WO2011132299A1

WO2011132299A1 - Reception device and reception method

Info

Publication number: WO2011132299A1
Application number: PCT/JP2010/057187
Authority: WO
Inventors: 義徳阿部
Original assignee: パイオニア株式会社
Priority date: 2010-04-22
Filing date: 2010-04-22
Publication date: 2011-10-27

Abstract

The disclosed reception device is highly resistant to spurious interference and is implemented with a simple circuit. Said reception device has: a time-domain processing unit (2) which transforms a received OFDM signal into a complex baseband signal; a window extraction unit (6) which extracts a segment of a prescribed length from the complex baseband signal or a signal obtained by further processing same; an FFT unit (7) which performs an FFT on the segment extracted by the window extraction unit (6); an adaptive-prediction difference filter (11) which outputs a prediction-residual signal for a signal obtained by downsampling the complex baseband signal by a factor of N (N being a natural number); and a symbol synchronization unit (21) which controls the window extraction position of the window extraction unit (6) on the basis of the prediction-residual signal. Performing symbol synchronization on the basis of a signal whitened by an adaptive-prediction difference filter (11) improves the resistance of the symbol synchronization to spurious interference.

Description

Receiving apparatus and receiving method

The present invention relates to a receiving apparatus and a receiving method for receiving and demodulating a digital modulation signal conforming to the OFDM system.

OFDM (Orthogonal Frequency Division Multiplex) modulation is known as a modulation system that is highly resistant to multipath interference, such as European DVB-T (Digital Video Broadcasting-Terrestrial), Japan's ISDB-T (Integrated-Terregarded-B). ), Etc., are adopted in the terrestrial digital broadcasting standards of each country.

The OFDM modulated wave transmitted from the broadcast station is received in a form in which various interference waves such as impulse noise, spurious interference, and co-channel NTSC interference are superimposed on the transmission path. In the configuration of the OFDM receiver, improvement of resistance to various interference waves is an important issue. In particular, in-vehicle receivers and portable receivers that are assumed to be used in various reception environments, high resistance to these various interference waves is required. The present invention relates to a technique for improving resistance to spurious interference among these interferences. The spurious interference mentioned here refers to an interference wave having a spectrum with a very narrow bandwidth (several kHz) compared to the transmission bandwidth of the OFDM modulated wave (about 5.57 MHz in the case of ISDB-T). Shall point to.

In OFDM modulation, since transmission data is distributed and transmitted over a plurality of subcarriers, it is easier to take measures against spurious interference than other modulation schemes. The most general and simple countermeasure is to reduce the influence of interference by treating subcarriers that are significantly affected by spurious interference among received subcarriers as erasures. In terrestrial television broadcasting, the number of subcarriers is on the order of several thousand. Sufficient transmission quality can be maintained even if tens of subcarriers that are significantly affected by spurious interference are treated as erasures. Normally, this measure can provide sufficient spurious interference resistance.

However, in severe spurious interference environments encountered by mobile reception, the above measures are often insufficient. For example, in a very severe spurious interference environment where the power of the spurious interference wave is higher than the power of the desired OFDM wave, synchronization processing such as symbol synchronization processing or narrowband AFC may not be performed properly. .

Fig. 7 shows the general configuration of an OFDM receiver. The receiving apparatus 300 includes a time domain processing unit 2, a window extraction unit 6, an FFT processing unit 7, a guard correlation calculation unit 4, a window position control unit 5, a frequency control unit 8, and a frequency domain processing unit 50. Have.

The time domain processing unit converts the input RF signal into a complex baseband signal and outputs it. The time domain processing unit generally performs processing such as RF filtering, frequency conversion to IF signals, IF filtering, A / D conversion, sampling frequency conversion, and quadrature detection. The window extraction unit extracts a window having an effective symbol length from the complex baseband signal. The FFT processing unit performs FFT processing on the window-extracted signal and outputs a received OFDM symbol composed of a plurality of received subcarriers. The frequency domain processing unit performs processing such as channel estimation, equalization, and error correction on the received OFDM symbol, and restores and outputs the transmitted information data. The guard correlation calculation unit calculates a guard correlation signal from the input complex baseband signal and outputs it. The guard correlation calculation unit will be described in detail later. The window position control unit controls the window extraction unit based on the guard correlation signal so that window extraction at an appropriate window position is performed. In general, this window position control processing is called symbol synchronization processing. Here, the symbol synchronization processing is achieved by a symbol synchronization unit 21 including a guard correlation calculation unit and a window position control unit. The frequency control unit controls the time domain processing unit based on the guard correlation signal so that the modulation frequency of the complex baseband signal becomes exactly zero. In general, this frequency control processing is called narrowband AFC processing.

