DE19818408C1 - Modulated signal recognition device for telecommunications or data processing - Google Patents

Abstract

The signal recognition device (1) has at least one reference unit (2), a number of correlation modules (5) and an amplitude discriminator (7). The reference unit supplies each reference point to each correlation module, for mixing with the multi-value modulated input signal via an integrated mixer (8), with parallel comparison of the correlation module outputs via the amplitude discriminator.

Description

The invention relates to a device for recognizing multivalued modulated signals for communication and data technology.

To increase the transmission rates in telecommunications and data technology higher-quality modulation methods are known, by means of which a lot is simultaneously number of bits can be transmitted. For the simultaneous transmission of two bits For example, a vector is transmitted that has four different angular positions can have, each angular position corresponds exactly to a 2-bit combination. This vector can also be understood as a symbol, the egg again at the receiving end ne associated bit combination must be assigned. The higher the quality of the modules tion takes place on the transmitter side, the more difficult it becomes for the receiver to unambiguously identify symbol due to distortion and noise. This is the signal an analog-to-digital converter on the receiving side and the output signal on finally evaluated by a digital signal processor. Especially at high Clock speeds and multi-stage modulation types increase the circuit complexity for the analog-to-digital converter, which leads to considerable manufacturing costs.

The invention is therefore based on the technical problem, a device for Er identification of multi-value modulated signals to create the circuitry is easy to implement and has an improved detection rate.

DE 39 22 972 and US 5,537,396 describe methods for evaluating Spread spectrum signals previously known. The disadvantage of this method is that agile circuitry implementation.

The solution to the technical problem results from the features of the patent Proverb 1. For this purpose, all are in a reference unit using reference bases possible changes in the symbols are stored and each in a correlation Module with a real input signal  correlated synchronously, the correlation results using the amplitude Discriminators are comparable, reducing the full energy of the receiving symbols is used in the recognition, resulting in a minimal error rate in the symbol recognition leads. By additional known methods of error correction, coding and tape storage, the detection accuracy can be further optimized. Another The main advantage of the device is the complete absence of analog Digital converter. The device is rather inexpensive with standard System modules can be assembled and can be easily integrated. About that In addition, the device is for any multi-value modulation method suitable. Depending on the system, there may also be a delay spread of the receive symbols be tolerated. This delay spread can even target another improved symbol recognition by means of scaling and an extended Decision logic can be evaluated.

Due to the comparatively low requirements regarding Processing speed, clock rate and signal processing have the Device also has a low specific power consumption. You can also store user-specific reference symbols in the Reference unit of the receiver adaptive to the concrete Transmission system or adapted to changing channel properties become. The device is also for so-called software receivers usable. Further advantageous embodiments of the invention result itself from the subclaims.

The invention is based on a preferred Embodiment explained in more detail. The figure shows:

Fig. 1 is a vector representation of a tetravalent symbol

Fig. 2 shows a voltage curve between two symbols

Fig. 3 is a block diagram for recognizing the symbol.

In FIG. 1, a symbol is shown, which can take four different states. This allows two bits to be represented with the symbol. The symbol can be understood as a rotating pointer that has a constant amount. How high-quality the pointer can be modulated is limited by the angle resolution at the receiving end. In addition to the state of the symbol, however, the change course from one symbol state to another symbol state is also characteristic. In Fig. 1, three possible change courses are shown as an example. The change from a state 21 of the symbol to a state 22 results, for example, in a change in the voltage profile, as shown in FIG. 2.

A device 1 for recognizing a symbol is shown in FIG. 3. The device 1 comprises a reference unit 2 , a microprocessor 3 , a digital-to-analog converter 4 , a multiplicity of correlation modules 5 , an input amplifier 6 and an amplitude discriminator 7 . The correlation modules 5 are constructed identically and each comprise a number of switches S ₁₁ -S _1n , a number of switches S ₂₁ -S _2n , a number of capacitors C ₁ -C _n , a mixer 8 and a low-pass filter 9 . The number of correlation modules 5 corresponds to the number of possible change courses of the symbol states, whereas the number of switches S _1n , S _2n or capacitors C _{n is} determined by the number of reference points of a change course. In the reference unit 2 , which is embodied, for example, as a ROM, n reference reference points are stored for each change course. For reference, five reference points are shown in FIG. 2.

