WO2001031932A2

WO2001031932A2 - Method and receiver for processing a signal generated according to the multi-frequency dialing method

Info

Publication number: WO2001031932A2
Application number: PCT/DE2000/003751
Authority: WO
Inventors: Gonzalo Lucioni
Original assignee: Siemens Aktiengesellschaft
Priority date: 1999-10-28
Filing date: 2000-10-24
Publication date: 2001-05-03
Also published as: DE19952027C2; DE19952027A1; WO2001031932A3

Abstract

The invention relates to a method for processing a signal generated according to the multi-frequency dialing method. According to said method, the signal is scanned and a time-discrete sequence of scanned values (xn) is produced. On the basis of said scanned values, frequency and amplitude dimensions(FM1, FM2, AM1, AM2) are detected, which correspond to a low frequency group (LG) or a high frequency group (HG). Said dimensions (FM1, FM2, AM1, AM2) are compared with predetermined set values. If the detected dimensions (FM1, FM2, AM1, AM2) match the set values, the signal is then identified as a signaling information carrier.

Description

description

Method and receiving device for processing a signal generated according to the multi-frequency selection method

The invention relates to a method for processing a signal generated by the multi-frequency selection method and a receiving device for performing the method.

The multi-frequency dialing method is a signaling method in which signaling characters such as e.g. Dialing digits, hereinafter referred to briefly as DTMF characters, are transmitted via analog voice channels. Each DTMF symbol is defined by the combination of two tone signals, the frequencies of which come from two different frequency groups. Four frequencies are contained in each of the lower and m of the upper frequency group. The lowest frequency of the upper frequency group is larger than the highest frequency of the lower frequency group. The maximum number of DTMF signs that can be displayed results from the number of possible combinations according to which distinguishable frequency pairs can be formed, each consisting of a frequency of the lower frequency group and a frequency of the upper frequency group. With four frequencies and frequency group, sixteen DTMF characters can be displayed.

Before the DTMF signal transmitted with the signal can be processed further, the signaling information on which the character is based must first be recovered from the signal in the receiving device. This is usually done taking into account specified target values, which are specified for DTMF signals, for example, by the Q.24 standard of the International Telecommunication Union, ITU-T for short. For the reception of DTMF signals, methods are known from the prior art which are based on comparatively complicated algorithms, so that the receiving devices which operate on them are technically complex. So need so- most conventional receiving devices operating terbänken with bandpass Fil ^¬, as well as those that perform a digital Fourier transform in accordance with the Goertzel algorithm, eight individual recipients that are assigned to the distributed to the two frequency groups of eight frequencies. This high level of technical complexity has hitherto made it difficult to design such a receiving device by means of a microprocessor with a comparatively simple processor architecture, such as is frequently used in digital private branch exchanges, for example.

The object of the invention is to provide a method or a receiving device operating according to this method, with which a DTMF signal can be received and processed efficiently with less technical effort than before.

The invention solves this problem by a method for processing a signal generated by the multi-frequency selection method, in which the signal is sampled and a time-discrete sequence of samples is generated, from this sequence a first digitally filtered sequence of samples, the permitted frequencies of which lie in a first frequency group , and at least one further, second digitally filtered sequence of samples, the permitted frequencies of which lie in a second frequency group and are greater than the frequencies of the first frequency group, are determined for the samples of the first filtered sequence in each case the phase difference compared to a previous sample as a measure of the signal frequency falling in the first frequency group and a measure of the signal amplitude present at this signal frequency, the phase difference in each case compared to a previous sample value as M for the sample values of the second filtered sequence ate for the signal frequency falling in the second frequency group and a measure for the signal amplitude present at this signal frequency, the measures determined from the samples of the first filtered sequence and the samples from the second ) co M h- »

Cπ o Lπ o <_π O Lπ

the frequencies of a frequency group are processed together. This distinguishes the process from the prior art, must be sequences wherein each of normally eight Fre ^¬ processed for itself. The inventive method thus allows for more efficient Signalverarbei ^¬ processing than before.

The sampled values of the first filtered sequence are advantageously fed to a first processing module, which generates a complex sampled value from the sampled values by assigning the respective sampled value unchanged to the real part and phase-shifted to the imaginary part of the complex sampled value, and which measures the phase difference as a measure of the signal frequency complex sample value and as a measure of the signal amplitude, the absolute value of the complex sample value is determined. Accordingly, according to this development of the invention, the samples of the second filtered sequence are fed to a second processing module, which generates a complex sample from the samples in each case by assigning the respective sample unchanged to the real part and out of phase with the imaginary part of the complex sample, and that as The phase difference of the complex sample value is determined as a measure of the signal frequency and the absolute amount of the complex sample value is determined as a measure of the signal amplitude.

While the conventional methods require a separate receiver for each frequency, the method developed in the manner explained above manages with only two processing modules. The technical outlay for carrying out the method is thus considerably reduced. The determination of complex sample values, ie an analytical sequence of sample values, has the advantage that the complete phase information is contained both in the real part and in the imaginary part of the corresponding complex sample value. This makes it possible, as it were, to hide, for example, the frequency which has a negative sign and which is in the real one LO) MMP "

Cπ o Cπ o Cπ o cπ

rjö φ Cn M> H? ⁾ rt αq d φ p- N uq H Ό ü <Φ tr? tr cn W rt 3 P. 3 N ^ ^ p. li rt P- Φ fc P- ⁾ d φ d P li cn s: ι-i P- li P- Φ li P- d Φ rt P- dd Φ P- Φ li o er d: cn Φ P cn n Φ P Hi li Φ cn rt d Φ o φ li N li M Φ PP ii rt P- φ d rt n rt tr Φ tr Hi PP rt * • P- Ό Hi? cn Φ P cn rt cn Φ Φ uq Hi rt o ^j ι- (J φ rt <Ό O. tr N Hi Φ rt TJ TJ Φ rnd φ P- rt O tr d P- cn

Φ P £ Φ PJ P. φ PJ Φ d: P d Φ Φ PJ cn Φ tr uq P <P d z α s d Φ φ Φ rt

P Φ P P- cn ⁾ P rt PPPP Go P- O Φ D. Φ P> P φ Φ φ uq PPP li Z

Z <P ⁾ P- Hi φ cn cn ^ Φ tu α tr P Φ PP α P- cn cn φ uq N 1 Φ φ φ P. rt P> cn Φ d α α ti Tl P PJ li Φ Φ ^ P cn rt P tr

P <li rt P li φ ^j J Φ tr tr O α Λ- P- O li φ φ tr P- s li

(- ^■ H P- Φ tr cn ω P Φ d P- Φ α P. rt P: PP ö Po PJ Φ 3 O. α Φ φ rt

PJ P l-f Er PJ

Φ P | Φ Φ Φ P> d: Φ nd cn Λ P ⁾ φ ^ uq s P ⁾ O uq P cn Φ 2 cn P cn P> P P.

