JP2006292710A - Broadband real-time digital spectrum equipment - Google Patents

Abstract

To provide a broadband real-time digital spectrum apparatus with a reduced amount of calculation.
A wide-band signal under measurement is divided on the frequency axis by a plurality of band-pass filters, and a signal that has passed through each band-pass filter is sampled by a corresponding analog-to-digital converter and converted into a digital signal. To do. The analog-to-digital converter 120 performs undersampling as necessary, and the output of each analog-to-digital converter 120 is connected to a plurality of digital down converters 130 and further divided on the frequency axis, and the output of the digital down converter 130 Is temporarily stored in the memory 140, but directly input to the FFT (Fast Fourier Transform) arithmetic circuit 150, and the output spectrum of each FFT arithmetic circuit 150 is concatenated on the frequency axis and stored in the display device or data recording device. The
[Selection] Figure 1

Description

  The present invention relates to a broadband real-time digital spectrum apparatus.

  Conventionally, there is a spectrum analyzer as an apparatus for measuring a certain band with respect to frequency. The spectrum analyzer includes a sweep type and a heterodyne type for measuring a relatively wide band, and an FFT (Fast Fourier Transform) type for measuring a specific narrow band.

  In addition, with the advancement of wireless technology and social demands, there are wide (high frequency) radio wave bands, overcrowding, and diversification of use technologies. This trend will continue to be deeply expanded.

Currently, there are the following products in the spectrum analyzer of the FFT method (IF sampling / digital quadrature detection method FFT analyzer) (see Non-Patent Document 1), and has the following features.
1. The RF down-converter converts a high-frequency signal into an IF frequency signal (IF signal) whose band is limited.
2. The IF signal is converted into a digital signal by a high-speed A / D converter.
3. The digitized IF signal may be subjected to quadrature detection processing.
This may be realized by a dedicated chip called a digital down converter, or may be performed by a calculation process on a memory by a high-speed arithmetic unit such as a DSP. The signal is a complex signal having a real part and an imaginary part.
The signal processed in 4.3 is converted into a spectrum by FFT calculation.
Without performing the processing in 5.3, it is possible to obtain a spectrum by directly performing an FFT operation on the digitized IF signal (which is a real signal).

Currently, there are the following products in the spectrum analyzer of the FFT method (analog quadrature detection method FFT analyzer) (see Non-Patent Document 2), and has the following features.
1. The RF down-converter converts a high-frequency signal into an IF frequency signal (IF signal) whose band is limited.
2. Perform quadrature detection with an analog circuit.
A CW signal having the same frequency as the IF frequency and having a phase shift of 90 degrees is mixed to obtain a complex signal having a frequency component of each difference as a real part and an imaginary part.
3. The analog real part and imaginary part signals are respectively input to the A / D converter and digitized.
4). An FFT operation is performed on the digitized complex signal to obtain a spectrum.

  For analog signals, a means called a filter bank composed of a plurality of filters is known (see Non-Patent Document 3).

Nippon Tektronix Co., Ltd. (2-15-2 Konan, Minato-ku, Tokyo), RSA series, as of March 10, 2005, Internet <URL: http: // www. tektronix. co. jp / Products / Measurement_Prod / SA /> Yokogawa Electric Corporation (2-9-32 Nakamachi, Musashino City, Tokyo), VN7100, as of March 10, 2005, Internet <URL: http: // www. yokogawa. co. jp / tm / Bu / VN7100 / welcome. htm> Hitoshi Kiya, “Multi-rate signal processing”, Shosendo Publishing (48 Yaraimachi, Shinjuku-ku, Tokyo), October 6, 1995, p. 89-148

    However, the maximum bandwidth of the spectrum obtained by the conventional IF sampling / digital quadrature detection method described above cannot theoretically be more than ½ of the A / D sampling frequency. Actually, about 1/4 to 1/3 is normal due to the transient characteristic of IF BPF.

  Similarly, in the analog quadrature detection method, using two A / D converters is equivalent to doubling the sampling frequency, and the handleable bandwidth is doubled.

Regardless of which method is used, the bandwidth that can be handled is
1) Handleable bandwidth ≤ A / D conversion sampling frequency / 2
Or
2) Handleable bandwidth ≤ A / D conversion sampling frequency.

