur when the FFSK signals change from F2 to F1. In some circumstan-ces, the signal F1 may be so distorted that it is represented by less than 6 samples and in that case the decision circuit DC is no longer able to recognise the signal F1 and may interprete it as F2. The same is also true if the number of samples representing F2 is greater than 5. From Table 1 it follows that the signal F1 (1200 Hertz) may still be recognised by the decision circuit DC when it reaches a maximum frequency of 1333 Hertz whilst the signal F2 (1800 Hertz) may still be recognised when it reaches a minimum frequency of 1600 Hertz. Within the range between 1333 and 1600 Hertz the response of the decision circuit DC is unpredictable.

Table 2 is now considered and shows that by oversampling the tolerance on the distortion of the- signals F1 and F2 is increased. Indeed, if in this case, 10 or less than 10 samples are counted, the frequency is F2 (1800 Hertz), whilst if more than 10 samples are counted, the frequency is F1 (1200 Hertz). The accuracy is thus increased and the probability of an error is decreased because the additional sample (11) corresponding to a frequency of 1455 Hertz is created between the frequencies 1333 and 1600 Hertz.

It is to be noted that to perform their functions the zero crossing detection circuit ZC, the decision circuit DC and the baud rate synchronisation circuit BD all make use of software programs called finite state machines (FSM, not shown).

As described above for binary data signals transmitted from MTX to the mobile stations MS, use is made of the modulator MD which forms part of MRT and mainly includes a Programmable Read Only Memory PROM and an associated control device (not shown). This PROM stores the values of sample signals of the two signals F1 and F2 as shown in Figure 5. The PROM is constituted by 4 tables (a/d) each containing 3 groups A, B and C of sample signals. The two first groups A and B each stores 7 samples and the third group C stores 6 samples so that these three groups contain a total of 20 samples. As already mentioned 20 sample signals correspond exactly to 3 binary data signals so that the signals generated by means of the samples stored in the PROM may be resynchronised by the 8 kilo-Hertz clock of the TDM signal after each occurrence of 20 sample signals. Each group of samples represents a signal F1 or F2 and the first sample signal of each of the groups of a same table has the same phase polarity.

More in detail, the 4 tables are designed as follows: - the first table (a) contains samples of the signal F1 starting with a positive polarity (F1 +); - the second table (b) contains samples of the signal F1 starting with a negative polarity (F1-); - the third table (c) contains samples of the signal F2 starting with a positive polarity (F2+); - the fourth table (d) contains samples of the signal F2 starting with a negative polarity (F2-).

According to the value of the binary data signal supplied to the terminal TCOUT of MRT (Figure 2) and transmitted to the PROM thereof via the control unit CU, and to the polarity of the last sample of the previous binary data signal (continuous phase-FFSK), the control device of the modulator MD retrieves a next sample signal in the PROM and sends it to the concentration station CS via the terminal CIN and the change-over switch S2. The operation of the modulator MD will become more clear from the following example wherein it is supposed that the binary data signals have the successive binary values 1, 0 and 1.

The first samples (1) collected from the PROM are the samples of the group A represented at table a in Figure 5, i.e. they represent about 1 cycle of the signal F1 starting with a positive polarity Fl + (assuming that the last sample of the previous signal had a negative polarity). Since this group A of Fl + ends with a sample of negative polarity, the second group of samples (0) must start with a positive polarity to ensure the phase continuity. Moreover, this group must represent about 1.5 cycle of F2. It thus corresponds to the group B of Figure 5 table c representing F2+. The third group of samples (1) is again about 1 cycle of F1 but must now start with a negative polarity F1- since the last sample of the previous signal had a positive polarity. Thus this last group of samples is the group C of Figure 5 table b.

Claims (11)

1. Radiocommunication system (CMR) for transmitting voice and binary data signals of the kind including at least one mobile station (MS), a telecommunication exchange (MTX), at least one base station (BS1/2, CS) coupled to the mobile station via a radiocommunication link and to the telecommunication exchange via a telecommunication link, the base station being able to transform the signals into digital signals and transmit them on the telecommunication link, characterised in that the mobile staton (MS) is able to transmit the binary data signals on the radiocommunication link as frequency-shift-keying (FSK) modulation signals using first (F1) and second (F2) frequencies, that the base station (Bs1/2, CS) includes sampling means to sample the modulation signals and the voice signals and to transmit the thus-obtained sample signals on the telecommunication link, and that the telecommunication exchange (MTX) includes a demodulator (DD) having noise-filtering means (FC) for the sample signals, oversampling means (OS) to derive additional sample signals from the sample signals, and zero crossing detection means (ZC, DC) to detect a change of polarity of the sample signals and the additional sample signals and to recover the binary data signals.

2. Radiocommunication system according to claim 1, characterised in that the frequency of the additional sample signals is twice the frequency of the sample signals which is the frequency required to recover the voice signals from the sample voice signals.

3. Radiocommunication system according to claim 1, characterised in that the duration of one bit period of the binary data signals is substantially equal to one cycle of the lowest (F1) of the first (F1) and second (F2) frequencies.

4. Radiocommunication system according to claim 1, characterised in that the frequency-shift-keying (FSK) modulation is a minimum-shift-keying (MSK) modulation B whereby F2 - F1 = 2 wherein F2 is the second 2 frequency, F1 is the first frequency and B is the baud rate of the binary data signals.

5. Radiocommunication system according to claim 1, characterised in that the zero crossing detection means (ZC, DC) include means for counting the number of sample signals and additional sample signals between two zero crossings.

6. Radiocommunication system according to claim 5, characterised in that the means for counting the sample signals and the additional sample signals are software finite-state-machines (FSM).

7. Radiocommunication system according to claims 2 and 4, characterised in that the first frequency F1 is 1200 Hertz, the second frequency F2 is 1800 Hertz the baud rate B is 1200 and the frequency of said sample signals is 8 kilo-Hertz.

8. Radiocommunication system according to claim 1, characterised in that the sample voice signals and the binary data signals are transmitted in time channels on the telecommunication link and that a first demodulator (DD) receives the sample signals and the additional sample signals during a time channel while simultaneously a second demodulator (DD) recovers the binary data signals from the sample signals and the additional sample signals received during a preceding time channel.

9. Radiocommunication system according to claim 1, characterised in that the telecommunication exchange (MTX) also includes a modulator (MD) to transform binary data signals generated by the exchange into sample signals representing the first (F1) and second (F2) frequencies and corresponding to the binary data signals.

10. Radiocommunication system according to claim 9, characterised in that the modulator (MD) includes a Programmable Read Only Memory (PROM) for storing the values of the sample signals.

11. Radiocommunication system substantially as described with reference to the accompanying drawings.