P tr P ι- (O li <tr P- φ d P- φ φ φ tP 3 P- 3 1 O Hi φ "<Φ a φ α <! Cn li cn t Φ o PPJ Φ rt rt P Hi li TJ rt P ⁾ N> CΛ 3 rt cn cn cn

P- P- PJ O P- φ P lf PJ: NP Φ Φ s P- P ⁾ rt Z d P- d er tr φ Φ uq rt tP H P- PJP rt cn α uq cn NP rt φ Φ P- Φ O Φ cn uq rt rt ⁾ rt <cn

P- d rt φ uq ds: Φ cn Φ h uq Φ -> r- ¹ rt P φ Hi P P- uq P s: PJ cn Φ o cn p- rt P α Φ d φ o Φ Fr P- P- M d tα ti P Φ • φ P- rt φ P> Φ cn φ P rt PJ vQ P- p- rt li tr rt uq Ό O d P- Φ fi d PP ^J N Φ cn ti rt P ^ N Φ d cn φ rt ⁾ N Ό O Ό ^ PP rt ff rt d cn rt cn rt Z d P Φ rt cn

Hi 3 tr P- Φ cn P- z φ t Ό ι-i P. Φ cn Φ P> Φ Φ P Φ P- cn Hi

P- o φ n P- P rt cn φ Ό φ Φ Φ P- P TJ er t- (<! <P ^J Φ cn H, rt rP o O:

P "α ff Hi tr Φ Φ α ri P ⁾ rt P- φ PJ Λ 3 uq ι- (rt φ Φ rt P- rt Φ d tr <P rt d cn rt • P P- J Ö3 • rt P to φ d P ^ ^d:) li ι- (d P li α Φ Φ Φ O 3

Φ rt Φ • ^ u rt Φ rt P φ Φ rt OP Cn PJ z ⁾ P P- φ P cn PP ι-f P- ι-i Φ rt d P- Φ Φ WP tr PJ rt P cn PJ uq rt cn φ Pi uq Φ cn ON cn tr uq

<Φ P rt cn P- JH tr N uq Z cn P tr d PJ tr ^• * J φ

N φ υq dn? P ⁾ tr P ^ σ P> Φ Hi Hi φ P- Φ Φ P. Φ P. rt> d P- PP d li φ cn t rt -> Φ tr ι- (rt Φ ¹ P d: P- PO ι - (P φ P- Φ P- O Hi φ uq IM uq PJ Hl uq P P- Hl rt Φ cn φ Φ P- rt P tr rt P rt! -I, uq 3 rt P. φ • φ S cπ φ P P- φ P- O P- ι- (P- O, PP rt Φ α. Rt r Φ Φ cn d Ό ι- (φ cn WP

Hi tr P- cn cn P r- ^> P- d Φ Φ P d α Φ O: P P- PW φ P- P- cn rt • d- 'Φ rt rt O rt Φ Φ Φ P rt Φ li P- li P P- O s uq Pi p, φ rt _^ PJ P- 5 1 tr P- Φ PJ tr Φ Φ tr P- tr P - P n Φ P c? P tr P- cn Hi 3 X rt rt P ⁾ cn CΛ

P rt ι-i P ^J φ P- ι- (Φ P Φ N • D tr Φ φ rt rt 3 P- P- φ • α rt d rt P- rt φ rt PPP φ cn uq M P- Tl r- ¹ P- li P rt Φ OP rt P Φ Φ Hl • uq rt Φ d Φ cn 1-1 P Φ P r P>: P 3 PJ P- P- Φ O α P. rt l ti rt P z P cn O z D d LQ D- φ cn P- s D. cn P Φ P- dd Φ tr

P- P. ⁾

• n uq σ d Z φ ι- (Ό d Φ Λ cn D. rt d rt Hi PP & Φ N cn P- P- li P- o Hi tα o li er d: O φ 3 d P- Φ rt PZ PJ Φ Φ uq rt rt P z O φ Φ cn

P r- ¹ α ≤ O tr rt n Φ P Φ uq p. Φ O Φ O P- α uq Φ 3 LJ. ⁾ cn p-

** PJ o r- ¹ Φ PJ N φ PP "tr

"* fi tr P Φ cn z cn O cn 3 ^ P φ cn P tr P cn Φ Φ φ cn N d Φ rt rt φ φ cn z P- Φ rt O P- P- Φ PJ α TJ rt PPP Φ PP Φ Φ PO φ Φ P,

Tr z disregards cn O

PJ <J P- r ι-i ≤ rt Hi P r Φ P> Φ rt cn P- t tr

Φ N cn α cn Z P- ⁾ r- ¹ P- tr

Φ PJ PJ φ tr d d: rt Φ P d S P O \ -> rt Φ f-i P p Φ Ό

P Φ H d cn P rt P> P H d Hi Φ p § φ PJ H- cn P rt

Φ P P- O φ

Hi Hi x

P Φ • n P Φ 1- ^ α P d: P

3 α φ P- - ¹ Φ <Φ N p cn uq Φ d: o PJ cn d • P Φ rt d α P Φ p 1 cn s: P Φ O cn p tr P <P Φ ti P- dz P- dph ^ P cn CD φ rt P- li d φ Φ rt p

N P 3:; v Φ α P n φ P α t-i - rt O P- p P P. P, p- d p- u? Ö

CΛ z cn Φ cn tr rt d H P- Φ tr O P- Φ Φ 0 cn P- rt Φ p- φ P- P n rt <! Φ φ Φ P- rjπ 3 uq P P- tr φ Λ α cn J

O: P- OOP PJ Φ tr 0> P- Φ PJ φ cn rt u li O φ P- P- pj: P d P> Ό α tr tr ¹ rt d uq p 3 d Φ

NPP li P rt ^ 1 cn pj: P- φ cn tr tr d Φ rt tα cn t-pj dö Hl Φ Hl p ^j cn Φ P- tr d Φ J ti cn α cn PP P- PJ li P- g φ Φ p

P- PP Φ P Φ Φ Φ P ¹ Φ P- Φ cn N

Ü Φ ^" PJ li cn n ^ ii - ¹ Φ Ό cn

1 P- W

1 P- o

3 p. uq li Hi 1 uq M P- α 1 - ^r φ Φ tr O Φ | ι- (P- rt p -i P P-

O. P P- d P- uq Φ cn NP ¹ 1 1 P- p-

Φ PJ 1 1 Φ φ

P φ PP "p. P- 1 uq 3 1 cn uq 1 3 P li 1 φ 1 1

gnale is impermeable. This allows low-frequency interference ^¬ signals, as for example, is on the fork of a remote ^¬ speaker reflected acoustic sound signal, easily and reliably be eliminated.

It is expedient to modulate the first filtered sequence before it is processed in the first processing module by multiplying every second sample of this sequence by -1. This modulation results in a band reversal in the frequency range of the sequence, which has proven to be favorable in test applications of the method with regard to the further processing of the sample values in the first processing module.

It is advantageous to match the sampling rates of the two filtered sequences to one another after their processing in the respective processing module and to supply the two filtered sequences to an identification unit which compares the dimensions determined by the processing modules with the target values. By adjusting the sampling rates, the DTMF signal can be evaluated in the identification unit for both frequency groups with the same sampling rate, that is, on a common time base.

According to a further aspect of the invention, a receiving device is provided for performing the method just explained.

Further advantageous embodiments of the invention are the subject of the subclaims and the following description.

The invention is explained in more detail below with reference to the figures. In it show:

1 shows a receiving device for a DTMF signal, FIG. 2 shows an example of a processing module provided in the receiving device according to FIG. 1,

FIG. 3 shows another example of the processing module,

FIG. 4 shows a special embodiment of the receiving device according to FIG. 1,

5a and 5b show two bridge wave digital filters provided in the receiving device according to FIG. 4,

6a to 6c the frequencies of the upper and lower frequency groups shifted due to a reduction in the sampling rate,

Figure 7a to 7c

Simulation results of the process.

The basic mode of operation of a receiving device 10 operating according to the method according to the invention is explained below with reference to FIG. 1, which is designed, for example, as part of a digital private branch exchange and for receiving DTMF signals.