Further, if a higher-speed A / D converter is used, the handleable bandwidth can be increased, but the higher the speed, the more expensive the A / D.
In addition, the A / D speed is limited by available devices from time to time.

Also, the spectrum bandwidth (span) obtained by discrete Fourier transform such as FFT is determined by the sampling frequency of the A / D converter.
In order to handle signals of a wider band than this, there is no choice but to simultaneously operate a plurality of IF sampling / digital quadrature detection system and analog quadrature detection system.
However, in that case, it is difficult to ensure the frequency matching of the down converter and the synchronization of the phase and the synchronization of the signals.

Further, in a system that simultaneously operates a plurality of IF sampling / digital quadrature detection methods and analog quadrature detection methods, a high-frequency analog circuit portion such as an RF down-converter unit is expensive.

  Further, when the sampling frequency becomes high, the data size becomes enormous. When recording in real time on the HDD, the storage cost is reduced to each stage, but the response time and write time of the storage element for recording become a bottleneck, and continuous recording of signals becomes impossible.

In addition, in order to continuously obtain the spectrum of the input signal as “real time / FFT / bandwidth” continuously without interruption, the calculation speed of the calculator determines this. In general, the amount of computation of FFT is
Complexity 演算 (log ₂ N) × N
There is a relationship. N is the size of the FFT. (Bringham Fast Fourier Transform) If the sampling frequency is the same, the larger N is, the more spectrum data can be obtained on the frequency axis (frequency resolution is improved). However, when the sampling frequency is high, the calculation is not in time unless N is less than a certain value.
After all, real-time, FFT, and bandwidth are proportional to the speed of the calculator. However, this is independent of the sampling frequency of the A / D converter. (If you spend money and quantity, it will be as fast as you want)

Considering the relationship between the bandwidth and the amount of calculation, it can be seen that the sampling frequency must be increased in order to obtain a wide bandwidth FFT. To obtain the same resolution on the frequency axis, a large number of data are sampled per unit time.
Real-time / FFT / Bandwidth ∝ Calculation speed / Computation amount FFT computation amount = ((log ₂ N) × N)
The table below compares this in several levels of size.

As shown in Table 1, the smaller size has better calculation efficiency.
Also, the smaller the FFT size, the better the calculation efficiency.
In DSPs and high-speed processors, the hit rate of the internal cache greatly contributes to the calculation speed. Therefore, a smaller FFT size than the numerical values in the above table is advantageous.

The present invention has been made in view of such problems,
A wideband signal to be measured is divided by a plurality of bandpass filters on the frequency axis, and the signals that have passed through each bandpass filter are sampled by individual analog-to-digital converters corresponding to each, and converted into digital signals. If the output signal exceeds the Nyquist frequency, the analog-to-digital converter performs undersampling as necessary, and the output of each analog-to-digital converter is connected to multiple digital downconverters. The output of the digital down converter is temporarily stored in a memory or directly input to an FFT (Fast Fourier Transform) arithmetic circuit, and the output spectrum of each FFT arithmetic circuit is connected and arranged on the frequency axis. Broadband characterized by being stored in device or data recording device An object of the present invention is to provide a real-time digital spectrum apparatus.

  It is another object of the present invention to provide a broadband real-time digital spectrum apparatus characterized in that the connection between a plurality of analog-digital converters and a plurality of digital down-converters can be switched.

  The wideband real-time digital spectrum apparatus of the present invention divides a wideband signal under measurement by a plurality of bandpass filters on the frequency axis, and the individual analog-to-digital converters corresponding to the signals passing through each bandpass filter respectively. Sampling into a digital signal, and if the output signal of the bandpass filter exceeds its Nyquist frequency, the analog-to-digital converter performs undersampling as necessary.

  According to the wide-band real-time digital spectrum apparatus of the present invention, a wide-band signal under measurement is divided on the frequency axis by a plurality of band-pass filters, and individual analog-to-digital conversions corresponding to the signals passing through the respective band-pass filters If the output signal of the bandpass filter exceeds the Nyquist frequency, the analog-to-digital converter performs undersampling as necessary to limit the analog-to-digital converter to the Nyquist frequency. In addition, a broadband spectrum can be observed without sweeping the local oscillator.

