AU666339B2 - Process for simultaneously transmitting signals from N-signal sources - Google Patents

Abstract

A process is disclosed for simultaneously transmitting signals from N-signal sources over a corresponding number of transmission channels. The individual signals are subdivided into blocks and the blocks are converted into spectral coefficients by transformation or filtering, and the latter are then subjected to a data reduction process. The invention is characterized in that the blocks belonging to the individual signals are subdivided into section. The momentary sections of all signals are processed together, the admissible disturbance is determined for each section by using a perception-specific model and the momentarily required total transmission capacity is calculated. The allocation of the maximum available transmission capacity for each individual signal is calculated from the total available transmission capacity and the total momentarily required transmission capacity. Each individual signal is coded and transmitted with the thus determined capacity.

Description

OPI DATE 07/06/93 .AOCJP DATE 05/08/93 PCT NUMBER PCT/DE92/00905 I111111111111 li 11 11 11 iili AU9228069 (51) Internationale Patentklassilikation 5 (11) Internationale Veroffentlichungsnummer: WO 93/09645 H3 7/2Verbffentlichungsdatum: 13. Mai 1993 (13.05.93) (21) Internationales Aktenzeichen: PCT/DE92/00905 Anwalt: M INIGH, Wilhelm; Wilhelm-Mayr-Str. 11, D- 8000 MOnchen 21 (DE).

(22) Internationales Anmeldedatum: 28. Oktober 1992 (28.10.92) (81) Bestiminungsstaaten: AU, CA, Fl, JP, KR, NO, RU, UA, Prioritatsdaten: US, europ~isches Patent (AT, BE, CH, DE, DK, ES, FR, P 41 35 977.1 3 1. Oktober 1991 (31.10.9 1) DE GB, GR, IE, IT, LU, MC, NL, SE).

(71) Anmelder (fir alic Bestimtnungssiaaten ausser US): Veroffentlicht FRAUNHOFER GESELLSC HAFT ZUR FORDE- Mit internationalem Recherchenberich.

RUNG DER ANGEWANDTEN FORSCHUNG E.V. Vor Ablauf der far Anderungen der Ansprflche :ugelasse- [DE/DEJ; Leonrodstr. 54, D-8000 Miinchen 19 izen Frist. Verdffentlichung wird wiederzol falls 4nde'rungen eintreff.

(72) Erfinder; und Erfinder/Anmielder (,iurliir US) :BRANDENBURG, Karl- Heinz [DE/DE]; Haagstr. 32: D-8520 Erlangen (DE).

GERHAUSEN, Heinz [DE/DE]: Saugendorf 17, D- 663 8551 Waischenfeld SElTZER. Dieter [DE/DE]:.

Humboldtstr. 14, D-8520 Erlangen SPORER, Thomas [DE/DE]; Wilhelmshavener Str. 29. D-8510 F~rth

(DE).

(54)Title: PROCESS FOR SIMULTANEOUSLY TRANSMITTING SIGNALS FROM N-SIGNAL SOURCES (54) Bezeichnung: VERFAHREN ZUR GLEICHZEITIGEN OI3ERTRAGUNG VON SIGNALEN AUS N-SIGNALQUEL-

LEN

SlnI q~ I SI nAIN11 ,NCfl 111 *nl,5Clt1U1 bn htl 51 t"Iil 8111 BIIOCI An 8,ruttvong dvr I~ltI~ Zltle lf An g'S I J9elr *r e9 1,cc ,lnllneflCoder JCdlcroog C~Il Codletg i, 6 IWI, Sign~ I Sn I Sig 1"t 1* LlIOCTO CAIALLATION Do Ot COOIINIVIUA CCOD-l ODN WhL lN OAPNW J (57) Abstract A process is disclosed for simultaneously transmitting signals from N-signal sources over a corresponding number of transmission channels. The individual signals are subdivided into blocks and the blocks are converted into spectral coefficients by transformation or filtering, and the latter are then subjected to a data reduction process. The invention is characterized in that the blocks belonging to the individual signals are subdivided into section. The momentary sections of all signals are processed together, the admissible disturbance is determined for each section by using a perception- specific model and the momentarily required total transmission capacity is calculated. The allocation of the maximum available transmission capacity for each individual signal is calculated from the total available transmission capacity and the total momentarily required- transmission capacity.

Each individual signal is coded and transmitted with the thus determinee! -apacity.

T

I Process for Simultaneous Transmission of Signals from N Signal Sources De s c r p ti o n Technical Field The present invention relates to a process fo simultaneous transmission of signals from N signal sources via a corresponding number of transmission channels.