First, the correlation modules 5 are loaded with their reference points. This is explained in detail below for the uppermost correlation module 5 . For this purpose, the switch S _{11 is} closed, whereas all other switches S ₁₂ -S _1n , S ₂₁ -S _2n remain open. Via the microprocessor 3 , the associated first reference base for the relevant change course is read out of the reference unit 2 . The digital reference base is converted by the digital-to-analog converter 4 into an analog, electrical signal, which charges the capacitor C ₁ via the switch S ₁₁ . Then switch S _{11 is} opened and switch S _{12 is} closed. The microprocessor 3 then reads out the second reference base, with which the capacitor C ₂ is then charged. This process is repeated until the capacitor C _n is also charged with its associated reference base. This process takes place successively for all correlation modules 5 . At the end of the charging process, all capacitors C ₁ -C _{n of} all correlation modules 5 are charged with a specific reference point and all switches S ₁₁ - S _{1n are} open. If an electrical signal is now received, which is intended to represent a specific symbol, this is amplified by the input amplifier 6 and passed on to all mixers 8 of the correlation modules 5 . In parallel, all switches S _{21 of} the correlation modules 5 are closed, so that the respective first reference reference points are present at the mixer 8 as precisely as possible in phase with the input signal. The subsequent switches S ₂₂ to S _2n are then closed synchronously in series, with only one switch S ₂₁ -S _2n per correlation module 5 being closed. The low-pass filter 9 connected downstream of the mixer 8 assumes two functions. On the one hand, high-frequency spectral components produced due to the folding at the mixer 8 are suppressed and, on the other hand, the low-frequency spectral components are integrated. This process runs in parallel for all correlation modules 5 , so that after the last capacitor C _{n has been} switched on, all results are present in parallel on the amplitude discriminator 7 . The amplitude discriminator 7 then uses the amplitudes to select the reference signal with the best correlation to the input signal, which can then be used to infer the transmitted symbol. This is followed by programmable symbol-bit decoding, which is not shown here.

The synchronous feeding of the reference points to the mixer 8 can be realized for example by means of a delay locked loop structure (DLL). Since the capacitors C ₁ -C _n discharge over time, they must be periodically recharged by the reference unit 2 .

Reference list

1

contraption

2nd

Reference unit

3rd

microprocessor

4th

Digital to analog converter

5

Correlation module

6

Input amplifier

7

Amplitude discriminator

8th

mixer

9

Low pass

Claims (5)

1. Device ( 1 ) for recognizing multivalued modulated signals, comprising at least one reference unit ( 2 ), a plurality of correlation modules ( 5 ) and an amplitude discriminator ( 7 ), each correlation module ( 5 ) Change course of the multi-valued modulated input signal is assigned and reference points of the change course can be supplied by the reference unit ( 2 ), each of which can be successively mixed and integrated with the modulated input signal at a mixer ( 8 ), with all correlation modules ( 5 ) are connected in parallel on the output side to the amplitude discriminator ( 7 ). 2. Device according to claim 1, characterized in that the correlation modules ( 5 ) are formed for each reference base with two mutually independent switchable switches (S ₁₁ -S _1n ; S ₂₁ -S _2n ), between which a capacitor (C ₁ -C _n ) is arranged, the capacitor (C ₁ -C _n ) via the first switch (S ₁₁ -S _1n ) with the reference unit ( 2 ) and via the second switch (S ₂₁ -S _2n ) with the mixer ( 8 ) is connectable. 3. Device according to claim 1 or 2, characterized in that the reference unit ( 2 ) is designed as fuzzy logic, which is connected to the output of the discriminator ( 7 ). 4. Device according to one of claims 1 to 3, characterized in that a low-pass filter is arranged between the mixers ( 8 ) and the amplitude discriminator ( 7 ). 5. Device according to one of the preceding claims, characterized in that the switches (S ₁₁ -S _1n ; S ₂₁ -S _2n ) are designed as MOS-FETs.