  In the broadband real-time digital spectrum device of the present invention, the output of each analog-digital converter is connected to a plurality of digital down converters and further divided on the frequency axis, and the output of the digital down converter is temporarily stored in a memory. It is stored or directly input to an FFT (Fast Fourier Transform) arithmetic circuit, and the output spectrum of each FFT arithmetic circuit is connected and arranged on the frequency axis and stored in a display device or data recording device.

  According to the wideband real-time digital spectrum apparatus of the present invention, a wideband signal under measurement is divided by a plurality of bandpass filters on the frequency axis, and the output of each analog-digital converter is connected to a plurality of digital downconverters. Divided on the frequency axis, the output of the digital down converter is temporarily stored in the memory or directly input to the FFT (Fast Fourier Transform) arithmetic circuit, and the output spectrum of each FFT arithmetic circuit is linked on the frequency axis By organizing and storing in a display device or data recording device, digital signals converted by the analog-digital converter can be connected and organized and observed as a broadband spectrum.

  According to the broadband real-time digital spectrum apparatus of the present invention, when observing a part of a narrow band, the connection between a plurality of analog-digital converters and a plurality of digital down-converters can be switched.

  By making the connection of a plurality of analog-digital converters and a plurality of digital down-converters switchable, not only a wide band but also a part of a narrow band can be observed.

The present invention has the following effects.
According to the wideband real-time digital spectrum apparatus of the first aspect, it is possible to sample a wideband signal while maintaining simultaneity without depending on the sampling frequency of the analog-digital converter. As long as even an analog intermediate frequency can be widened, an unlimited bandwidth can be obtained by using a plurality of analog-digital converters.
Also, for frequency division, the analog down converter (combination of mixer, local oscillator and bandpass filter) is minimized (this is costly) and by using a digital down converter, compared to analog The total cost is low, no adjustment, chip making, and cost reduction by mass production.
In addition, by dividing the acquired signal on the frequency axis, it can be divided into signals with a relatively low sampling rate that are easy to handle in time series, and the hardware for storage and calculation can be easily configured (a high-speed signal system can be The division will be a large-scale, ultra-high-speed logic circuit). Further, the acquired signal is divided into a low-rate signal in a narrow band of a plurality of channels at the end. For this reason, it is possible to record on an inexpensive, large-capacity HDD or the like in real time.

  According to the broadband real-time digital spectrum apparatus of the present invention as claimed in claim 2, the acquired signal is divided into a plurality of narrow-band and low-speed signals at the end, but they are synchronized and phase continuity. Therefore, it is possible to restore the complex spectrum of each channel to the time axis of the input wideband signal by inverse Fourier.

The best mode for carrying out the broadband real-time digital spectrum apparatus of the present invention will be described below.
First, the best mode for carrying out the first invention according to the broadband real-time digital spectrum apparatus will be described.

  The 10 analog modules in FIG. 1 are composed of 60 amplifiers, 100 RF downconverters, and 70 switches, and signals from 50 input signals are routed through 105 wideband IF signals and 100 RF downconverters. This is a device having a configuration that can be switched by 70 switches.

  The signal processing module 20 in FIG. 1 is a device that receives signals output from 10 analog modules and has the following configuration.

110 band pass filters are composed of a plurality.
One 110 band-pass filter is connected to one 120 analog-digital device. Therefore, the number of 120 analog-digital devices is the same as the number of 110 band-pass filters.
A plurality of 130 digital down converters are connected to one 120 analog-digital devices via 190 crossbar switches. In FIG. 1, four 130 digital down converters are connected via 190 crossbar switches per 120 analog-digital devices.
The 140 memory temporarily stores data acquired by the 130 digital down converter.
150 FFT operation devices are devices that perform FFT operations. The 150 FFT arithmetic unit may be a DSP, but it is preferable to construct a circuit that can perform high-speed and real-time (input signals are continuously calculated without exception) using an FPGA. In the present invention, either a dedicated circuit such as a DSP or an FPGA may be used.
There are many existing IPs for FFT operation circuits.
130 digital down converters, 140 memories, and 150 FFT arithmetic units are connected and connected one by one.
Reference numeral 160 denotes a concatenated arithmetic device.
The above is the configuration of the 20 signal processing modules.

Further, the 30 data processing modules in FIG. 1 are composed of 170 display units and 180 data recording devices.
The display unit 170 may be a pancon screen or an image display device having a memory.
The 180 data recording device may be a commercially available HDD.

