JPH01125153A - Digital fm demodulator - Google Patents

Abstract

PURPOSE:To decrease the error rate by varying a code discrimination level depending on a preceding data and offsetting a timing of the code discrimination. CONSTITUTION:Means 2, 5, 6 varying the code discrimination level of succeeding bits in a direction cancelling the inter-code interference by the date in response to the output of a code discrimination circuit 3, and a means 8 generating a sampling pulse whose phase is led with respect to a clock signal extracted from the FM detection output by a clock synchronization circuit 7 are provided. Thus, in the presence of the inter-code interference, a frequency detection waveform varies with the data of preceding/succeeding bits. Then accurate code discrimination is applied by discriminating the detection waveform depending on the preceding data. Thus, the bit error rate is reduced.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention is utilized for detection demodulation of a signal modulated by a narrowband digital frequency modulation method. Narrowband digital frequency modulation is a modulation method that effectively utilizes a limited frequency band, and is particularly widely used in mobile communication devices.

[Conventional technology]

In the narrowband digital frequency modulation method, in order to narrow the spectrum of the modulated wave, high frequency components of the baseband signal are removed before modulation is performed.

That is, a low-pass filter with a bandwidth Bb is inserted before the frequency modulator, and the digital data signal that has passed through this low-pass filter is frequency-modulated. A GMSK modulation method is known as a typical example of such a modulation method. Regarding the GMSK modulation method, see Nitride, Hirata,
"Transmission Characteristics of rGMSK Modulation System", Transactions of the Institute of Electrical Communication Engineers (B), Vol. J64-B, No. 10, pp. 1123 to 1130 (October 1982).

[Problem that the invention seeks to solve]

However, in the GMSK modulation method, high frequency components of the data signal are removed, so a waveform in which intersymbol interference occurs is supplied to the input of a frequency modulator, and this waveform is frequency-modulated. If this waveform is frequency-detected on the receiving side, the eye opening at the time of sampling becomes small, and even if the code is determined and the transmitted data is reproduced, the code error becomes large.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above problems and provide a digital FM demodulator that can reproduce transmitted data with small code errors.

[Means for solving problems]

The digital FM demodulator of the present invention includes means for changing the sign determination level of subsequent bits in a direction to cancel intersymbol interference caused by the data according to the output of the sign determination circuit;
The clock synchronization circuit is characterized in that it includes means for generating a sampling pulse whose phase is advanced with respect to the tarok signal extracted from the FM detection output.

[For production]

When there is intersymbol interference, the frequency detection waveform changes depending on the data of the previous and subsequent bits. Therefore, if the detected waveforms are distinguished based on previous data, accurate sign determination can be performed. Possible changes in the detected waveform due to intersymbol interference include changes in voltage level and deformation of the eye. Therefore, in the digital FM demodulator of the present invention, the code determination level is changed depending on the previous data, and the timing of code determination is offset. As for the amount of offset, it is desirable to set the timing at which the eye opening is maximum.

〔Example〕

FIG. 1 is a block diagram of a digital FM demodulator according to an embodiment of the present invention.

An FM detection output is supplied to the input terminal 1 from the FM detection circuit at the previous stage. Input terminal 1 is connected to sign determination circuit 3 via adder circuit 2 . The sign determination circuit 3 is connected to the output terminal 4 and branched to the shift register 5.
connected to. Shift register 5 is connected to voltage generation circuit 6. Voltage generating circuit 6 is connected to adder circuit 2 .

Input terminal 1 is also connected to clock synchronization circuit 7 .

Clock synchronization circuit 7 is connected to sampling pulse generation circuit 8 . The sampling pulse generation circuit 8 is connected to the sign determination circuit 3.

The sign determination circuit 3 determines the sign of the FM detection output and reproduces digital transmission data. The clock synchronization circuit 7 is
A clock signal is extracted from the same FM detection output. The sampling pulse generation circuit 8 supplies sampling pulses to the sign determination circuit 3 in response to the output of the clock synchronization circuit 7.

Shift register 5, voltage' generation circuit 6 and addition circuit 2
constitutes a feedback circuit, and changes the voltage level of subsequent bits in a direction to cancel intersymbol interference caused by the data according to the output of the sign determination circuit 3, thereby changing the sign determination level. The sampling pulse generation circuit 8 generates a sampling pulse whose phase is advanced from the clock signal supplied from the clock synchronization circuit 7 so that the sign determination can be performed at the time when the eye opening is maximum.

In this embodiment, the sign determination circuit 3 is configured to determine the sign based on the sign of the output of the adder circuit 2, and the sign determination level is changed by changing the voltage level of the FM detection output. On the other hand, the present invention can be implemented in the same way even if the determination threshold of the sign determination circuit 3 is changed without changing the voltage level of the FM detection output.

FIG. 2 shows the waveform after receiving the transmitted frequency modulated wave and detecting the frequency using a limiter disk, and FIG. 3 shows the frequency detected waveform when the previous bit data is a mark.

If previous bits are not taken into account, the eye opening becomes smaller due to intersymbol interference, as shown by diagonal lines in FIG. On the other hand, when the data one bit before is a mark, the eye is wide open.