The receiving device 10 has two processing branches 12, 14, each of which contains a processing module 16, for which two exemplary embodiments are shown in FIGS. 2 and 3. The processing branch 12 processes the DTMF signal in a frequency range which is defined by a lower frequency group provided in the multi-frequency selection method and which contains four predetermined frequencies. Correspondingly, the second processing branch 14 processes the DTMF signal in a frequency range which is assigned to an upper frequency group. The upper frequency group also contains four frequencies, which, however, are larger than the highest frequency of the lower frequency group. The lower and the upper frequency group are m Figure 6a shows a concrete example with reference to ge ^¬ and designated there with LG and HG.

The processing module 16 provided in the first processing branch 12 is preceded by a low-pass filter 18, which is permeable to the frequencies of the lower frequency group LG and impermeable to the frequencies of the upper frequency group HG. Accordingly, the processing module 16 provided in the second processing branch 14 is preceded by a high-pass filter 20 which is impermeable to the frequencies of the lower frequency group LG and permeable to the frequencies of the upper frequency group HG.

The processing module 16 of the first processing branch 12 determines, later to be explained, a measure of the frequency and a measure of the amplitude of the portion of the DTMF signal processed by the processing branch 12. These two measures are referred to below as frequency measure FM1 and as amplitude measure AMI. Accordingly, the processing module 16 of the second processing branch determines a frequency measure FM2 and an amplitude measure AM2 of the portion of the DTMF signal processed in the second processing branch 14.

The dimensions FM1, AMI and FM2, AM2 are fed to an identification unit 22, which evaluates these dimensions in order to thereby transmit a signaling information transmitted with the DTMF signal, i.e. a DTMF sign to identify. For this purpose, the identification unit 22 compares the dimensions FM1, AMI with target values specified for the lower frequency group LG, while it compares the dimensions FM2, AM2 with target values specified for the upper frequency group HG. This comparison is carried out in the receiving device 10 presented here in accordance with the ITU-T standard Q.24 specified for the multi-frequency dialing method.

The setpoints used for evaluation in the identification unit 22 mean that for the m two frequency Group-assigned frequencies defined tolerance ranges within which the frequency measures FM1 and FM2 must lie so that the corresponding frequencies are recognized as being present in the received DTMF signal. In addition, the setpoints stipulate minimum levels for the amplitude measures relating to the frequencies under consideration.

The identification unit 22 uses these setpoints to check whether the dimensions FM1, AMI or FM2, AM2 are within the predetermined tolerance ranges both for one of the four frequencies of the lower frequency group LG and for one of the frequencies of the upper frequency group HG. If this is the case, the received DTMF signal is identified as the transmission signal of that DTMF symbol which is determined by the combination of these two frequencies contained in the different frequency groups LG, HG.

FIG. 2 shows a first example of the processing module 16 used in the two processing branches 12, 14 of the receiving device 10. The processing module 16 receives a sequence of samples x _n , which has been generated in a manner known per se by sampling the DTMF signal at a predetermined sampling rate and then digitizing. The sampled values x _n are fed to a digital filter 24 of the processing module 16. From the respective sample value x _n, this generates a complex sample value x ' _n corresponding to this. n is a running index which indicates that x _n or x ' _{n is} the nth sample within the time-discrete sequence of samples. The complex sample x ' _{n is} the nth sample within the discrete-time sequence of samples. The complex sample x ' _n is by the relationship

(1) x ' _n = Ae ^ωnJ = Ae ^3P "= Re (x' _n ) + j ■ Im (x ' _n )

given, with A the constant amplitude and ω the constant frequency of the sinusoidal DTMF signal, with T the Sampling period corresponding sampling period, with p _n the phase of the complex sample x ' _{n /} mi Re the real part and Im the imaginary part of the complex sample x'"and j the imaginary unit. In the digital filter 24, the received sample value x _n is supplied on the one hand to a real branch 26 and on the other hand to an imaginary branch 28. Over the real part of branch 26, the sample value x _n substantially unchanged as a real part x of the complex sample _'n output while the sample x _n in the Imaginärteilz- weig 16 by a phase shift unit 30 by -90 ° with respect to the data transmitted via the real part branch 26 sample x _{n is} phase shifted. The digital filter 24 thus outputs a value via the imaginary part branch 28, which represents the imaginary part of the complex sample value x ' _n .

A Hilbert transformer can be used as the digital filter 24, which carries out the above-described phase shift and thus the generation of the complex sampling values x ' _n . The digital filter 24 outputs the complex sample value x ' _n, divided into the real part and the imaginary part, to a phase determination unit 32 and to an amplitude determination unit 34. The phase determination unit 32 contains an arithmetic unit 34, which determines the phase p _n from the complex sample x ′ ″ for each sample n and outputs it. The CORDIC algorithm known from the prior art can be used to determine the phase.

Delay elements 36 and an adder 38 are connected on the output side to the arithmetic unit 35. The phase p _n determined by the arithmetic unit 35 is supplied to the adder 38 directly on the one hand and on the other hand via the delay element 36. The delay element 36 delays the phase p _n by m times the sampling period T. m is a positive integer and is equal to 1 in the example according to FIG. 2. The signal p _n _ _m output by the delay element 36 is the phase of the ( nm) -th complex sample x ' _n - _m . The adder 38 calculates the phase difference p _n _ _m between the nth complex Sample x ' _n and the (nm) th complex sample x' _n - _m . The result of the subtraction carried out by the adder 38 is output as frequency measure FM1 or FM2.

The amplitude determining unit 34 includes two multiplied as ^¬ rer 40, 42 and an adder 44. The two inputs of the multiplier 40, the real part Re (x _'n) of the complex sample x' _n is in each case supplied while the two inputs of the multiplier 42.jeweils the Imaginary part Im (x ' _n ) of the complex sample x' _{n is} supplied. The multipliers 40, 42 each form the square of the values supplied to them and output the result to the adder 44. The addition carried out by this provides the square of the absolute value of the complex sample value x ' _n and thus, as the amplitude measure AMI or AM2, the square of the constant amplitude A of the sinusoidal DTMF signal.

FIG. 3 shows a further example of the processing module 16. This differs from the processing module according to FIG. 2 only in the design of the digital filter used, which is designated 46 in FIG. 3, and in the two additional averaging units 48 and 50. The others Components are identical to those of the processing module according to FIG. 2, so that they are not described here.

As in the first example of the processing module 16, in the example shown in FIG. 3, the received sample value x _{n is} supplied to a real branch 52 and to an imaginary branch 54, the structure of which is described in detail below. After signal processing, the digital filter 46 are over the real part of branch 52 to the real part of the complex sample _'n aufzufassenden value Re (x' _n) x, while it to the imaginary part of the complex sample value x on the Imaginärteilzweig 54 _'n aufzufassenden value Im (x ' _n ). The imaginary part Im (x ' _n ) ω Cd r ιv) P-> P- ¹ cπ o cπ o Cπ o Cπ cn α Λ ^ <P o P- PJ> P. <J CTi 3 Cπ p. N d Lπ P- P d C tr CX uq i rt ö V Φ n Φ d P o φ d PPP \ -> P. φ Φ d OP co Φ o P 00 φ JP d PJ o P er d Φ Φ φ P tr P Φ uq PPP φ Φ Φ cn α 3 for rt P uq α 3 Pi α tr rt tr rt PP pf J rt