Next, the operation of FIG. 1 will be described.
The RF down converter 100 in FIG. 1 converts an arbitrary frequency into an intermediate frequency signal (hereinafter referred to as an IF signal) that can be analog-digital converted within a certain range.
Details will be described with reference to FIGS.
The output of 100 RF down converters is generally equipped with a band pass filter (BPF), and the output signal is band limited.
100 RF downconverters shall respond to a wide band. For example, a wide band such as 200 MHz that cannot be handled by the conventional IF sampling method may be used.
In the present invention, a plurality of 110 band pass filters are provided. Each passband is shown in FIG.
As shown in FIG. 5, the pass band of each 110 band pass filter covers the band of 100 RF down converters when connected.

The 120 analog-digital device groups each correspond to an individual 110 bandpass filter.
The 120 analog / digital devices sample the IF frequency higher than the Nyquist frequency by the undersampling method with some exceptions.
The sampling frequency of 120 analog-digital devices is not one, but two or more are assumed. Details will be described with reference to FIG.

There are a plurality of 130 digital down converters, and details are shown in FIG.
In the case of 130 digital down converters, many have four channels or even more channels implemented on a single chip. Each figure will be described with 4 channels.
By making a plurality of 130 digital down converters correspond to the 120 analog digital devices, the signals detected by the 120 analog digital devices can be divided into a plurality of bands and subjected to FFT processing. As a result, it becomes possible to reduce the individual FFT size and increase the calculation efficiency. Also, the configuration of the calculation circuit by hardware becomes easy.

The spectrum obtained by each of 150 FFT arithmetic units corresponds to the band of 120 analog-digital devices and 130 digital down converters. If this is connected on the frequency axis, a broadband spectrum can be obtained.
Reference numeral 160 denotes a connection calculation device that performs a calculation for joining the above-described frequency axes.
The broadband spectrum thus obtained is displayed as a broadband spectrum on 170 display units (PC screen or image display device having a memory).
If necessary, the above-mentioned broadband spectrum data is recorded in 180 data recording devices.

  190 crossbar switches are devices that arbitrarily switch the output of 120 analog digital devices and the connection of 130 digital down converters. It can be easily realized with existing logic circuit technology. Details are shown in FIGS. 8 and 9 and the description thereof.

Next, FIG. 2 will be described. FIG. 2 depicts the signal passed from the RF downconverter 100 in FIG. 1 to the 110 bandpass filter as a change in the frequency domain.
The input signal of the 210 RF down converter 210 is an input signal to the 100 RF down converter. Among these, within a range that can be set by the apparatus of FIG. 2, centering on an arbitrary 220 center frequency f ₀ , a portion of a certain bandwidth, that is, a portion determined by the characteristics of the band pass filter of 215 of 100 RF down converters It is moved in parallel on the frequency axis of the appropriately configured intermediate frequency f _d a. (Translated to 240 band-limited, A / D capable intermediate signal frequencies determined by the characteristics of the band pass filter of 215 100 RF downconverters).
The 110 bandpass filters finely divide the 240 band-limited A / D-capable intermediate signal frequencies, as shown by the thin solid and dashed trapezoids shown by the 100 RF downconverter bandpass filter signals. The passband of the 110 bandpass filter may take some overlay.

Next, FIG. 3 will be described. FIG. 3 illustrates the details of 100 RF downconverters and is an existing technology.
f _IN : Input frequency (frequency of 50 input signals)
f _Lo : Local oscillator frequency (frequency emitted by 320 local oscillators)
f _IF : intermediate frequency (intermediate frequency signal band-limited via a bandpass filter of 100 RF downconverters via 310 mixers)
f _IF = f _IN ± f _Lo (1)
The frequency of the 50 input signals is converted to an intermediate frequency according to equation (1).
The 100 RF downconverter has a bandpass filter of 330 100 RF downconverter, and the intermediate signal is band limited.