I State of the Art Processes in which the individual (time) signals are divided into blocks and these blocks are transformed by Stransformation or filtering into spectral coefficients which for their part undergo a data reduction process or which are coded according to a data reduction respectively, are known. In this connection, reference is made to, by way of illustration, the overview article "Perceptual Audio coding" by J6rg Houpert in Studio-Technik or the article "Daten-DiAt, Datenreduktion bei digitalisierten Audio-Signalen" by Stefanie Renner in Elrad, 1991. These overview articles as well as the PCT A-document WO 88/01811 is explicitly referred to with regard to the explanation of any terms and process steps not made more apparent herein.

In a number of cases, it is necessary to transmit signals from several signal sources simultaneously via a corresponding number of transmission channels. The transmission of stereo signals via two transmission channels is mentioned as the most simple example therefor.

lhI S-2- The transmission of signals from N signal sources via a corresponding number of transmission channels presents the problem of dimensioning the transmission channels: If each indivilual transmission channel is dimensioned in such a manner that it transmits the "maximum incident bit flow" (German: Bit-Strom), comparatively large transmission capacity remains unused "on the average" In the transmission of signals from numerous signal sources via a corresponding number of transmission channels, it is known from digital telephone technology to design the transmission channels for only "average demand" and to balance short-term increased demand on individual channels by allotment from other channels. The allotment e0sreexclusively via signal statistics.

For the state of the art, reference is made to the following literary sources "Ein digitales Sprachinterpolationsverfahren mit pradiktionsgesteuerter Wortaufteilung" by Dr. H. Gerhauser (1980), "Ein digitales Sprachinterpolationsverfahren mit momentaner Priorit&tszuteilung", by R. Woitowitc (1977) and "Ein digitales Sprachinterpolationsverfahren mit blockweiser Priorit&tszuteilung" by G.G. Klahnenbucher (1978).

An element of the present invention is that it was understood that the usual processes in digital telephone technology for balancing fluctuating demand in the Stransmission of numerous signals via a corresponding number of transmission channels does not have good results if the digital signals to be transmitted previously underwent data reduction, by way of illustration, according to the so-called OCF process.

r

-B

-3- Description of the Invention The object of the present invention is to provide a process for simultaneous transmission of signals from N signal sources via a corresponding number of transmission channels with which "data-reduced signals" can be transmitted via transmission channels that are only dimensioned for "average demand" without any perceptable, i.e. by way of illustration audible loss in signal capacity.

A solution to this object in accordance with the present invention is set forth in claim 1. Further improvements of the invention are the subject matter of the subclaims.

The present invention is based on the fundamental idea not to conduct the allotment to the individual signals according to statistical considerations in balancing the fluctuating requirements during simultaneous transmission of signals from N signal sources via a corresponding number of transmission channels, but rather already to balance the fluctuating demand by appropriate means in the process step in which the signals are coded for data I. reduction.

This inventive fundamental idea is explained in the following using a preferred embodiment with reference to the accompanying drawings, showing: Fig. 1 a block diagram to explain the invented process, and Fig. 2a and 2b the invented signal build-up (German: Signalaufbau) i-

I-

-4- In the invented process the individual signals are divided into blocks and the blocks are transformed into spectral coefficients by transformation or filtering.: To balance the fluctuating demand, the blocks belonging to the individual signals are divided into sections, and the respective current sections of all signals are processed simultaneously. This is graphically illustrated in Fig. 1 by the corresponding "function blocks".

Employing a perception-specific model which, by way of illustration, when transmitting audio-signals can be a psycho-acoustic model, the permissable interference is determined for each section and from it is calculated the request for the currently required overall transmission capacity. This calculation of the overall transmission capacity, d.h. the required number of bits, occurs in all the blocks simultaneously. From all the transmission capacity at disposal and the currently required overall transmission capacity, the allotment of maximum transmission capacity at disposal is calculated for each individual signal. With each "number of bits" alloted to each i signal, the coding of the individual signal occurs and accordingly the transmission of this indivdual signal. In Sthe simplest case, balancing or equalizing can only occur in the re-spective required transmission capacity between the channels. 4 In the further improvement described in claim 2, there is a transmission capacity reserve, a so-called bit reservoir, from which, in the event that the required overall transmission exceeds the transmission capacity on the average at disposal, an allotment of transmission capacity occurs.

This bit reservoir is filled whenever the requested transmission capacity is less than the transmission capacity at disposal (claim 3).

In any case, it is necessary, in order to prevent an increase in the bit reservoir boointo great, if the transmission capacity is much smaller than the transmission capacity at disposal, that there is a forced allotment of bits to the individual channels (claim This forced allotment occurs preferably only to the channels, respectively the signal sources, that have reported a need that is greater than average demand. A substantially greater demand than the average demand, noteably, means that these signals are su' tantially more difficult to code than usual signals.