P ^ PP d rt uq PN tr cn Φ Φ P rt rt d rt) uq P- ¹ cn

P N P P Φ cn 13 p to φ α O O rT f P Pi Φ cn> Φ cπ Pi Φ P • cn cn ö V rt Ό

Φ N 3 tr P Φ tr φ 3 fi Φ uq P. Z 3 P rt 3 rt α P 00 JP α rt uq P Φ Φ Pi t *> d J üü PPP ⁾ O n ⁾ φ P Φ Φ P tr P o Φ Ό α tr P ö O H- * uq JPP

Φ <! εo Φ P cn tr uq P Pi 3 φ Φ rt tr Hl 3 PN ι_ι Φ Φ cn Φ Φ PP r- ¹ O

P o uq P Φ P "φ P Φ P cn rt N rt rt PJ P Φ z Φ PP ro Pf pf φ rt rt tr li 13 φ cn Φ P rt PP α φ> p. Φ Φ P rt Pi φ ZP p . ⁾ Φ α rt

Φ * N φ P Φ d φ PJ cπ α Q. Φ P <uq Φ P Φ Φ £ cn PH O r- ¹ P φ φ Pi Φ rt ι-i P. P fi 00 P> α 3 O rt φ P rt p, uq PPPO 3 α cπ Hl cn

Φ d ^ PN li 3 φ P rt Φ α P t Φ pf rt Φ P- ¹ Φ PZP) Φ CTi PN PJ

Pi ι- (uq φ cn Cd PP tr Φ P> i α Φ> PNP cn 3 cπ cπ cn P d P uq "• ¹ Z Ϊ tr cn n i-il rt d rt 4-» rt φ Φ P α Φ P Pi Φ CΛ ooo CD _-> P Φ φ PH rt Φ o Φ rt tr ^ Φ uq Φ rt ti P - ^> α Pi rt φ α rt d uq P

P> N uq P P Φ P P

Φ> P rt P PJ 3 rt PP Pi Φ tr Φ cn α dd N £ P ⁾ p. φ J <P ⁾ p (uq ä V

Φ Φ ι- (d P φ CSX O Φ rt i Φ J cn P d Φ P φ P pf Φ cn cn Ξ

£ P tr NP tu tr P rt co 3 fi

P σs Pi CΛ Pi rt Pi cn uq α uq P uq PP rt ρ <cπ φ P

PPN z Φ Φ uq rt Cπ rt cn K Φ n P ⁾ Φ uq PJ Φ • «• φ φ Φ N j-> MX φ rt φ £ φ> PP cn P * α Φ Pi tr PP φ d PN Hl CΛ P 3 PO Φ CTi φ α rt P d ω rt Φ d α X d PJ p, a P rt tr 3 uq ddd ZP uq P α P Φ

Φ ⁾ rt cn n Φ PP Φ - P cn uq J σi Φ P g PP, t § PJ uq <N Φ tr Φ p. * P ^ Φ uq PPP uq H α Φ cn N> tr o P a P. p, y tr PJ Φ Z Pf P Φ cn gdd 3 P PJ φ tr cn tr tr Φ P rt cn J Φ PJ rt PJ P ( P PI tr rt tr cn 3

P α TJ P p- ¹ z Φ Φ T) P. ⁾ P d N o cn PJ cn cn P Φ P Φ uq NP 3 rt PJ rt uq PP ¹ tu uq cn PPP tr P. PJ P yd Ό Φ cn P . α PO uq d Φ PJ α cn φ P Φ cn Φ rt P PJ Φ cn tr cn uq Φ P | rt Φ rst <Φ Φ d uq PN tr cn cn Φ

Φ rt P cn α tr cn cn P α φ Pi uq Z cn Z Φ Pf er P P φ cπ d rt rt

P TJ d cn P φ α Φ Φ cn Φ α PJ P PJ Φ Φ φ P. Φ rt Φ α α Pf * »uq tα Z 3

P tr Ό uq cn PMPPO Pi φ cn d rt tr Pi PNPJPP d ^M t> d PJ Φ PJ P φ tr J Φ PP Φ rt Φ O rt cn tr PP r P- Hi O cn P φ Φ P σ P Hl d P- ¹ f 3

Φ cn P Φ P ⁾ <d ti Φ Φ N Pi OP P- P ⁾ uq N rt uq P uq Φ d cn tr rt

P φ 3 -> Φ PP cυ 3 pf> Z rt P cn XP n Φ ≤ PJ tr P cn cn tr Φ Φ 1 rt P PJ Φ P tr P.>t> PP ¹ 3 CΛ α Φ. Φ tr Pf 3 Φ P Φ Φ uq rt PP Φ P {cn CD φ tP P P> P rt Φ PX d α P tr Z pf Φ d P> uq PP P- ¹ Φ φ rt P rt o

> fc> l-i P Z l-i Φ P. Z rt rt tr P Hi s PJ P PJ Φ

1-1 tr P rt Φ α P tr Φ P rt X o co 3 fe O Φ r PP Φ P- Φ rt rt φ PJ P P- ¹ P p. α uq α Φ 3 φ rt to P Φ 1-1 -

P s tr fi Φ Φ P d Φ Pi P (n φ tr α • cn φ cn α X tr PP φ 3 PJ Ό tr rt P "rt P. P φ rt PP ^J PJ Φ tr P uq P PJ rt PJ rt _^ n P • tr

N φ rt P & Φ uq er rt Ϊ Pi Φ P- Φ ö σx Φ tr PJ σi d tr er φ cn ⁾ z "d P d rt π fi cn Φ rt cn φ p. D PJ o P Pf uq Φ P _- »cn Φ rt P Hl cn

• d t £ p P cn -p φ rt P α 3 rt P (cn er φ Φ φ P α 1 • P PJ P o Φ

P g 3 l_J uq Ό Φ P ⁾ O σx Φ nd rt P. Pi P ¹ • Φ 2 φ cn NPP α uq IV) P φ cn ii PPP Cn P o P cn> d tr P Φ pι d rt PPP PJ Pi ö P! P rt P 3 α rt 3 Φ rt rt rt p. cn P. cn cn CTi PZ uq o Φ tr φ φ P oo Z uq J Φ li Φ <! nd Φ φ P Φ N g α P. Φ rt o φ P Φ f P tr φ Φ rt Pi n P φ Φ o α tr α rt P Φ rt Pi d PP P- N Ό CT) • rt Pi N φ PP o cn tr P li tr Φ Φ • rt Pf P uq uq rt φ P φ P .C- 3 Φ>Φ>> P rt Pf tr n

Z tr p- ¹ PP rt O Φ Φ PJ P Φ φ φ Pi ö N α α P α α rt tr

Φ φ φ PJ tr cn P. Φ CJ Φ P tr Hl cn Φ 3 tr P o

P P l-i Φ Φ l-i> Cn cn P d rt P (P> <φ d Φ α O α X φ J

P rt O Φ P! 3 3 P P P Cd Pi d tr

P rt α P P- ¹ P 3 Φ P

X tr O P P- α h Φ p φ Φ cπ N cn Φ

Φ Φ PJ Ti P cn p | cn P Φ cπ P- Φ tr i φ PH t-i Φ Pi Pf d O uq p rt P. • n «rt O" uq rt Pi 00 * 1 P J d

Φ o rt Z Φ »P P P Φ H P a P 3 Φ rt φ Φ (D 3 uq Φ Φ

PJ r ^ c P. P PJ Pf Φ PI P d Φ tr p

P 3 σ, ä 1 P PJ er Φ O Z Φ uq i o P φ uq P P cn P. 1 P cυ cn ^ P Ό φ P P cn N P- P Pi d Φ P <Φ φ P P cn Cπ 1 P