Next, FIG. 4 will be described. FIG. 4 illustrates the details of 130 digital down converters and is an existing technology.
410 is a digital signal data stream converted by 120 analog-digital conversion devices, and the input is the converted digital signal data stream.
The digital signal generation device 420 generates a digital signal.
The signal output from the digital signal generation device 420 of 421 generates a digital signal of cos (2πf _IF t).
The signal output from the digital signal generating device 420 of 422 generates a digital signal of sin (2πf _IF t).
The 430 mixer and the 440 mixer are the signals input by the digital signal data stream converted by the 410 analog / digital conversion device 410 and cos (2πf _IF t) output from the 420 digital signal generation device 421. And the digital signal of sin (2πf _IF t) output from the digital signal generating device 420 of 422 are respectively multiplied. As in FIG. 3, the multiplication result, that is, the output to the 450 low pass filter and the 460 low pass filter, is the two sums and differences of the frequencies of f _IF , cos (2πf _IF t) and sin (2πf _IF t). Become an ingredient. The difference is a so-called baseband signal with a frequency of zero.
The 450 low-pass filter and the 460 low-pass filter are low-pass filters that remove the sum component from the above signals and pass only the difference baseband signal.
Of the resulting frequencies, the product of input multiplied by cos (2πf _IF t) corresponds to the “I component” (signal output from the low-pass filter of 470 450), and similarly to sin (2πf _IF t). Is generally called a “Q component” (a signal output from a 450 low-pass filter of 480).
Hereinafter, the I component and the Q component are as described above.

Next, FIG. 6 will be described. FIG. 6 illustrates processing of a signal that has passed through a band-pass filter (hereinafter referred to as BPF).
In addition, FIG. 5 will be described together. FIG. 5 is an example showing the correspondence between sampling frequencies of 120 analog-digital devices and bands in which signal processing is possible.
The signal shown in FIG. 6 passes through the 110 band-pass filter group and the signal observed in the 120 analog-digital device group is a signal obtained by sampling any band, as shown in FIG. To the Nyquist frequency (fs / 2).
Further, a low-pass filter (hereinafter referred to as LPF) from 0 Hz such as a signal sampled through a low-pass filter from 0 MHz to 50 MHz in the 80 MHz band of 500 in FIG. 5 and a signal sampled through a low-pass filter from 0 MHz to 70 MHz in the 120 MHz band of 550 in FIG. ) Has an aliasing band near the Nyquist frequency.
If this signal is subjected to FFT calculation with the sampling frequency kept at fs, the spectrum of the above band can be obtained, but the calculation efficiency is poor as shown in Table 1. Also, the higher the fs, the greater the difficulty in designing and manufacturing the electrical circuit. This is especially true when using the latest high-speed analog-to-digital converters.

Further, if the signal processing band 610 in FIG. 6 is divided into a plurality of bands as in the state of FIG. 6B,
1) Dividing the sampling frequency can reduce circuit difficulties.
2) The number of FFT samples for obtaining the same resolution is reduced, and the calculation efficiency is good.
There are advantages such as.
In order to divide the signal band, an LSI called a digital down converter (DDC) has already been marketed. There is also an FPGA IP source. These DDCs are convenient because some DDCs have a plurality of channels mounted thereon.
Band division and sampling frequency decimation correspond.
Dividing number = decimating number In other words, if dividing into four, the sampling frequency can be reduced to ¼.
Further, as shown in FIG. 5, if signals obtained at different sampling frequencies are decimated to the common divisor sampling frequency, the sampling frequencies become the same as each other, which makes it very easy to handle.
For example, in FIG. 5, there are two sampling frequencies of 80 MHz and 120 MHz, and their greatest common divisor is 40 MHz.
Also, for example, four 110 band pass filters are provided, and each pass band is sampled from a lower frequency through a band pass filter of 50 MHz to 130 MHz in the 120 MHz band of 560 in FIG. The signal sampled through a bandpass filter of 90 MHz to 190 MHz in FIG. The signal sampled through the band pass filter from 50 MHz to 130 MHz in the 120 MHz band of 560 in FIG. 5 and the signal sampled through the band pass filter from 90 MHz to 190 MHz in the 120 MHz band of 570 is output sampling frequency if each decimated to 1/3 Becomes 40 MHz.
By synthesizing the spectrum obtained from the divided band signals in this way, a spectrum with a bandwidth of 140 MHz can be obtained.