In any event, it is preferable according to claim 9 if an overall block is formed from all the separately coded i signals from the signal sources. This overall block is composed of a fixed region containing information from which the separation of signals can be determined and of several regions of more flexible length which receives the coded signals. This is diagrammatically shown in Fig. 2a.

Further saving in transmission capacity is achieved in that identical input signals are recognized and are transmitted only once by a suited transmission format (claim This is diagramatically shown in Fig. 2b.

In any case, the currently required transmission capacity can be accurately determined or only estimated (claims 7 and 8).

IL.-

r -6- Moreover, in a great extend the invented process can be conducted parallel. To do this, it is preferable if according to claim 10 coding of the individual signals already occurs during calculation of the allotment of the trans-mission capacity for each signal.

Another preferred realization of the invented fundamental idea is set forth in claim 11: If the required transmission capacity exceeds the transmission capacity at disposal and no allotment from the bit reservoir can occur, the value of the permissible interference for all the signals can be raised in such a manner that the required overall transmission capacity does not exceed the transmission capacity at disposal (claim 11).

In the following a numerical example of a manner of proceeding for audio signals is given. It is explicitly pointed out that the invented fundamental idea is not restricted to audio signals, but rather also video signals or other signals underlying a perception-specific assessment can be treated similarly.

Example of a possible manner of proceeding for audio signals: Assuming that y(t) are the sampling values of the audio <d signal.

1) The audio signal y is broken down or separated in a known manner into the sampling values which are digitized. The digitized sampling values are broken down or decomposed into blocks of the length 2n, which in the ioI -7selected embodiment are overlapping blocks having an overlapping of n: x(k,b) y(b*n+k) for k=o..2n (b block number).

2) Each block of length n is transformed into spectral coefficients by transformation, by way of illustration Fast Fourier Transformation or a cosinus transformation: x(j,b) SUM(1=0..2n; x(l,b)*f(1)*cos(pi*(21+1+n) (2j+l) for j=0..n with f(1) sqrt(2)*sin (pi (2n) 3) Each of the blocks is divided Lnto sections and the energy density is calculated for each section: E(i,b X(k,n) 2 for i=l c, with the coefficients a(i) being taken from the following Table 1.

4) The permissible interference is calculated for each section with a suited psycho-acoustical model, for which the literature is to be referred. The masking between the bands is yielded from the permitted interference T(i,b) MAX(k=1 i-1; E(k,b)*z(i-k) the masking in the band: s(i,b) max E(i,b) e(i) T(i,b) and the masking between the blocks: I L -8ss(i,b) max s(i,b-1)/16 s(i,b) thereon follows the calculation of the required number of bits for each block.

Calculation of the required number of bits for the block: a) for a coding like in the case of OCF (Huffman coding): p p0 SUM(i=1..C; b) for PCM coding (SNR 6dB/bit) A scaling factor and the number of bits per sampling value as additional information are transmitted for each section p pO SUM(i=l..c; 10/6 log( E(i,b) ss(i,b) The pertinent values for the individual values, respectively for the individual constants, are given in the form of tables in the following: n 512 c 23 pO 1200 for OCF (average number of bits per block) pO 345 for PCM (scaling factors: 10 bit/section, coding 4 of the number of quantization steps: 5 bits/ section TABLE 1 i 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 a(i) 0 4 8 12 16 20 24 28 32 36 40 46 52 60 70 82 SIII I i a(i) -9- 17 18 19 20 21 22 23 24 96 114 136 164 198 240 296 372 TABLE 2 1 2 3 4 5 6 7 8 9 e(i) le-1 le-2 le-3 le-4 le-5 le-6 le-7 le-8 le-9 a(1) 0 for i>9 TABLE 3 i e(i) i e(i) i e(i) 1 2 3 4 0.0004 0.0004 0.Q004 0.0004 0.0004 0.0004 0.0004 8 0.002 16 1 0.06 0.