P O Pi P φ o P d φ p, Φ Pi P | Φ P P 1 CTt co cπ φ

P 1 P φ P 1 cn p. p. Pi φ N 00 φ

1 1 H Cd 1 Pi d 1

Cd Cd tv> rv) P- ¹ P- ¹ cπ o Cπ o Cπ O cπ

> cn P t P Tl rt ü tJÜ o PJ rt cn ?? Φ z ^ rt σ P- z Φ φ P Φ α rt P- **! uq PJ d rt d φ d P- P- J Pf er P Φ O Φ Φ P- p- PJ PJ P- P- P- P- J P- J d P P- φ d 3 cn Pi Φ PP uq O cn d: Φ P tr Pf PPP> x cn cn φ Pl PP o PPP uq uq cn P TJ tr P> P ≤ uq d POP Φ s Φ P. tu rt α tr Φ tr P ⁾ uq P- d φ uq - ¹ rt N φ Pi a Ϊ Φ z <P- Pi φ N Pf 3 P- φ Pf N d tr Pi er ^ cn P-

Φ Φ uq P- ¹ α Pi Φ PP: Φ Φ "PJ P- P p- α d Ph P» Pi Φ Pi Φ rt

PPP! Φ P »Φ d: P tr P tP -j P- P- o <rt rt d rt Φ uq o Φ 4-> P Φ P- d α d Φ p. cn os: * J 1-1 P ⁾ rt IV) φ P- Φ Φ α Pf fr P φ PJ • 3 PP α

Φ Φ TJ P uq? Φ P! Φ f P rt φ Pi P 00 Φ O IV ⁾ rt Hl d N s: d P ^J Φ

P pf TJ P- tr φ P- φ 1— 'Φ P er cn d α Φ Pi <Pi tr ro ⁾ Φ d: cn ö TJ Φ Φ Φ Φ P

3 φ P- Φ NPP ^j α Φ P- P P- Pi cn P- Φ? cn er tr uq P- i P- PP P- φ l_l. P- uq P- φ z Φ d P- n α uq φ OPP tu M 3 p ⁾ rt Φ P φ φ P- uq α N rt P

Φ P t- ¹ P- α P- ^ φ P Φ cn rt tr P- P- uq 3 N 3 Φ d pf P rt P- cn P rt Φ 3 3 α α Ω rt Φ uq P- P- ¹ α P P- d -o rt P Φ φ P- Ό tr cn PJ Φ φ Φ NP ⁾ Cπ P-

Φ Φ PJ P rt uq P- NOP P- 4-. PJ tr f PP PJ Ph Pf uq rt Φ uq P- Φ • £ o rt cn Pf tr r- ¹ Φ d Φ uq uq tr uq P Φ N rt α er PJ Φ φ Φ p, S rt> J P- rt

Φ Φ Hl ^ P Pf P P- P cn Φ Hl P- φ Φ uq P Pi Z P- ¹ ^

5 d P- P lτ] d:

NP! N P- Pi P. rt d P. N P. P P- rt d P Φ uq Φ Φ <p: cn Φ φ s er dzo Φ t P- ¹ P- PJ • 0 Φ Z Φ rt ¹ uq P- Φ rt cn f O d 1 α Hi P-> Φ P φ -o uq rt Λ O uq TJ P Φ 3 tr rt 00 rt P f PJ Φ rt P rt CΛ PJ 0 (- • cn ** Pi uq

P- ro Φ φ dn uq P- Hi φ P- pi: Φ tv) φ α cn P- fr Φ Φ P- & tr Φ J cn rt P Pf Φ tr φ rt P- uq P. p (Φ -j rt P rt rt P Pi uq PP Φ hrj 3 Φ

Φ N Φ P TJ NJN er φ rt P ⁾ "d 3 o Φ p (J d rt P P. d N ^ Φ p-

P d co NJ φ rt d rt P- ¹ Pi • 00 d O rt P- cn Z Pf Φ PJ PJ P P- IV) tr P uq <! 4. uq cn P- Hl φ uq PJ Λ cn LJ. N Φ n rt p- P r- "cn uq d φ PI tr φ O Pi cn uq P- Pf φ cn d σ uq uq φ d tr pf Pf 3 cn Ph P 'er Φ φ er Hi P ¹ J d Φ rt P-- o rt P Φ cn Φ φ α uq S rt J φ π3 P P- Ph er P- ¹ N Pf P- rt d: uq d TJ cn Φ rt 00 Pf 1-1 Φ rt Pf O uq Φ d rt tr Pi PJ PT O PJ Φ uq Φ ro Pi P rt f P- er O d: 5 s Φ rt

PJ tr φ cn TJ P Φ 4-> J 3 Φ Z Φ Pf Hi P Φ P P- n y P d cn PS 1 uq P Ö P tr 3 rt rt <φ P o H J Φ Φ ΓJÖ Φ fr φ Φ>

PN tr CΛ 0 P- P ¹ N 3 ■ α d PJ for 4-, ^ω

Z - o uq • P- 00 P rt P Pi P rt P z Pf P- Φ φ P- φ P- P. PJ cn rt φ P φ t → φ 4-. cn Φ <d: PJ « _* φ rt uq O pf cn f Ό P- rt φ P (Ph cn PJ PJ

P α er CD M Hi O P Φ O ^ cn P- er P- ^ P 3 Φ tα P- P p | uq d: cn P rt Φ Φ cn TJ d: tr PI * - Pi rt φ rt rt PJ Φ P- N φ φ Φ Φ tr J> rt cn

Φ P! er P d PJ er <3 φ Φ pf P- PJ Φ r- ¹ PPPP (p- rt w cn Pi d O cn Φ P cn cn Φ P- Φ P- PP PJ tr Φ cn Φ rt α for P- P 3 rt d TJ P tr

P. PJ α O cn Pi φ Pi P sd rt Φ rt f Z tr Φ rt P PJ CΛ Φ TJ φ P P- uq P- ¹

P- cn pf Φ P tP ⁾ φ Φ Φ P- Φ Z er P- J Pi J er P- p, Hl P- "uq P- Φ P-

Φ rt PJ tu uq cn P- rt pf Φ P-- P P α P- Φ rt α Pi Φ cn

• α P- ¹ uq J P- "cn rt P Φ pf cd Q rt | 3 tr PN <Φ PPP ⁾ P- α rt Pf rt Φ PJ: PP rt Hl d &>

£ Hi • P- ¹ 3 Φ Φ φ uq Φ P rt cn Φ Φ Pi PI d PJ & uq Φ O P- uq φ

^ Pi PJ 00 rt P- Pi P pf Pf er Φ rt NP -j P ⁾ Hi rt cn PP φ φ p

O rt rt P. σ P- ¹ rt

P- cn α d 3 P- P S d o rt rt PJ Φ

P-r PJ Φ P- PJ cn d ^ j P- n P- 3 3 α

& p. Φ for P cn Φ _* φ cn P- M g P- Φ uq cn P φ φ OOP -t. uq JP rt Φ Z p. rt PJ: P- Φ p, rt P 3 J rt φ rt er P cn Z uq P- Φ α PJ N H- rt PJ OP P- HP cn J pf Pi TJ Φ tP rt N

Pi Z d Ph o cn P- ct Φ cn P- P cn tr a cn Φ rt O P PJ p- Hl P Φ

<PJ P- P o: J tr NP PJ tr ¹ Pf rt z

Ό P. P> rt cn φ rt Pi rt P Pf d - Z ¹ Q Φ α P- tr er rt O PJ Jμ r- 'o M Pi d pf rt 00 d Ph Φ Φ tr p P- p rt P-

