DE1815962C3 - Synchronization method for address-coded telephone systems - Google Patents

Description

The invention relates to a .Synchronisierverfahren for operating an address-coded telephone system by means of analog or digital pulse modulation, with each information pulse being divided into a group of shorter pulses and coding in the time-frequency level (= matrix) the complete address of the Partners can carry and in order to ensure the high station density due to the many foreign impulses difficult address recognition by the partner through suitable processing of the information - for example modified delta modulation - a low pulse frequency is provided.

Such systems (which are often referred to in the English-language literature as »RADA« = »Random Access Discrete Address«) are essentially based on the The fact that a broadband channel is potentially available to a large number of participants at the same time, however, only a small number is actually used at any given moment. The relatively low canal load by the individual participant is not only achieved by the fact that he is only a small one Actually speaks for a fraction of the time, but particularly advantageously because he - even in the active state - i.e. during the transmission process - the channel is only briefly occupied in percentage terms.

The conversion of speech into a binary train of impulses must therefore take place in such a way that the middle Frequency of the resulting pulses remains small. This can be done e.g. B. with a modified delta modulation reach:

With the usual delta modulation, steep positive edges are reproduced using the pulse train (... 1111 ...), the negative edge through (... 0000 ...), while constant input signals through the fast changes (... 0101 ...) whose clock frequency is far above the pass band of the receiving low-pass filter is to be reproduced.

The addition modulo-2 of an alternating signal (... 010101 ...) to the delta signal leaves the following impulse groups arise:

... 101010.
... 010101.
... 111111.
... 000000.

ascending
descending

constant

Input signal

Constant signal levels can therefore be characterized either by the 0 sequence or the 1 sequence.

If they are always represented in an advantageous manner by the 0 sequence on the transmit side, then they arise on the receive side after the same modulo-2 conversion probably here and there small errors of the size of a quantization level, however, the number of pulses to be transmitted is greatly reduced: at a clock frequency of the Delta modulator of 40 kHz, the mean pulse frequency of the modified delta signal drops to values of 2

... 3 KHz. In addition, this has modified Delta modulation has the advantage that two 1-pulses can never occur in direct succession: they are always separated by at least one zero - this fact allows the use of the Double interval for pulse broadening, increasing the number of addresses or reducing the peak power.

Cluster-shaped disturbances which simulate impulses in successive clock intervals, are converted into a via the modulo-2 adder stage of the receiver after conversion (... 010101 ...) - Alternating signal converted and strongly suppressed in the receive low-pass filter

The modulation system for the message that has just been outlined in its basic features must still be supplemented by the ability to transmit addresses. The principle of the time-frequency matrix offers itself as a simple solution, among other things because the nature of the resulting systemic interference only slightly affects the voice transmission with this modulation system. With this conception, the entire available bandwidth B is divided into N channels ("channel raster"); each channel comprises η discrete frequencies ("frequency raster"), of which h frequencies are used to identify the address. Each original message pulse is replaced by a group of h pulses, each of which is broadcast at one of k possible time slots within the 50 ^ s interval. The receiver monitors the h him allocated frequencies and feeds y> the incoming pulses in the delay lines of the correct length, so that all pulses of a particular group for him occur simultaneously at the output of all the delay lines and thus mark the original message pulse.

The allocation of different addresses to a single device as well as the receipt of the priority call is without Difficulties possible.

When the transmission medium is heavily loaded by many active devices and / or by multipath propagation there is a certain probability that incorrect address pulses will arise in each receiver, whereas the probability of suppressing a correct impulse is much smaller and practical may be neglected (phase and amplitude of an interferer should have exactly the right values).

The increased false address probability has two effects, the one without countermeasures make such a transmission system unusable:

1. If the receiver demodulation channel is constantly open, all false impulses that arise from the coincidence of the h outputs of the delay lines are demodulated and lead to loud noise.

2. It is also no longer possible, a tight time gate, which scans the clock of 40 kHz for some \ is provided for this purpose the coincidence circuit and so the majority of all disturbances eliminated, to synchronize to the desired station. no

With a satisfactory solution to the synchronization problem for the time gate, the effect of the Incorrect address probability increased due to the high station density be significantly reduced; even if about 14-20% of all (> s Sampling periods due to wrong impulses would still be intelligible with exact synchronization over 85%, with simultaneous (unlikely) failure of 13% of all useful pulses (range limit) you would still have a word intelligibility of over 60% (corresponding to a signal-to-noise ratio of = 16 dB or «8 dB).

The invention lies; the task is based on the shortcomings of previously known systems of the above described type, in which practically not with a high system load by numerous active interferers was calculated for the FaI! the existence of a high station density and the effect a high probability of disturbed addresses by additional synchronization. According to the invention this is achieved in that, in order to increase the as a result of the information processing reduced frequency of address transmission by the calling subscriber station and thus in order to improve the synchronization of the station called with this address with the calling station Station each calling station periodically and at larger time intervals a longer, dense pulse train (e.g. 20th Pulses in the shortest permissible time grid spacing) emits which pulse sequence in the time-frequency level with the address of the called station is coded, and that the relative duration of this pulse train is so it is chosen briefly that neither the total system load increases significantly, nor the corresponding one Loss of information is practically perceived.