9 10 11 12 13 14 0.002 0.002 0.004 0.01 0.015 0.025 0.04 .7 06 18 0.06 19 20 21 22 23 0.08 0.08 0.11 0.14 0.18 This is followed by the allotment of the number of bits to the individual signals. For this it is assumed that k(k)bits are requested for coding the K input signals while psoll number of bits are at disposal.

psum Sum((p)k)) Now it is necessary to differenciate: x 1) If psum=psoll Each signal receives the requested number of bits: z(k) P(k) 2) If psum<psoll Each signal receives more than the requested number of bits: z(k) (psoll/psum) p(k) K=2, psoll=1600, p(1)=540,p(2)=660 psum=1200 z(1) 1600/1200 540 720 (180 bits more) z(2) 1600/1200 660 880 (220 bits more) 3) If psoll>psum Each signal receives less than the requested number of bits: a) for OCF: z(k) (psoll/psum) p(k) b) for PCM: The minimum number of bits for each signal must not be undercut: z(k) pO (psoll-K*pO) (p(k)-pO) K=2, psoll=1600, p0=500, p(1)=600, p(2)=1200 then Psum=1800 z(l)=500+(1600-2*500)/(18C0-2*500)*(600-500)=575 bits less) z(2)=500+(1600-2*500)/(1800-2*500)*(1200-500)=1025 (175 bits less) In order to correct the permissible interference, the following differenciation is necessary if p bits are requested for each signal but z bits are alloted: 1) If the alloted number of bits equals the requested number: no correction is required.

2) If more bits were alloted than were requested: S• i I II -11- For OCF: no correction is required.

For PCM: The number of bits at disposal for quantization in each section is increased by (z-p)/512.

3) If the number of bits being alloted is less than being requested: For OCF: ss(i,b) s(i*b) (z-pO)/(p-oO for p>pO ss(i,b) s(i,b) for p<=pO For PCM: The number of bits at disposal for quantization in each j section are increased by (z-p)/512.

In the case of PCM, a rounding off bit per ATW to an integer is required: in order to do this all bits/ATW are rounded off to the next lowest integer and the resulting bit sum is determined therefrom.

If there are still bits at disposal,, one bit/ATW more is placed at disposal in a first estimation run for each band beginning with the lowest band until the number of bits at disposal is reached.

Example: 104 bits are at disposal Section: 1 2 3 4

I-

-12- ;-12trl.

Width: Bits/ATW: rounded off: *Wid:h 4 4.2 6 5.2 8 3.4 3 24 12 2.4 2 24 still to be alloted: 10 bits +1 +1 Result: 5 6 3 2 The present invention has been described hereinbefore using preferred embodiments. There are, of course, very many different variations possible within the scope of the overall inventive idea: A fixed overall block length can be employed, with filling bits being used or there 'is a transfer of not yet ended coders. Furthermore, a flexible block length can be employed which prescribes a maximum block length and in addition time averaging occurs.

Claims (11)

1. A process for simultaneous transmission of signals of N signal sources via a corresponding number of transmission channels, in which the individual signals are divided into blocks and said blocks are transformed into spectral coefficients by transformation or filtering, said spectral coefficients undergoing a data reduction process, characterized by the following features: said blocks belonging to said individual signals are divided into sections, the respective current sections of all signals are processed simultaneously, the permissible disturbance for each section is determined utilizing a perception-specific model and a request of currently required overall transmission capacity is calculated, the allotment of maximum transmission capacity which can be provided for each individual signal is calculated from the overall transmission capacity which can be provided and the currently required overall transmission capacity and each of said individual signal is coded and transmitted with the thus determined capacity.

2. Process according to claim 1, characterized by there being a reserve of transmission capacity (bit reservoir) from which an allotment occurs if the required overall transmission capacity exceeds the average transmission capacity at disposal.

3. Process according to claim 2, characterized by said bit reservoir being filled if the requested transmission capacity is smaller than the transmission capacity at disposal.

4. A process according to claim 3, characterized by a forced allotment occurring in order to prevent said bit reservoir from increasing too greatly if the requested transmission capacity is very much smaller than the transmission capacity at disposal.

I(* i 4 I -14- A process according to claim 4, characterized by said forced allotment only occurring if there is a need greater than the average need.

6. A process according to one of the claims 1 to 5, characterized by identical input signals being recognized and being transmitted only once by a suited transmission format.

7. A process according to one of the claims 1 to 6, characterized by precisely determining the currently required transmission capacity.

8. A process according to one of the claims 1 to 6, characterized by estimating said determination of said currently required transmission capacity.

9. A process according to one of the claims 1 to 15 8, characterized by an overall block being formed from all the separately coded signals from the signal sources, said overall block being composed of a fixed section containing information from which the separation of said individual signals can be determined and composed of several regions of flexible length.

A process according to one of the claims 1 to 9, characterized by said coding of said individual signals already occurring during the calculation of the allotment of the transmission capacity for each signal.

11. A process according to one of the claims 1 to characterized by, if the currently required overall transmission capacity exceeds the overall number of bits at disposal, the permissible interference for all the signal sources being increased so that a reduced bit 30 request is yielded. DATED this 24th day of November 1995 FRAUNHOFER GESELLSCHAFT ZUR FORDERUNG DER ANGEWANDTEN FORSCHUNG E. V. By their Patent Attorneys CULLEN CO. II--