P Φ α Φ O Φ P. Ph Pi PJ O cn PJ o N P- P- P- rt uq P P- ¹ rt

Pi α α tr P- Φ P- N rt>? PJ uq rt α P- N Φ rt <d cn _^ . Φ

4-> Φ Φ P- P- uq 3 d -j Φ φ P Φ Φ Φ Φ Φ P- EP d O p Φ er "1 d Φ 3 φ er α φ rt uq t P cn P

* "Pi 3 p. P Φ r- ¹ pp uq P- P- jjj φ

3 1 φ P ¹ α Φ Φ

Pi a H cn J _> fc _> <PJ pf d P, P- uq p uq $ <p s- tu O Φ Pi o Φ t ω PJ φ d Φ P dn 00 P ¹ r Pf d IV) O rt

N Φ Pf Pi dp, h Pi Pf PP Ph er P- uq P tr o P- P (Pf rt er 1: Φ P- P 00 α 3 d: P. 1 n uq P.? S- n ⁾ Φ

PJ rt t P P f Φ 4- »P P- Λ" φ tr cn

PJ 1 1 1 3 P- tu P- tr P- ¹ d cn 1 d 3 φ P 1 rt rt 1 P- φ N cn 1 cn P-

Ph 1 1 1 rt rt 1 φ

Cd d to t P- ¹ P- ¹ cπ o Cπ o Lπ o Cπ

P-> PH H α -S Hl <rt ö N rt rt rt - ¹ σ PJ NP uq n er S α er to N α er ≤ P ro o P α Φ Φ d: φ Φ P- P- Φ P- J PJ P- d £ PP tr Φ PJ P- fr

P P- tr P P Φ Φ o cn P φ Ph Φ rt PJ Φ H- tP z J P- Φ φ P-

Φ P- ¹ rt C- φ P φ P 3 fr

Hl P. P 3 cn f P ⁾ P rt P rt uq P- Φ P- ¹ P cn 1 α P> n P- P Φ r-> 3 td: φ rt PJ Φ rt Pi Z <Φ Φ Z rt < P- ¹ rt P cn J CΛ φ Φ cn r rt <Φ rt N tr 3 P- tP Φ er P- OP Ph Φ O Φ <tα p- P- P- rt rt Φ Φ Φ O j P

P Ph d Φ PP d: PNP Λ- P Φ ι-3 α φ uq PPPP P- P o α P rt N P- fe P Tl P- CD tu tr rt dt P Φ (- ^■ P φ PJ z P P - Φ Φ z N [2J P. O rt α OJ Pi N £ 3 er 3 cn ⁾ P- P- rt φ Φ P. uq uq P P. φ P- d α Φ P rt P. Φ Φ P- Φ ≤ PP Φ P- P φ φ P- d

Φ P- Φ • φ uq P Φ 3 PJ P. P- P- f <rt P Hi S Φ φ P P- '3 rt cn s N

Pi rt p Φ uq 3 d d N Φ P O PJ α P p- er • n P φ rt H- PJ P-

Φ d ö 3 ^

P- P- P- 3 P- ds cn 3 d cn Φ NP tn Φ Φ cn Φ P- Φ P Φ P o t- ¹ rt Φ

H- P P uq? Φ 3 z P rt Pf P- 1 Φ cn uq tα Pi α Pi er Pf Hl PJ P

P uq φ rt P O P Φ Φ φ P Φ P- Z Φ rt CΛ φ d o 3 P- Φ rt • P- cn rt

<cn cn rt P. rt uq tr P- S P- g Φ P Φ P- P- P n P- uq P- Φ J P ^ rt •

Φ Φ Φ Φ d P. p- Φ P P ^ d rt 3 Pi uq d rt tr P P- P 3 tn rt P

Φ P P- P- ¹ Φ P- ¹ φ TJ er φ o - * Φ rt rt -j τι PP P- σx TJ α rt tr rt Φ JW i PJ PZ rt P 3 Φ α P P- ¹ rt P _* Φ CTi o PJ o uq PJ Φ P ⁾ Φ PP rt P- i tr Φ P- ¹ P- P Φ uq P- cn P 'tr rt 3 tP p P ^ P- PP ⁾ Φ P

Φ er φ pf cn PJ σx φ N Φ cn cn Φ TJ -n N uq P- Φ Ph rt CΛ rt 00 P φ

Pf Φ P- α OP p P- P TJ P ^ p- Φ d Φ d: 'Z rt P- rt Φ 4- »cn P- rt Φ N cn Φ ^ pi: 3 P- P- ¹ P- uq er Φ P- P. rv> Φ P Φ φ rt rt P α tα d rt P ^ P rt P- N rt P φ <Φ Ph Pi tu Φ -j rt CTi P <Hl HP

Φ d tv ⁾⁾ PJ uq Φ Φ o φ Φ rt P- Φ Φ ho P P- ¹ α P3 cn σx Φ Φ d: 3 rt

P P ro cn cn P φ P P P Φ P d: P ^ φ. cn »P IM P P P- cn uq O P. Hl uq d 1 P Φ tr J X er Φ d Φ 3 P TJ

"d cn PJ tr Tl d- '<SO φ Φ PJ P-> Φ p- ^ P f? uq P- Ö Φ P O 00 uq P P- α α P

P- N d P α tr Φ O • P P P P rt rt Φ O P- N PJ 00 φ PJ P P- Φ Φ

P- ¹ Z cn P- Φ Φ PP IN 3 tr PJ rt PP φ Φ CΛ cn _! Ph I P- Φ PO rt Φ uq φ PP rt J 3 er tα d CD Φ cn ⁾ rt cn P- TJ cn, d = er Φ Φ tr

Φ P- φ EP d • Pf o Φ J! - ^■ O p- rt cn PJ Φ n Φ Öd O tr Φ P o H Φ pf uq uq Φ • n er P- P rt ** rt P Φ rt z P cn tr PZP P- φ er P

Φ φ P- P P- o Φ d Φ α P- ⁾ PZ φ PP d: P- rt rt P Φ -j α p- P- ¹ er o. N uq p- P- P- ¹ P- TJ P . 00 rt Φ Pi φ cn Φ er O φ. d 4i Φ

P ^ Φ tr 3 d Φ rt P- o α (- ^■ φ CΛ Φ P. f P- rt rt Φ? P ^ 1 Φ tr P PJ f cn d φ tr Φ P- PP d rt Φ PO: Φ Φ ö uq ro Ph φ α • NP Φ Φ P Pi cn PN P><Pf Φ P φ p cn Φ P Φ P- cn • nd: p- Φ d P cn uq rt P- CJ. d Φ o P »P Φ P- £ P φ NP ⁾ p PB rt rt P Pf Prt cn Φ P. ι Φ Φ cn P tr P tα n Φ cn Z d Φ PJ

[<, tv ⁾ h N Φ P- P α uq Pi Z o cn α O: tr P- ¹ α Φ Φ cn P α cn

00 w> P ¹ p- ZJ uq rt Φ φ P-> P- P- PJ Φ PP ⁾ P- uq d P- rt ro § dd P- ¹ Φ O d • P uq P d Pf φ er rt cn Φ P Φ uq φ Φ φ ι-1

TJ cn d P rt P- P- ¹ <Pi φ uq cn P- Φ tα OOPP ⁾ P uq PJ

<Ph £ P α Φ uq uq o ö N er φ er <P- ¹ O: P P. P rt P- ¹ φ N <rt o PJ Φ P. Pi Φ P σx P- z Φ α o pf cn Ph P- tr ro er uq o Φ