At this position of the eye, the voltage level is high due to the inter-frame interference caused by the data from the 1-pit old man, and the eye opens earlier. Furthermore, when the data one bit before is a space, the voltage level of the eye decreases.

The eye opening timing is the same as when the previous bit data is a mark. Therefore, by offsetting the sampling timing and changing the sign determination voltage level depending on the previous data, accurate sign determination can be performed.

This will be explained in more detail using mathematical formulas.

If the pulse response of the low-pass filter inserted before the frequency modulator is g (t), what is the frequency detection waveform when using the GMSK modulation method? '(t) is expressed as. Here, ah represents the transmission data series, and its value is "+1
” or “-1”. Further, T represents the bit length.

When the bandwidth B of the low-pass filter is narrowed, intersymbol interference increases, and the tail of the pulse response g (t) becomes longer. However, if B, T ≧0.2, the pulse response g (t) converges to approximately zero with l t−nT l >3T/2. Therefore, in order to determine a, the frequency detection waveform! When '(t) is sampled at t=nT, weight'(nT) = -Can-+g(T) + a, g (0) 10a,,. 1g (-T) )T. ah-1, an and a,. , are independent data, the eye opening is π.

On the other hand, if the sampling timing is advanced by xT bits (0<X<1> and t=(n-x)T, then ((2-X)T) + an
-1g ((1-x)T)T + a, g (-xT) + a h+l g (
(-1-X)T) : ]・ (3) It becomes. In this equation, an-2 and ah-1 are obtained as past determination results, so by using these values to change the code determination voltage level, it is possible to remove the influence of intersymbol interference. can. therefore,
In order to determine a, , -Ca, g (-xT)
It is only necessary to make a determination on the waveform of +ag+I g ((-1-X)T) :)T. The eye opening at this time is 2 ・-[: g (-xT) -g ((-1-X)
T) ]T.

FIG. 4 shows an example of the relationship between sampling timing and eye opening. In mobile communications, B, T = 0.25 is used in consideration of the spectrum concentration of modulated waves. Therefore, in Figure 4, BbT is 0.2
An example of case 5 is shown.

As in the past, when the sampling timing is set to nT without feeding back the previous data, the eye opening is 2.
π/2Tx0.31. On the other hand, if data from one bit earlier is fed back, the influence of intersymbol interference caused by that data can be removed, and the eye opening increases. Furthermore, the sampling timing is set to (n-0,2
5) If T, the eye opening is 2・π/2TX0.52
This is approximately 1.7 times the conventional value. If the data from 1 bit before and the data from 2 bits before are fed back, the eye gap becomes even larger. Therefore, the bit error rate characteristics can be greatly improved.

FIG. 5 shows measurement results of bit error rate versus carrier-to-noise power ratio. This figure shows 16kb/s G MSK
When using the device of this embodiment with a modulation method and not feeding back previous data, or feeding back 1 bit previous data but not offsetting the sampling timing,
Also, the measurement results are shown when the data one bit earlier is fed back and the sampling timing is offset by 174 bits. Furthermore, 16kb/sM S
The bit error rate with respect to the carrier-to-noise power ratio in the case of the K modulation method is also shown.

As shown in FIG. 5, the bit error rate can be greatly reduced simply by feeding back previous data, and the bit error rate can be further reduced by advancing the sampling timing by 1/4 bit.

〔Effect of the invention〕

As explained above, the digital FM demodulator of the present invention can reduce the error rate by changing the code determination voltage level using previous data, and by offsetting the sampling timing at that time. , it is possible to further reduce the bit error rate. The present invention is particularly effective when applied to a digital FM mobile communication device.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a digital FM demodulator according to an embodiment of the present invention. FIG. 2 is a diagram showing a frequency detection waveform. FIG. 3 is a diagram showing a frequency detection waveform when the data one bit before is a mark. FIG. 4 is a diagram showing an example of the relationship between sampling timing and eye opening. FIG. 5 is a diagram showing measurement results of bit error rate versus carrier-to-noise power ratio. DESCRIPTION OF SYMBOLS 1... Input terminal, 2... Addition circuit, 3... Sign determination circuit, 4... Output terminal, 5... Shift register,
6... Voltage generation circuit, 7... Clock synchronization circuit, and... Sampling pulse generation circuit. Patent Applicant Nippon Telegraph and Telephone Corporation Agent Patent Attorney Nao Takashi Ide W:J 1st In-117nT (n,
I IT 2nd π 3 times × 2 × Uma T X ; ji! ! i4 Figure CNR (dB) 5th port

Claims (1)

[Claims] (1) A code determination circuit that determines the sign of the FM detection output and reproduces digital transmission data; a clock synchronization circuit that extracts a clock signal from the FM detection output; and a clock synchronization circuit that extracts the clock signal from the FM detection output; In a digital FM demodulator equipped with a sampling pulse generation circuit that supplies sampling pulses to a determination circuit, the sign determination level of subsequent bits is changed in accordance with the output of the sign determination circuit in a direction to cancel intersymbol interference caused by the data. A digital FM demodulator, characterized in that the sampling pulse generation circuit includes means for generating a sampling pulse whose phase is advanced with respect to the clock signal supplied from the clock synchronization circuit.