Here, the internal configuration of the guard correlation calculation unit will be described. FIG. 8 shows a general configuration of the guard correlation calculation unit. The delayed correlator 12 performs a complex correlation operation between the input complex baseband signal and a signal obtained by delaying the complex baseband signal by the effective symbol length, and outputs the result. The moving averager 13 calculates a moving average of the guard length width for the output signal of the delay correlator and outputs this as a guard correlation signal. Other methods for configuring the guard correlation calculation unit include performing averaging between different OFDM symbols after taking a moving average, or using a moving average width different from the guard length, etc. There is.

The function of the guard correlation calculation unit, symbol synchronization processing, and narrowband AFC processing will be described in detail with reference to FIG. As shown in FIG. 9A, the OFDM modulated signal is composed of a guard interval and an effective symbol interval in the time domain. The rear end portion of the effective symbol section is copied to the guard section arranged at the symbol head. By performing a complex correlation operation between such an OFDM signal and a signal obtained by delaying the OFDM signal by an effective symbol length, that is, the signal shown in FIG. 9B, the output of the delayed correlator is as shown in FIG. As shown, it is expected that a highly correlated section with a guard width will appear periodically. Further, at the output of the moving averager, it is expected that a triangular wave peak is observed at the boundary of the transmission symbols as shown in FIG. 9 (d). The window position control unit detects the triangular wave peak, and controls the window extraction unit based on the timing at which the peak is detected. On the other hand, the frequency control unit controls the conversion frequency of the time domain processing unit based on the angle of the detected peak value on the complex plane.

Suppose here that spurious interference is superimposed on the received signal. In such a case, it is known that a DC offset occurs in the guard correlation signal. This phenomenon is caused by spurious interference waves having a very high time domain correlation. When this DC offset component is a negative real number, the guard correlation signal is offset downward as shown in FIG. In this case, since the symbol synchronization unit and the frequency control unit cannot detect the peak point, the synchronization process is not appropriately performed.

A reception method (see Patent Document 1) that solves the above problem has already been proposed. In this prior art, an interference wave is detected after FFT processing, and a component near the interference frequency is removed from the complex baseband signal by a notch filter. And the above-mentioned problem is solved by performing a synchronous process using the signal which removed this interference wave. However, in this method, it is necessary to prepare a disturbance detector and a notch filter separately, and in order to cope with a plurality of spurious disturbances, it is necessary to prepare a plurality of notch filters. It ’s hard to say. Further, in order to efficiently remove spurious interference waves, complicated control is required. For example, if the number of spurious jamming waves is larger than the number of notch filters and the power of each spurious jamming wave fluctuates, the notch filters must be adaptively allocated to high power jamming waves. Very complex control is required.

JP 2006-174218 A

As described above, a general OFDM receiver may not be able to properly perform symbol synchronization in a severe spurious interference environment. Although a method for solving this problem has been proposed, it has a problem to be improved from the viewpoint of cost or complexity of control.

An object of the present invention is to realize a receiver having high resistance to spurious interference waves with a simple circuit configuration.

In order to solve the above-mentioned problems, the invention according to claim 1 is characterized in that a time domain processing means for converting a received OFDM signal into a complex baseband signal, and the complex baseband signal or a signal obtained by further processing it. A window extracting means for extracting a section of a predetermined length from
FFT processing means for performing FFT processing on the window extracted by the window extraction means, and a prediction residual signal for a signal obtained by thinning the complex baseband signal with a thinning rate of 1 / N (N is a natural number) And an adaptive prediction difference filter means for performing, and a symbol synchronization means for controlling the extraction window position of the window extraction means based on the prediction residual signal.

In order to solve the above problem, the invention according to claim 6 is a time domain processing step for converting a received OFDM signal into a complex baseband signal, and further obtained by processing the complex baseband signal or the complex baseband signal. A window extraction step for extracting a section of a predetermined length from the obtained signal, an FFT processing step for performing FFT processing on the section extracted by the window extraction means, and the complex baseband signal N: 1 (N is An adaptive prediction difference filter step for outputting a prediction residual signal for a signal thinned out at a thinning rate of a natural number), and a symbol synchronization step for controlling the extraction window position of the window extraction means based on the prediction residual signal. Have.