P P i Cd rt <! er Φ φ P d Φ PJ P Φ rt P- φ uq J Qj: φ P P P uq uq rt p. O p. tn P- φ P P P P- O uq cn P- P P d P

Φ cn d PJ Φ P- Pi O φ P Φ rt uq α P α PPP rt rt cn rt 0 φ TJ α φ cn Φ P cn P p- uq P 3 P>: Φ 1 Φ Φ Φ cn J uq PJ OP TJ Φ α

Φ P- uq P- ¹ φ φ tr SP o <P α 3 cn p- r- ¹ Φ tr Φ uq ^ φ tr 3 d

P t? pj: cn S Φ d P- 'O Φ uq P- ¹ Ph P- "P- φ ON

Φ Pi ⁾ d p- φ fr d P

PJ rt Φ p- g uq P l_l. 3 X -n P PJ Φ P- H- P cn PJ ω P-

P P- P TJ rt tT tr rt P- ¹ rt φ φ O: Φ PJ d uq P-- φ φ uq d P Φ φ O - * Φ φ 1 ro P. tr P- Pf P cn P- P TJ uq Φ s Pf P - 'cn rt rt EN t φ cn d: P rt P Φ φ rt TJ P- ¹ φ Λ- cn rt O cn rz Φ uq od rt P- rt rt N Φ fr P, P- PP φ PZ φ ^ P d φ d Φ d PZ P- ¹ PJ: P- 1 tu PJ P TJ PJ Φ cn α P- φ P- uq Φ uq

1 P 1 P uq Φ P- d? N Cn cn P P 3 rt 00 Φ rt P P φ P uq Φ Pf 1 1 PJ rt P- Φ P ^ 00 P Φ P. 1 1

1 1 1 1 1 1 cn

tastratenverminderer 74 reduced in its sampling rate of 4 kHz to 2 kHz sequence of samples which the upper Fre ^¬ quenzgruppe associated processing module to 16th The proces ^¬ beitungsmodul 16 determines a sampling rate of 2 kHz, the dimensions FM2 and AM2. To match the sampling rate of the dimensions AMI and FM1 determined in the first processing branch, FM2 is fed to the sampling rate reducer 78 and AM2 to the sampling rate reducer 80. These each hide the second sample value and then feed the dimensions FM2 and AM2 to the identification unit 22 at a sampling rate of 1 kHz.

FIGS. 5a and 5b show specific embodiments of the two bridge wave digital filters 84 and 88. These embodiments are based on the prior art set out in "Explicit Formulas for Lattice Wave Digital Filters", L. Gazsi, IEEE Trans, on Circuits and Systems, Jan. 1985, pages 68 to 88, so that the mode of operation depends does not need to be discussed here. The bridge wave digital filter 84 has three arithmetic blocks 92, 94 and 96, also referred to as adapters, whose coefficients γl, γ3 and γ5 are defined as follows:

y ₁ = - (2 ^{~ 2} - 2 ^{~ 4} + 2 ^'6 ) γ ₃ = - (2 ^"1 + 2 ^{" 3} ) γ ₅ = - (1 - 2 ^"3 + 2 ^{" 5} )

The bridge wave digital filter 84 thus represents a filter of the seventh degree. In contrast, the bridge wave digital filter 88 is a filter of the fifth degree, the arithmetic blocks or adapters 98, 100 of which are assigned the following coefficients γi and γ ₃ :

γi = - (2 ^_1 - 2 ^"3 - 2 ^{" 5} - 2 ^'1 ) γ ₃ = - (1 -2 ^-2 + 2 ^' - 2 ^'1 )

It should be noted that both the bridge wave digital filter 84 and the bridge wave digital filter 88 each weils can be operated at the lower sampling rate, that is, the bridge wave digital filter 84 to 2 kHz, and the bridges ^¬ wave digital filter 88 at 1 kHz. The sampling periods T shown in FIG. 5 also relate to these sampling rates.

FIG. 6 shows the frequency shifts of the two frequency groups LG and HG, which are caused by the signal processing of the DTMF signal in the receiving device 10 according to FIG. 4. The diagram in FIG. 6a shows the frequency spectrum of the DTMF signal supplied to the first filter unit 82. The length of the arrows is intended to illustrate the signal amplitudes occurring at the corresponding frequencies. As can also be seen in FIG. 4, the sampling rate at this stage of the signal processing is 4 kHz.

FIG. 6b shows the frequency spectrum of the DTMF signal as it appears at the input of the processing module 16 provided in the first processing branch 12 of the receiving device 10. 6b that the audio signal HT has been removed from the frequency spectrum by the bridge wave digital filter 88, which acts as a high-pass filter.

The fact that the frequency spectrum shown in FIG. 6b is only shifted but not "flipped" is due to the fact that the modulation carried out by the multiplier 90 is added to the reduction in the sampling rate from 4 kHz to 1 kHz, which in turn influences the frequency spectrum.

FIG. 6 c shows the frequency spectrum which the DTMF signal shows at the input of the processing module 16 provided in the second processing branch 14. The sampling rate reduction from 4 kHz to 2 kHz leads to a mirror-symmetrical "flipping" of the frequency spectrum of HG. The mirror axis of this flip is determined by the sampling frequency of 2 kHz. FIGS. 7a to 7c show simulation results which have been achieved with the receiving device 10 according to FIG. 4. The following boundary conditions have been taken into account in this simulation:

Pulse / pause ratio: 40 ms / 40 ms DTMF transmission level for HG and for LG each -29 dBm nominal frequencies for the lower frequency group: 697 Hz, 770 Hz, 852 Hz, 941 Hz

Nominal frequencies for the upper frequency group 1209 Hz, 1336 Hz, 1477 Hz, 1633 Hz

Broadband noise of -41 dBm as interference source dial tone at 330 Hz of -7 dBm.

FIGS. 7a and 7b show the time dependency of the frequency of a DTMF signal to be tested with the receiving device 10 according to FIG. This DTMF signal is intended to transmit all sixteen possible DTMF characters in the time range shown. The time diagram according to FIG. 7a is assigned to the lower frequency group, while the time diagram of FIG. 7b is assigned to the upper frequency group. Both the FIG. 7a and FIG. 7b clearly show the nominal frequencies of the respective frequency group.

FIG. 7c shows the simulation result as it is output by the identification unit 22 of the receiving device 10. It can be seen that all sixteen transmitted DTMF characters are recognized by the identification unit 22. Furthermore, the simulation shows that the receiver 10 according to FIG. 4 has to execute about 0.9 million commands per second in order to recognize the DTMF characters. Conventional DTMF receivers require 1.5 to 2.5 million commands per second for this.

Claims

claims

1. Method for processing a signal generated by the multi-frequency selection method, in which the signal is sampled and a time-discrete sequence of samples x _n ) is generated, from this sequence a first digitally filtered sequence of samples, the permitted frequencies of which in a first frequency group (LG) lie, and at least one further, second digitally filtered sequence of samples, the permitted frequencies of which are in a second frequency group (HG) and are greater than the frequencies of the first frequency group (LG), are determined for the samples x _n ) of the first filtered The phase difference from a previous sample value as a measure of the signal frequency (FM1) falling in the first frequency group and a measure of the signal amplitude (AMI) present at this signal frequency are determined for the sample values x _n ) of the second filtered sequence the phase difference from a previous sample a As a measure of the signal frequency (FM2) falling into the second frequency group and a measure of the signal amplitude (AM2) present at this signal frequency, the dimensions determined from the sample values x _n ) of the first filtered sequence and the sample values from the second filtered The subsequently determined dimensions are compared with target values specified for signaling information and the signal is identified as the carrier of the signaling information if the determined dimensions match the target values.