Next, FIG. 7 will be described. FIG. 7 is a specific example of band division related to spectrum synthesis, and is a more detailed view of FIG.
The band of the output bandpass filter of 10 analog modules of 700 is assumed to have a flat output over 80 MHz from 50 MHz to 130 MHZ.
A pass band of three band pass filters, a pass band of 710 band pass filter, a pass band of 720 band pass filter, and a pass band of 730 band pass filter.
The passband of the 710 bandpass filter is connected to 120 analog-to-digital conversion devices with a sampling frequency of 80 MHz, the passband of 720 bandpass filter is connected to 120 analog-to-digital conversion devices with a sampling frequency of 100 MHz, and 730 The pass band of the band-pass filter is connected to 120 analog-digital conversion devices having a sampling frequency of 120 MHz.
The output of 120 analog-to-digital conversion devices corresponding to the pass band of 710 band-pass filter, the pass band of 720 band-pass filter, and the pass band of 730 band-pass filter are respectively input to 130 digital down converters. Is done.

The 130 digital down converters corresponding to the pass band of the 710 band pass filter have Decimation = 8, an output rate of 10 MHz, an output of 4 channels (hereinafter referred to as CH), and an output bandwidth of 1 CH of slightly over 5 MHz. The band of each output CH has a slight overlap.
The bandwidth of 4CH connection is 20 MHz, and covers 50 MHz to 70 MHz.
In addition, 130 digital down converters corresponding to the pass band of the 720 band pass filter have Decimation = 10, the output rate is 10 MHz, the output is 8 CH, the bandwidth of each output CH is 5 MHz, and 40 MHz from 70 MHz to 110 MHz. Cover.
In addition, 130 digital down converters corresponding to the pass band of the 730 band pass filter have Decimation = 12, the output rate is 10 MHz, the output is 4 CH, the bandwidth of each output CH is 5 MHz, and 40 MHz from 70 MHz to 110 MHz. In addition, the 130 digital down converter corresponding to the pass band of the 730 band-pass filter can perform processing higher than the band of the output rate corresponding to 730 of 734, but the output band pass of the 10 analog modules of 700 Since it exceeds the filter bandwidth, there is no point in handling higher frequencies.
Further, if the output rate of each of the 130 digital down converters is a common divisor of the sampling frequencies of the three 120 analog to digital conversion devices, the output rate of each of the 130 digital down converters can be made the same.
Each of the 110 band-pass filters and each of the 120 analog-digital conversion devices corresponds to each of the 130 digital down converters, and each of the 130 digital down converters has a plurality of output CHs.
In addition, each output of each of the 130 digital down converters is continuous with no gaps while having a slight crossover.
As a result, it is possible to obtain a spectrum obtained by dividing the flat band-pass bandwidth 50 MHz-130 MHz = 80 MHz on the frequency axis in the band of the output bandpass filter of the ten analog modules 700.
Furthermore, if a large number of 110 band-pass filters, 120 analog-digital conversion devices, and 130 digital down converters are connected, an unlimited wide-band spectrum can be obtained.

つまり、演算速度は周波数スパンに大きく依存する。分解能にも若干依存する。
したがって、演算は間に合わないが大容量メモリに格納したあとに、演算すれば擬似的リアルタイムにはなる。
そのために必要なメモリサイズは、
データ取得時間×ＤＤＣ出力チャンネル数×ＤＤＣ出力周波数
となり、表２の結果が得られる。Next, the calculation speed in the FFT arithmetic unit 150 in FIG. 1 will be described.
The maximum calculation speed in 150 FFT arithmetic units is determined by circuit design in FPGA. Although DSP is skillful in coding, it is not faster than a certain speed.
Therefore, the calculation speed required to obtain the real-time FFT is
Time required: FFT window function FFT window function: inversely proportional to resolution; k / Rbw, k is a proportional constant 2-3
FFT size: window function time x sampling frequency Measurement frequency span: proportional to sampling frequency; real number Fs / 2, complex number Fs
Then,
Complexity: FFT size x log (FFT size)
And
Calculation speed: Calculation amount / Time required
= FFT calculation amount / window function time length

In other words, the calculation speed greatly depends on the frequency span. It also depends somewhat on the resolution.
Therefore, the calculation is not in time, but if it is calculated after being stored in the large-capacity memory, it becomes pseudo real time.
The memory size required for that is
Data acquisition time x number of DDC output channels x DDC output frequency, and the results in Table 2 are obtained.