The block diagram which shows the structure of 1st Embodiment. Simulation results of adaptive prediction difference filter Table showing simulation conditions The block diagram which shows the structure of 2nd Embodiment. The block diagram which shows the structure of 3rd Embodiment. Simulation results of decimation part and adaptive prediction difference filter The block diagram which shows the structure of the conventional general receiver Block diagram showing a configuration example of the guard correlation calculation unit Diagram showing the calculation process of guard correlation signal

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(First embodiment)
A configuration example of a receiving apparatus according to the first embodiment of the present invention is shown in FIG. Components having the same functions as those in FIG. 7 are denoted by the same reference numerals as those in FIG. The receiving device 100 includes a time domain processing unit 2, a symbol synchronization unit 21, a window extraction unit 6, an FFT processing unit 7, a frequency control unit 8, an adaptive prediction difference filter 11, and a frequency domain processing unit 50. is doing. Further, the symbol synchronization unit includes a guard correlation calculation unit 4 and a window position control unit 5. Here, the only difference from the conventional receiver of FIG. 7 is that an adaptive prediction difference filter is placed in front of the symbol synchronization unit. The following description focuses on the function of the adaptive prediction difference filter, which is a new element, and its effect.

In addition, the arrow which shows the flow of the signal in a figure shows the flow of the main signal between each component, For example, regarding signals, such as a response signal and a monitoring signal accompanying such a main signal, The case where it is transmitted in the direction opposite to the arrow in the figure is included. Furthermore, the arrows in the figure conceptually indicate the flow of signals between the components, and in an actual device, it is not necessary for each signal to be faithfully exchanged along the path indicated by the arrows. . Moreover, in an actual apparatus, it is not necessary that each component is divided faithfully as shown in FIG.

The time domain processing unit corresponds to the time domain processing means described in each claim, the adaptive prediction difference filter corresponds to the adaptive prediction difference filter means, and the window extraction unit corresponds to the window extraction means. The symbol synchronization unit corresponds to symbol synchronization means, the FFT processing unit corresponds to FFT processing means, and the frequency control unit corresponds to frequency control means.

First, a known adaptive prediction difference filter will be described. The prediction difference filter is a filter that predicts a current value of an input sequence from an input sequence other than the current value, and outputs a difference between the predicted value and the current value, that is, a prediction residual. Among the prediction difference filters, the adaptive prediction difference filter has a function of autonomously controlling the filter characteristics so that the power of the prediction residual is minimized.

Here, the example of a structure of a prediction difference filter is shown. In general, the predicted value is calculated by an FIR filter. In this case, the predicted value p _k of the input x _k at time k, the input sequence of past M taps _{_{{x k-1, x k}} -1, ..., x kM} and the prediction coefficients {c- _1, c- _2, ..., c- _M } is calculated as follows:
p _k = c ₁ x _k-1 + c ₂ x _k-2 +… + c _M x _kM
The prediction residual y _k is given by the following equation.
y _k = x _k -p _k

An adaptive prediction difference filter can be configured by updating the prediction coefficient with an appropriate algorithm. As the coefficient update algorithm, the LMS algorithm represented by the following equation is most commonly used.
c _n ← c _n + u y _k x _kn ^* (n = 1,2,…, M)
Where u is the appropriate update step and * is the complex conjugate operator.

Note that the configuration of the adaptive prediction difference filter described above is merely an example. In the above, the so-called forward prediction that predicts the current value from the past series is exemplified, but backward prediction that predicts the current value from the future series or bidirectional prediction that predicts the current value from both the past series and the future series is used. Even in this case, there is no difference in the effect of the present invention. The same applies to the case where an IIR filter is used for the filter structure and the case where an RLS algorithm is used for the coefficient update algorithm.

It is well known that the adaptive prediction difference filter has the property of whitening the signal. That is, the adaptive prediction difference filter operates so as to minimize the autocorrelation within the range of the tap length of the output signal. In the frequency domain, this is nothing other than controlling the frequency characteristics so that the power spectrum of the output signal is made as flat as possible, that is, whitened. Due to this whitening property, the power of the spurious interference wave is suppressed at the output of the adaptive prediction difference filter.