2. The method according to claim 1, characterized in that the sample values x _n ) of the first filtered sequence are fed to a first processing module (16) which generates a complex sample value (x ' _n ) from the sample values x by the respective sample value x _n ) unchanged from Real part and phase-shifted is assigned to the imaginary part of the complex sample (x '"), and that as a measure of the signal frequency (FM1) the phase difference of the complex sample (x' _n ) and as a measure of the signal amplitude (AMI) the absolute amount of the complex sample (x ' _n ), and that the samples x _n ) of the second filtered sequence are fed to a second processing module (16), each of which produces a complex sample from the samples x _n )

(x ' _n ) is generated by assigning the respective sample value unchanged to the real part and phase-shifted to the imaginary part of the complex sample value (x' _n ), and as a measure of the signal frequency (FM2) the phase difference of the complex sample value (x'n) and the absolute amount of the complex sample value (x ' _n ) is determined as a measure of the signal amplitude (AM2).

3. The method according to claim 2, characterized in that a Hilbert transformer is used as the first and the second processing module (16).

4. The method according to any one of the preceding claims, characterized in that the two filtered sequences are determined by a first bi-reciprocal bridge wave digital filter (84) which works as a low-pass filter by being permeable for the frequencies contained in the first frequency group (LG) and for the frequencies contained in the second frequency group (HG) is opaque and works as a high-pass filter by being impermeable to the m contained in the first frequency group (LG) and permeable to the frequencies contained in the second frequency group (HG).

5. The method according to claim 4, characterized in that the sampling rates of the two sequences determined by the first bridge wave digital filter (84) are reduced by masking out sampling values are processed before the sequences are processed in the respective processing module (10).

6. The method according to claim 5, characterized in that the sampling rate of the first filtered sequence is further reduced compared to the sampling rate of the second filtered sequence by masking out sample values before the first filtered sequence is fed to the first processing module (16).

7. The method according to claim 5 or 6, characterized in that the first filtered sequence, before it is processed in the first processing module (16), is fed to a second bi-reciprocal bridge wave digital filter (88) which is impermeable to interference signals (HT) ,

8. The method according to any one of claims 5 to 7, characterized in that the first filtered sequence is modulated before it is processed in the first processing module (16) by multiplying every second sample of this sequence by -1.

9. The method according to any one of the preceding claims, characterized in that the sampling rates of the two filtered sequences are matched to one another after their processing in the respective processing module (16, 18) and the two filtered sequences are fed to an identification unit (22) which compares the dimensions (FM1, AMI, FM2, AM2) determined by the processing modules (16) with the target values.

10. The method according to claim 9, characterized in that the identification unit (22) the determined dimensions with those target compares values that are specified by the ITU-T standard Q.24 specified for the multi-frequency selection process.

11. The method according to any one of the preceding claims, characterized in that it is carried out by a processor of a digital private branch exchange.

12. Receiving device for carrying out the method according to one of claims 1 to 11, with a first filter unit (82) which, from a time-discrete sequence of sample values x _n ) of the signal, a first digitally filtered sequence of sample values, their permitted frequencies in a first Frequency group (LG) lie, and at least one further, second digitally filtered sequence of samples, the permitted frequencies of which lie in a second frequency group (HG) and are higher than the frequencies of the first frequency group, are connected downstream of one of the first filter units (82) first processing module (16) which, for the sample values x _n ) of the first filtered sequence, each has the phase difference with respect to a previous sample value as a measure of the signal frequency (FM1) falling into the first frequency group and a measure of the signal amplitude present at this signal frequency ( AMI) determines a second Ver downstream of the first filter unit (82) Work module (16), which determines the phase difference for a sample x _n ) of the second filtered sequence compared to a previous sample as a measure of the signal frequency (FM2) falling in the second frequency group and a measure of the signal amplitude (AM2) present at this signal frequency , and an identification unit (22) connected downstream of the processing modules (16), which also includes the dimensions (FM1, AMI) determined from the sample values x _n ) of the first filtered sequence and the dimensions (FM2, AM2) determined from the sample values of the second filtered sequence for a signaling information specified target values and compares the signal identified as carriers of this signaling information, when the dimensions (FML, AMI, AM2 FM2) coincide with the target values over ^¬.

13. Receiving device according to claim 12, characterized in that the first processing module (16) generates a complex sample value (x ' _n ) from the sample values x _n ) of the first filtered sequence by changing the respective sample value x _n ) unchanged Assigns the real part and the phase shift to the imaginary part of the complex sample (x ' _n ), and determines the phase difference of the complex sample as a measure of the signal frequency (FM1) and the absolute amount of the complex sample as a measure of the signal amplitude (AMI), and the second processing module (16 ) a complex each from the sample values x _n ) of the second filtered sequence

Sampled value (x '") is generated by assigning the respective sampled value x _n ) unchanged to the real part and out of phase with the imaginary part of the complex sampled value (x' _n ), and as a measure of the signal frequency (FM2) the phase difference of the complex sampled value and as a measure for the signal amplitude (AM2) the absolute value of the complex sample value (x ' _n ) is determined.

14. Receiving device according to claim 13, characterized in that the first and the second processing module (16) are each designed as Hilbert transformers.

15. Receiving device according to one of claims 12 to 14, characterized in that the filter unit (82) contains a first birecursive bridge wave digital filter (84) which works as a low-pass filter by being permeable to the frequencies contained in the first frequency group (LG) and for the frequencies contained in the second frequency group (HG) is opaque, and that works as a high pass filter by containing the frequencies contained in the first frequency group (LG) frequencies are impermeable and permeable to the frequencies contained in the second frequency group (HG).

16. The receiving device according to claim 15, characterized by a first sampling rate reducer (72) connected between the first bridge wave digital filter (84) and the first processing module (16), which reduces the sampling rate of the first filtered sequence _. of sampling values, and a second sampling rate reducer (74) connected between the first bridge wave digital filter (84) and the second processing module (16), which reduces the sampling rate of the second filtered sequence by masking out sampling values.

17. Receiving device according to claim 16, g e k e n n z e i c h n e t by between the first sampling rate reducer (72) and the first processing module

(16) switched third sample rate reducer (76), which further reduces the sample rate of the first filtered sequence by masking out sample values before it is fed to the first processing module (16). ^'

18. Receiving device according to claim 15 or 17, characterized by a multiplier (90) connected upstream of the first processing module (16), which modulates the first filtered sequence before it is processed in the first processing module (16) by taking every second sample of these Sequence modulated with -1.

19. Receiving device according to one of claims 15 to 18, characterized by a between the first bridge wave digital filter (84) and the first processing module (16) connected second bridge wave digital filter (88), which is designed as a high-pass filter impermeable to interference signals (HT).

20. Receiving device according to one of claims 17 to 19, characterized by a between the second processing module (16) and the identification unit (22) connected sampling rate reducer (80, 82), the scanning councils of the second processed in the second processing module (16) Align the sequence to the sampling rate of the first sequence processed in the first processing module (16) before the two sequences are fed to the identification unit (22).

21. Receiving device according to one of claims 12 to 20, characterized in that the identification unit (22) works on the basis of those setpoints which are specified by the ITU-T standard Q.24 defined for the multi-frequency selection method.

22. Receiving device according to one of claims 15 to 21, characterized in that the first bridge wave digital filter (84) is a seventh degree filter.

23. Receiving device according to one of claims 19 to 22, characterized in that the second bridge wave digital filter (88) is a fifth degree filter.