・Ｔｏｔａｌ Ａ／Ｄ Ｓａｍｐｌｉｎｇ Ｆｒｅｑ＞＝ＤＤＣ出力周波数＊ＣＨ数

Total A / D Sampling Freq> = DDC output frequency * number of channels

From the above, according to the first best mode for carrying out the present invention, a wide-band signal under measurement is divided by a plurality of bandpass filters on the frequency axis, and the signals that have passed through each bandpass filter are respectively If the output signal of the bandpass filter exceeds the Nyquist frequency, the analog-to-digital converter performs undersampling as necessary. It is not limited to the Nyquist frequency of the converter, and a wide spectrum can be observed without sweeping the local oscillator.
The output of each analog-digital converter is connected to a plurality of digital down converters and further divided on the frequency axis, and the output of the digital down converter is temporarily stored in a memory or directly FFT (Fast Fourier Transform) Since the output spectrum of each FFT arithmetic circuit is input to the arithmetic circuit and connected and arranged on the frequency axis and stored in the display device or data recording device, the digital signal converted by the analog-to-digital converter is connected and arranged. It can be observed as a simple spectrum.

  The present invention is not limited to the best mode for carrying out the above-described invention, and various other configurations can be adopted without departing from the gist of the present invention.

  Next, the best mode for carrying out the second invention according to the broadband real-time digital spectrum apparatus will be described.

8 and 9 will be described. 8 and 9 are switching examples of the analog-digital conversion device and the digital down converter.
When performing spectrum measurement, it is also necessary to measure a narrower band with higher frequency resolution. In such a case, it is convenient to provide the crossbar switch 190 in FIG. 1 and switch the combination of the analog-digital conversion device and the digital down converter.
Hereinafter, symbols and simplified notations used in FIGS. 8 and 9 are as follows.
The 120 analog-digital conversion devices are denoted as A / D1.
The analog-digital conversion device 121 is denoted as A / D2.
The analog-digital conversion device 122 is denoted as A / D3.
The analog / digital conversion device 123 is denoted as A / D4.
The 120 analog-digital conversion devices, 121 analog-digital conversion devices, 122 analog-digital conversion devices, and 123 analog-digital conversion devices are collectively referred to as A / D.
The 130 digital down converter is denoted as DDC1.
The digital down converter 131 is denoted as DDC2.
The digital down converter 132 is denoted as DDC3.
The digital down converter 133 is denoted as DDC4.
The 130 digital down converters, 131 digital down converters, 132 digital down converters, and 133 digital down converters are collectively referred to as A / D.
FIGS. 8 and 9 are examples of four analog-digital conversion devices, four digital down converters, and four channels of digital down converters (denoted as CH).
In FIG. 8, there is a one-to-one correspondence between A / D and DDC. This is the connection to obtain the widest band spectrum. In the example of FIG. 8, if each DDC has a 4CH down converter, the entire band can be divided into a maximum of 16 partial bands.
In FIG. 9, A / D2 and A / D3 are connected to DDC1, DDC2, DDC3, and DDC4, respectively. The frequency range of A / D1 and A / D2 cannot be processed naturally, but the A / D2 and A / D3 bands can be divided up to 16, and the band corresponding to A / D1 and A / D2 can be processed. It is.
Similarly, a high resolution spectrum of a wide band can be obtained by switching each FFT process by a combination of A / D and DDC switching. The number of DDC channels may be increased or decreased as necessary. However, complete real-time characteristics are lost.

  From the above, according to the second best mode for carrying out the present invention, when observing some narrow bands, the connection between a plurality of analog-digital converters and a plurality of digital down-converters is switched. Since it is possible, not only a wide band but also a part of a narrow band can be observed.

  The present invention is not limited to the best mode for carrying out the above-described invention, and various other configurations can be adopted without departing from the gist of the present invention.

It relates to the broadband real-time digital spectrum equipment industry.

It is a block diagram which shows embodiment of the invention concerning a broadband real-time digital spectrum apparatus. It is a frequency (RF) downconverter signal schematic diagram showing an embodiment of the invention according to a broadband real-time digital spectrum apparatus. It is a block diagram of RF down converter which shows embodiment of the invention concerning a broadband real-time digital spectrum apparatus. It is a block diagram of the digital down converter which shows embodiment of the invention concerning a broadband real-time digital spectrum apparatus. It is an example figure of the frequency distribution in A / D conversion which shows embodiment of the invention concerning a wideband real-time digital spectrum apparatus. It is an illustration figure of the process of the signal which passed the optimal band pass filter which shows embodiment of the invention concerning a wideband real-time digital spectrum apparatus. (A) Before band division (b) After band division It is a specific example figure about the synthesis | combination of the optimal spectrum which shows embodiment of the invention concerning a broadband real-time digital spectrum apparatus. (A) Primary division (b) Secondary division of 700 and 730 (c) Secondary division of 720 (d) Band division table of 710 (e) Band division table of 720 (f) Band division table of 730 It is an example of switching of A / D and DDC which shows embodiment of the invention concerning a broadband real-time digital spectrum apparatus. It is an example of switching of A / D and DDC which shows embodiment of the invention concerning a broadband real-time digital spectrum apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Analog module 20 Signal processing module 30 Data processing module 50 Input signal 60 Amplifier 70 Switch 100 RF down converter 105 Wideband IF signal 110 Band pass filter 120 Analog digital conversion device 121 Analog digital conversion device 122 Analog digital conversion device 123 Analog digital conversion Device 130 Digital downconverter 131 Digital downconverter 132 Digital downconverter 133 Digital downconverter 140 Memory 150 FFT processing unit 160 Connection processing unit 170 Display unit 180 Data recording unit 190 Crossbar switch 210 100 RF downconverter input signal 215 100 Special characteristics of RF downconverter band-pass filter 220 RF downconverter bandpass filter signal 240 at arbitrary frequency 230 100 Bandlimited A / D convertible intermediate frequency 310 Mixer 320 Local oscillator 330 100 RF downconverter bandpass filter 410 120 Analog to digital conversion Data stream 420 of digital signal converted by the device 420 Signal output from the digital signal generation device 421 420 Digital signal generation device 421 420 Signal output from the digital signal generation device 440 Mixer 450 Mixer 450 Low pass filter 460 Low pass filter 470 Signal 480 output from 450 low-pass filter Signal output from 450 low-pass filter 500 Signal output from 0 MHz to 50 MHz in 80 MHz band A signal sampled through a pass-pass filter 510 A signal sampled through a band-pass filter from 30 MHz to 90 MHz in the 80 MHz band 520 A signal sampled through a band-pass filter from 70 MHz to 130 MHz in the 80 MHz band 530 A band-pass filter from 110 MHz to 170 MHz in the 80 MHz band Signal 540 sampled through 150 MHz to 210 MHz bandpass filter in 80 MHz band Signal 550 Sampled through low pass filter from 0 MHz to 70 MHz in 120 MHz band 560 Sampled through bandpass filter from 50 MHz to 130 MHz in 120 MHz band Signal 570 120MHz band Signal 580 sampled through a bandpass filter from 90 MHz to 190 MHz 580 Signal sampled through a bandpass filter from 170 MHz to 250 MHz in the 120 MHz band 610 Output bandpass filter output 710 of the analog module in the signal processing band 700 10 Passed through the bandpass filter An output rate corresponding to the band 711 710 An output rate corresponding to the output rate 712 710 corresponding to the output rate 712 710 corresponding to the output rate 720 corresponding to the band 711 710 An output rate corresponding to the pass band 721 710 710 An output rate corresponding to the passband 721 720 Output rate corresponding to 723 720 output rate corresponding to 724 720 output rate corresponding to 720 720 output rate corresponding to 726 720 output rate 727 7 Output rate 730 corresponding to 20, output rate 730 corresponding to the passband 731 730 corresponding to the pass band 731 730 output rate corresponding to the output rate 733 corresponding to the output rate 733 corresponding to the output rate 733 corresponding to the output rate 734 corresponding to the output rate 734

Claims (2)

  A wideband signal to be measured is divided by a plurality of bandpass filters on the frequency axis, and the signals that have passed through each bandpass filter are sampled by individual analog-to-digital converters corresponding to each, and converted into digital signals. If the output signal exceeds the Nyquist frequency, the analog-to-digital converter performs undersampling as necessary, and the output of each analog-to-digital converter is connected to multiple digital downconverters. The output of the digital down converter is temporarily stored in a memory or directly input to an FFT (Fast Fourier Transform) arithmetic circuit, and the output spectrum of each FFT arithmetic circuit is connected and arranged on the frequency axis. Broadband characterized by being stored in device or data recording device Real-time digital spectrum apparatus.   2. The broadband real-time digital spectrum apparatus according to claim 1, wherein the connection between the plurality of analog-digital converters and the plurality of digital down-converters is switchable.

Images