Fig. 2 shows the computer simulation result showing the spurious interference suppression effect of the adaptive prediction differential filter. The conditions used for the simulation are shown in FIG. In this simulation, a signal obtained by superimposing three spurious interference waves on an OFDM modulated signal is input to an adaptive prediction difference filter. The power spectrum of the input signal obtained by the simulation is shown in FIG. In addition to the rectangular OFDM spectrum that spreads in a band of about 5.57 MHz centering on DC, three linear spectra due to spurious interference waves can be seen. FIG. 2B shows the frequency characteristics of the prediction difference filter after the coefficient is updated for a sufficient period in a state where such a signal is input. It can be seen that zeros are formed corresponding to the frequencies of the three spurious interference waves. At this time, the output spectrum of the adaptive prediction difference filter, that is, the power spectrum of the prediction residual signal is as shown in FIG. It can be seen that the spurious interference wave component is greatly suppressed compared to the input power spectrum. As can be seen from the above simulation results, the adaptive prediction difference filter autonomously, i.e., externally, filters the input signal including a plurality of spurious interference waves to suppress these interference waves. Can be formed without the need for control from

As described above, the adaptive prediction difference filter has the property of suppressing spurious interference waves. In the receiving apparatus according to the present embodiment, an output signal of the adaptive prediction difference filter, that is, a signal in which spurious interference waves are suppressed is input to the symbol synchronization unit. As a result, the DC offset generated in the guard correlation signal is reduced as compared with the conventional receiver that does not have an adaptive prediction difference filter, and symbol synchronization processing and narrowband AFC processing can be performed appropriately.

Note that the number of spurious interference waves that can be suppressed by the adaptive prediction difference filter increases according to the number of taps. However, it is not preferable to increase the number of taps more than necessary. This is because when the impulse response of the adaptive prediction difference filter is long, the waveform of the guard correlation signal itself changes, which may adversely affect the synchronization processing. For the same reason, it is understood that an FIR filter that can limit the length of the impulse response within the number of taps is desirable as the structure of the adaptive prediction difference filter. In the above computer simulation, a 16-tap FIR type adaptive prediction difference filter is used. In this number of taps, the impulse response is limited to be sufficiently smaller than the effective symbol length, so the above-described adverse effect is extremely small, and substantial performance degradation can be ignored. On the other hand, it will be apparent from the simulation results that this tap number has a sufficient spurious interference suppression effect.

(Second Embodiment)
An example of the configuration of a receiving apparatus according to the second embodiment of the present invention is shown in FIG. In addition, about the component which has a function equivalent to FIG. 1, the code | symbol same as FIG. 1 is attached | subjected, and description is abbreviate | omitted. The difference between the receiving apparatus shown in FIG. 1 and the receiving apparatus shown in FIG. 4 is only the input source of the window extraction unit. In FIG. 1, the output signal of the time domain processing unit is input to the window extraction unit, whereas in FIG. 4, the output signal of the adaptive prediction difference filter is input to the window extraction unit. In the configuration of FIG. 4, since FFT processing and frequency domain processing are performed based on a signal in which spurious interference is suppressed, there is a possibility that the influence of interference can be further reduced as compared with the configuration of FIG. It will be apparent that the configuration of FIG. 4 also reduces the influence of spurious interference on the synchronization processing, as in the configuration of FIG.

(Third embodiment)
FIG. 5 shows a configuration example of a receiving apparatus according to the third embodiment of the present invention. In addition, about the component which has the same function as FIG. 1, the code | symbol same as FIG. 1 is attached | subjected, and description is abbreviate | omitted. The configuration of FIG. 5 is different from FIG. 1 in that a decimation unit 15 is inserted before the adaptive prediction difference filter. The decimation unit thins out the input signal sequence every other sample and halves the sampling frequency.

There are two main effects of decimation. The first is the effect of reducing circuit scale and power consumption. This is because the operation speed of the guard correlation calculation unit is reduced by half due to decimation, and the number of operations is reduced accordingly.