In an embodiment of the invention that can be easily implemented in practice, the procedure is as follows: triggered by pressing the transmit button and from then on the transmitter emits every 3 seconds automatically the appropriately addressed pulse train 1010 ... 10, which consists of 20 "ones" and 20 "zeros". Such a "binary word" takes up a time interval of 20 times 50 μβ = 1 ms. It gets every 3 seconds Inserted into the language on the transmit side, after detection through a corresponding gap on the receive side replaced. In the receiver takes a binary pass-through memory (shift register with taps) with a Resolving power of 50 μ5 (approx. 200 cells) that received signal.

As soon as there are 15 or more one at the right distance of 50 μβ in the pass-through memory (shift register) it considers the signal as intended for it, deletes the memory contents and is synchronized at the same time. If one demands that there are at least 5 zeros, for example, between the ones, so will Incorrect synchronization due to permanent disturbance excluded.

For the purpose of a further drastic reduction of the time and frequency-related system load the actual information pulses are completely freed from the address, so that the address recognition and the associated synchronization exclusively by means of the nested address-coded dense pulse trains (binary words).

In many cases it can also be useful to have means for recognition in the receiver of the called station several different addresses to be provided so that the recipient recognizes one of these addresses is automatically synchronized to the recognized address.

After all, you can - in order to receive a kind of "receipt" for correct To be able to receive and check reception on the other side - together with the synchronization of the recipient of the called station, its sender automatically changes to the received address

can be set so that the received address is reproduced when sending from this station.

The probability of correct recognition P _c of the word given by the address-coded, dense pulse train is iir. essential determined by the expression:

- P) " ¹

Here, P is the probability that a message pulse will not be recognized. Assuming a word intelligibility of 80% for monosyllabic words, P = 0.135 (range limit). With this value you get:

P ₁ >0.95;

m = 20
P = 0.1 ^ 5

This means that in over 95% of all extreme cases the synchronization takes place with the first binary word.

Conversely, such a word can also be faked by incorrect addresses. If one calculates with an average value of Pr \ of the false address probability (systemic disturbances and range limit) of 17%, the result for the probability Pv of the pretense is:

= Σ

I m = 20

<■ J

P _1Λ = 0.17

2 K)

In comparison with the resolution of the memory (5 με), this means that a false start of the synchronization occurs on average every 4 minutes if the system is heavily loaded per address. After such a start, Reception! - is blocked for 3 seconds, but opens again immediately if the next packet is missing.

To make a possible loss of synchronization during the call unlikely, one could meet the requirements in the synchronized state reduce and, for example, only require the reception of 10 or less one.

With such a synchronization device it is possible to establish connections even when the system is heavily loaded build and maintain. At the same time, it can be used to synchronize any encryption devices to serve.

In special cases it is even possible to let the transmitter - in agreement with the receiver - do without the complete addressing of all message pulses after synchronization and to use only two or in extreme cases only one pulse instead of three or four. The "wrong address probability" does not increase significantly. since the system load is correspondingly smaller (Ργα ~ 20 ... 25% with the same number of speakers and only one impulse). The main advantage is a corresponding reduction in the average transmission power.

Claims (6)

Patent claims: 1. Method for operating an address-coded telephone system using analog or digital Pulse modulation, where jt the information pulse by dividing it into a group of shorter ones Pulses and coding in the time-frequency level (= matrix) the complete address of the partner can wear and to ensure the high ι ο Station density due to the large number of external impulses made address recognition difficult by the partner through suitable processing of the information - for example modified delta modulation - for low pulse frequency is provided, characterized in that in order to increase the as a result of the information processing with a reduced frequency of address transmission the calling subscriber station and thus for the purpose of improving the synchronization with it Address called station with the calling station each calling station periodically and in larger A longer, dense pulse sequence (e.g. 20 pulses in the shortest permissible time grid spacing) emits which pulse sequence in the time-frequency level is encoded with the address of the called station, and that the relative duration of this pulse train is chosen so short that neither the overall system load increased significantly, nor the corresponding loss of information practically table is perceived. 2. The method according to claim 1, characterized in that the address code comprising Pulse train (binary word) on the transmitting side in certain, compared to the duration of the pulse train large Time intervals (e.g. every 3 seconds) inserted into the voice signal and on the receiving side after recognition is replaced by a corresponding gap. 3. The method according to claim 1 or 2, characterized in that for the purpose of further, more drastic Reduction of the temporal and frequency-related system load the actual information impulses be completely freed from the address, so that the address recognition and associated Synchronization exclusively by means of the activated address-coded dense pulse trains (Binary words) is made. 4. The method according to claim 1 or 2 and 3, characterized in that the receiver called station a binary pass-through memory in the form of a tapped shift register is provided that picks up the received signal and in the presence of a larger one Number of binary "one" signals following one another at the correct distance automatically deleted which simultaneously synchronizes the recipient of the called station to the address he follows. 5. The method according to claim 1 or 2, 3 and 4, characterized in that the receiver called station means for recognizing several different addresses are available and that the Recipients are automatically synchronized to the recognized address when one of these addresses is recognized will. ds 6. The method according to claim 1 or 2, 3 and 4 and according to claim 5, characterized in that that together with the synchronization of the recipient of the called station also their Sender is automatically set to the received address, so that when sending from this Staiion is reproduced from the received address.