問題 Prior to explaining the second effect of the decimation part, we will clarify the problems when not decimating. Looking at the power spectrum of the complex baseband signal in the computer simulation described above, that is, FIG. 2A, the OFDM spectrum is distributed on a rectangle in a band of about 5.57 MHz centering on DC. On the other hand, outside the band, that is, outside the OFDM band, it can be seen that the power density is lower by about 20 dB than in the band. In this region, there is no effective signal and only a noise component is distributed. As described above, the adaptive prediction difference filter has the property of flattening the power spectrum of the output signal. As a result, in the output spectrum shown in FIG. 2C, it can be seen that the power density outside the band is relatively increased with respect to the inside of the band, and the power density ratio of both is reduced to about 15 dB. That is, the adaptive prediction difference filter relatively amplifies the noise component outside the band. As a result, the purity of the input signal to the guard correlation calculation unit may decrease, and the accuracy of symbol synchronization processing and narrowband AFC processing may decrease. In addition, since unnecessary bands are amplified, the spurious interference suppression performance that is essential may be lowered.

The above problem can be solved by performing decimation before the adaptive prediction difference filter. FIG. 6 shows a simulation result of the adaptive prediction difference filter in a state where the decimation unit is placed in front. An OFDM signal in which three spurious interference waves are superimposed is input to the decimation unit as in the above-described simulation. The power spectrum after decimation is shown in FIG. When compared with the power spectrum before decimation shown in FIG. 6A, it can be seen that the frequency domain aliasing caused by the decimation eliminates the low power density region and increases the flatness of the spectrum. As a result, the prediction difference filter can apply its spectrum flattening capability only to suppression of spurious interference without amplifying unnecessary bands.

In the above description, the thinning rate is set to 1/2. However, the same effect can be expected when the thinning rate other than this is set to 1/3 or 1/4.

As described above, by inserting the decimation unit into the input of the adaptive prediction difference filter, the amount of calculation of the adaptive prediction difference filter can be reduced, and spurious interference can be efficiently performed without amplifying noise components outside the OFDM band. It becomes possible to suppress well.

In addition to those already described above, the methods according to the above embodiments may be used in appropriate combination.

2 Time domain processing unit (equivalent to time domain processing means)
4 Guard correlation calculation part 5 Window position control part 6 Window extraction part (equivalent to window extraction means)
7 FFT processing unit (equivalent to FFT processing means)
8 Frequency controller (equivalent to frequency control means)
11 Adaptive prediction difference filter (equivalent to adaptive prediction difference filter means)
13 Moving averager 15 Decimation unit 21 Symbol synchronization unit (symbol synchronization means)
50 Frequency

Domain Processing Unit

100, 200 Receiver

Claims

Time domain processing means for converting the received OFDM signal into a complex baseband signal;
Window extraction means for extracting a window of a predetermined length from the complex baseband signal or a signal obtained by further processing the complex baseband signal;
FFT processing means for performing FFT processing on the window extracted by the window extraction means;
Adaptive prediction difference filter means for outputting a prediction residual signal for a signal obtained by thinning out the complex baseband signal at a thinning rate of 1 / N (N is a natural number);
Symbol synchronization means for controlling the extraction window position of the window extraction means based on the prediction residual signal;
A receiving apparatus comprising:
The receiving apparatus according to claim 1, wherein N of the thinning rate 1 / N is given as a natural number of 2 or more.
The adaptive prediction difference filter means is configured as a linear variable coefficient filter,
3. The receiving apparatus according to claim 1, wherein the coefficient value of the linear variable coefficient filter is updated by an LMS algorithm so that the power of the prediction residual signal is reduced.
The symbol synchronization means includes
The extraction window position of the window extraction means is controlled based on a complex correlation between the prediction residual signal and a signal obtained by delaying the prediction residual signal by an effective symbol length. The receiving device according to any one of claims.
The receiving device is:
5. The receiving apparatus according to claim 4, further comprising frequency control means for controlling the time domain processing means so that a modulation frequency error of the complex baseband signal becomes zero based on the complex correlation.
A time domain processing step of converting the received OFDM signal into a complex baseband signal;
A window extraction step for extracting a predetermined length section from the complex baseband signal or a signal obtained by further processing the complex baseband signal;
An FFT processing step for performing FFT processing on the section extracted by the window extraction means;
An adaptive prediction difference filter step of outputting a prediction residual signal for a signal obtained by thinning out the complex baseband signal at a thinning rate of N: 1 (N is a natural number);
A symbol synchronization step for controlling an extraction window position of the window extraction means based on the prediction residual signal;
A receiving method comprising: