JP2010127645A - Arbitrary signal generator - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To allow a signal having an arbitrary waveform up to a high frequency area to be generated by expanding a frequency band of the signal to be generated while using a sampling frequency of a relatively low frequency. <P>SOLUTION: A complex spectrum in a frequency area in which a target signal is subjected to discrete Fourier transform is divided into P sets of signal components. They are subjected to split processing by distributing them to P signal modules SG<SB>1</SB>, SG<SB>2</SB>, SG<SB>3</SB>to SG<SB>P</SB>operating at a low sampling clock, and the respective outputs are combined to generate an arbitrary signal. Each signal module subjects the distributed signal component to inverse discrete transform by an inverse discrete Fourier transformer 10, and the inverse discrete Fourier transform output is up-converted by a quadrature modulator 17, with respect to a high-frequency component having difficulties in digital signal processing. Accordingly, each of the signal modules SG<SB>2</SB>, SG<SB>3</SB>to SG<SB>P</SB>is operated with a low-speed clock inversely proportional to a parallel number P. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an arbitrary signal generator that generates a target signal having a wideband arbitrary waveform.

  Conventionally, as a technique for generating a signal having a small pulse width used in a pulse radar or the like, a technique for generating a target signal having an arbitrary waveform using a step recovery diode, an avalanche diode, a high-speed logic gate, or the like is known. These technologies generate a signal with a small pulse width by analog signal processing, and the performance may change due to changes in circuit characteristics due to environmental changes such as temperature and humidity, and changes in the characteristics of elements and materials over time. There is.

  Further, once an analog signal processing circuit is created, subsequent changes are difficult, and it cannot be said that the degree of freedom is necessarily high with respect to changes in specifications and purposes. For example, in the case of an application as a signal generator for radar, even if the assigned frequency of the radar is changed, it is difficult to cope with an analog circuit. Further, when changing the spectrum of the transmission wave, it is necessary to take measures such as changing an external filter. However, it is difficult for the analog circuit to adaptively change the shape and level of the transmission spectrum.

For this reason, in recent years, a technique for generating a target signal by digital signal processing is often employed. For example, a direct digital synthesizer (DDS) as disclosed in Patent Document 1 is known as a signal generator using this digital signal processing. DDS is a technique for generating an arbitrary waveform by updating a specified address in a memory containing waveform data and converting the data at the specified address into an analog waveform by D / A conversion.
Japanese Patent Application Laid-Open No. 11-2225022

  However, the signal generator using DDS as disclosed in Patent Document 1 performs A / D conversion after digitally synthesizing a signal waveform having a period that is a rational multiple of the reference clock in synchronization with the reference clock frequency. It is carried out. Therefore, according to the sampling theorem, the D / A converter must be operated at a sampling frequency that is twice or more the highest frequency included in the target signal. For this reason, if the pulse width of the target signal is reduced, the required sampling frequency increases in inverse proportion to the pulse width. Therefore, there is a limit to the frequency of the signal that can be generated.

  The present invention has been made in view of the above circumstances, and can expand the frequency band of a signal to be generated while using a relatively low sampling frequency, and can generate a signal having an arbitrary waveform up to a higher frequency range. An object is to provide an arbitrary signal generator.

  In order to achieve the above object, an arbitrary signal generator according to the present invention divides a frequency domain spectrum calculated from a time domain signal into a plurality of signal components and distributes the spectrum from the signal distributor. A signal converter that converts each of the plurality of signal components into a signal in a time domain, and a frequency converter that upconverts the frequency of at least one of the plurality of signals converted by the signal converter And a signal synthesizer that synthesizes a plurality of signals including the signal up-converted by the frequency converter and outputs a target signal.

  According to the present invention, the frequency band of a generated signal can be expanded while operating at a sampling frequency of a relatively low frequency, and a signal having an arbitrary waveform up to a higher frequency can be generated.

  Embodiments of the present invention will be described below with reference to the drawings. 1 to 12 relate to a first embodiment of the present invention, FIG. 1 is a configuration diagram of an arbitrary signal generator, FIG. 2 is an explanatory diagram showing a basic algorithm for generating an arbitrary signal by division processing, and FIG. FIG. 4 is an explanatory diagram showing the spectrum after inverse discrete Fourier transform in the low-frequency signal module, and FIG. 5 is the spectrum after interpolation in the low-frequency signal module. FIG. 6 is an explanatory diagram showing a spectrum after filtering by a complex bandpass filter in the low frequency signal module, FIG. 7 is an explanatory diagram showing a signal extracted by the low frequency signal module, and FIG. FIG. 9 is an explanatory diagram showing the spectrum after inverse discrete Fourier transform in the high-frequency signal module, and FIG. 9 shows the spectrum after interpolation in the high-frequency signal module. FIG. 10 is an explanatory diagram showing a spectrum after filtering by a complex bandpass filter in a high-frequency signal module, FIG. 11 is an explanatory diagram showing an output spectrum of the quadrature modulator, and FIG. 12 is an output of the mixer It is explanatory drawing which shows a spectrum.

The arbitrary signal generator 1 shown in FIG. 1 is supplied with data obtained by discretizing a target signal in an arbitrary frequency band to be generated in the time domain as an input signal xin. The input signal xin is digitally processed to generate an actual target signal xout. In the conventional digital signal processing, since the accuracy is high but the frequency that can be handled is relatively low, it is difficult to process an arbitrary broadband signal. In the present invention, an arbitrary target signal is divided into P sets of signals capable of digital signal processing, and parallel processing is performed. For the convenience of signal processing, the number of divisions is preferably a power of two. That is, N / P = 2 ^a (a is an arbitrary natural number) is preferable. However, it is possible to divide and process by an arbitrary number without being limited to a power of two.

The arbitrary signal generator 1 includes a serial-parallel converter 2 that converts an input signal xin into N parallel data, a discrete Fourier transform of the N parallel data, and divides it into P signal components in the frequency domain. data (complex spectrum) P-number of signal module SG ₁ for each frequency _{range, SG 2, SG 3, ...} , a discrete Fourier transformer 3 for distributing the SG _P, parallel complex spectrum dispensed from discrete Fourier transformer 3 in processes, P-number of signal module SG ₁ for generating signals of different frequencies from each _{other, SG 2, SG 3, ...} , SG P, the signal module _{SG 1, SG 2, SG 3} , ..., each output of the SG _P The mixer 4 is configured to synthesize and output a target signal xout having an arbitrary frequency band.

  In the present embodiment, the P signal modules generate signals having different frequencies from each other, in other words, do not handle signals in the overlapping frequency domain, but the present invention is not limited to this. Even if there are frequency regions that are handled in duplicate by a plurality of signal modules, there is no problem with the signal processing of the present invention by setting the target signal in consideration thereof.

  In addition, if there is a specific frequency region that you do not want to output in the frequency band of the target signal, do not handle that specific frequency region in any of the signal modules. A target signal not included can be obtained. For example, when there is a frequency region whose output is restricted due to laws and regulations, a target signal corresponding to the laws and regulations can be obtained by not handling the frequency region.

Each signal module _{SG 1, SG 2, SG 3} , ..., SG P is a module for a low frequency one module SG ₁ generates the lowest frequency of the signal, the other modules SG _2, SG _3, ... , SG _P is a high frequency signals than module SG _1, a module for high frequency generated in the different bands respectively. These signal module _{SG 1, SG 2, SG 3} , ..., SG P is the difference in the frequency band of the signal generated by each, and a different components in separate and common components shown below .

That is, each signal module _{SG 1, SG 2, SG 3} , ..., a common structure of the SG _P, inverse discrete Fourier transformer 10 to inverse discrete Fourier transform of the complex spectrum which is dispensed from discrete Fourier transformer 3, an inverse discrete A parallel-serial converter 11 that converts the parallel data subjected to Fourier transform into serial data (serial data), an interpolator 12 that multiplies the output data of the parallel-serial converter 11, and a spectrum in a desired frequency band is extracted from the multiplied signal. Complex bandpass filter (complex BPF) 13, IQ separator 14 that separates the output from complex BPF 13 into an I (In-Phase) signal and a Q (Quadrature) signal, an IQ-separated digital signal D / A converter 15 for converting the analog signal into an analog signal, and the D / A converted analog signal is filtered to produce spurious (unnecessary generation). ) Is a low pass filter (LPF) 16 for removing is provided with.

In addition, the following configurations are different for each signal module. Since the low frequency signal module SG ₁ takes out only the I signal (real number component) by the IQ separator 14 from the signal band-limited by the complex BPF 13, a set of D / A converters 15 corresponding to the I signal. And LPF 16. An analog signal from the LPF 16 is output to the mixer 4. On the other hand, the signal module SG ₂ for high-frequency, SG _3, ..., SG _P is the operating frequency band is different only in the same configuration, I signal (real component) from the IQ separator 14 and Q signals (complex component 2), two D / A converters 15a and 15b as the D / A converter 15 and two LPFs 16a and 16b as the LPF 16 are provided. Further, a quadrature modulator 17 that quadrature modulates the I signal from the LPF 16a and the Q signal from the LPF 16b is provided.

A signal output from the LPF16 signal module SG ₁ for low-frequency, signal module SG ₂ for high-frequency, SG _3, ..., the signal output from the quadrature modulator 17 of the SG _P is input to a mixer 4 Are combined and finally the target signal xout is output. The signal module SG ₂ for high-frequency, SG _3, ..., when the presence of a DC component in the output from the SG _P is concerned, signal module for high frequency SG _2, SG _3, ..., the SG _P A BPF or LPF may be added between each quadrature modulator 17 and the mixer 4 as necessary.

Next, the operation of the arbitrary signal generator 1 will be described. Schematically, any signal generator 1, the complex spectrum in the frequency domain, P-number of signal module SG ₁ operating at a low sampling clock, SG _2, SG _3, ..., and division processing by SG _P, each output To generate an arbitrary signal. High-frequency components that are difficult to handle with conventional digital signal processing are low-speed clocks that are inversely proportional to the division P, which is the number to be processed in parallel, by orthogonally modulating the inverse discrete Fourier transform output in each module with analog signal processing. each signal module SG _2, SG _3, ..., can be processed by operating the SG _P.

  Here, a basic algorithm for generating an arbitrary signal by division processing will be described. As schematically shown in FIG. 2, an analog signal that is temporally continuous is assumed as a target signal, the analog signal is converted into a spectrum in the frequency domain by a discrete Fourier transform, and the output spectrum is divided into a plurality of components. . In FIG. 2, the example divided into four components A, B, C, and D is shown. Of these four components, the A component having the lowest frequency is a frequency band in which digital signal processing is possible.

  The component A signal is subjected to inverse discrete Fourier transform and D / A conversion. The signal of component B is shifted to the frequency band of component A, subjected to inverse discrete Fourier transform, converted to an analog signal by D / A conversion, and further orthogonally modulated to shift the frequency band to the high frequency side adjacent to component A To do. Similarly, the component C signal is also shifted to the frequency band of component A, subjected to inverse discrete Fourier transform, converted to an analog signal by D / A conversion, and shifted to the high frequency side adjacent to component B by orthogonal modulation. To do. Similarly, the signal of component D is also shifted to the frequency band of component A, subjected to inverse discrete Fourier transform, converted to an analog signal by D / A conversion, and the frequency band is shifted to the high frequency side following component C by orthogonal modulation. . Finally, a signal similar to the target signal can be obtained by synthesizing the component A signal converted into an analog signal and the components B, C, and D converted into analog signals and shifted to the high frequency side. .

This is the signal module _{SG 1, SG 2, SG 3} , ..., the SG _P, with given by dividing the complex spectrum of the frequency domain of the target signal, processes the digital signal can be processed component in the signal module This means that a wideband arbitrary signal proportional to the parallel number P can be generated (synthesized) by orthogonally modulating high-frequency components that are difficult to process digitally by analog signal processing. When the frequency bandwidth that can be processed by one module is Bw, the frequency band of a signal that can be generated is Bw · P. That is, as long as there is a signal module whose frequency band that can be processed is Bw, a target signal of the frequency band Bw · P can be obtained by preparing a plurality of signal modules. Thereby, in the arbitrary signal generator 1 of this form, the arbitrary signal of a high frequency can be produced | generated, suppressing the sampling frequency of each module to a comparatively low frequency.

  In addition, the frequency band of a signal that can be generated can be increased or decreased by increasing or decreasing the division number P with respect to a predetermined frequency band Bw, and a highly scalable signal generator can be realized. That is, by dynamically changing the number of Ps according to the frequency band of the target signal, unnecessary signal module operations can be suppressed and power consumption can be reduced.

  For example, when the frequency bandwidth of the target signal is 1 GHz and the number of divisions is 4, the target signal can be obtained by handling the frequency band of 250 MHz in each signal module. In addition, even if the frequency bands of 1 GHz can be handled as a whole by the four signal modules without being divided equally, the signal modules may handle the unequal frequency bands. For example, it can be set to 100 MHz, 200 MHz, 300 MHz, or 400 MHz. However, for the convenience of signal processing, it is preferable to treat it equally. In the following, an example of equally distributing to each signal module is shown.

但し、ｎ，ｍ＝０，１，２，・・・，Ｎ−１ Specifically, a real continuous signal is x (t), and a signal obtained by discretizing the signal x (t) is x (n). If the result of discrete Fourier transform of this signal x (n) with the total number of points N is X (m), the relationship between x (n) and X (m) is expressed by the following equation (1).

However, n, m = 0, 1, 2,..., N−1

Next, the equation (1) is transformed into the following equation (2).

Here, since the signal X (m) is a discrete Fourier transform of the real signal x (n), X (1) = X ^* (N−k) holds, and the expression (2) is expressed by the following (3) It can be expressed by an expression. However, * shows a conjugate complex number and is k = 1, 2, ..., m.

In equation (3), when x (n) is a real number signal, X (0) and X (N / 2) are real numbers, and can be represented by the following equation (4). Furthermore, (5) It can be expanded like an expression. However, P represents the number of divisions, that is, the number of parallels, and takes a natural number.

Here, a condition for generating the signal X (n) using the equation (5) is considered. In order to simplify processing, this condition is that X (N / 2) is zero among the frequency components included in the signal x (n), and appropriate low-pass filter processing is performed on the signal x (n). It can be achieved if done. That is, the target signal x (n) is obtained when the signal represented by the equation (5) is processed at the sampling frequency fs, and the spectrum of the target signal x (n) is a spectrum as shown in FIG. , The data of the total number of points N from the series-parallel converter 2 is divided into P sets of discrete Fourier transforms of the number of points N / (2P) by the discrete Fourier transformer 3 and P signal modules SG _{1. , SG 2, SG 3, ...} , by parallel processing by distributing the SG _P, it is possible to obtain a target signal x (n).

  Specifically, the discrete Fourier transform in the discrete Fourier transformer 3 is a fast Fourier transform (FFT) that uses the symmetry of the discrete Fourier transform to reduce the amount of computation and perform transformation at high speed. The discrete Fourier transformer 3 is the second half of N / 2 of the signals {X (0), X (1), X (2), X (3),. The term is removed using the symmetry of Fourier transform, and the first half of the signal components {X (0) / 2, X (1), X (2), X (3),..., X (N / 2-1 )} Is divided into N sets of N / (2P) pieces.

Divided P sets of first signal component {X (0) / 2, X (1), ..., X (N / (2P) -1)} is distributed to the signal module SG ₁ for low frequencies. Next signal component {X (N / (2P) ), X (N / (2P) +1), ..., X (2N / (2P) -1)} is distributed to the signal module SG ₂ for high-frequency . Furthermore, the signal component {X (2N / (2P) ), X (2N / (2P) +1), ..., X (3N / (2P) -1)} is distributed to the signal module SG _3. Hereinafter, the signal components {X ((i-1) N / (2P)), X ((i-1) N / (2P) +1),..., X (iN / (2P) -1)} are high frequencies. It is distributed to the signal module SG _i of use. Thereafter, the signals are distributed in the same manner, and the last signal component {X ((P-1) N / (2P)), X ((P-1) N / (2P) +1), ..., X (PN / (2P) -1)} is distributed to the signal module SG _P.

That is, the N / 2P term component is distributed to each signal module as follows.

  In this case, a plurality of target signals are subjected to discrete Fourier transform in advance, their values are stored in a memory, and a complex spectrum is read from the memory to the inverse discrete Fourier transformer 10 of each signal module according to the type of the target signal. The target signal may be generated. As a result, it is not necessary to always operate the discrete Fourier transformer 3 to obtain the complex spectrum of the target signal each time, so that the circuit scale can be reduced and the power consumption can be reduced.

Processing of the signal module SG ₁ for low frequency (5) corresponding to the first term of expression. Signal module SG ₂ for high-frequency, SG _3, ..., each process in the SG _P corresponds to (5) below the second term of the equation. Hereinafter, the signal module _{SG 1, SG 2, SG 3} , ..., is described signal processing in SG _P.

[Signal processing in signal module SG ₁ for low frequency]
The sampling frequency of the signal module SG ₁ is fs / (2P), and the input components {X (0) / 2, X (1),..., X (N / (2P) -1)} to the module are inverse discrete Fourier transforms. An inverse discrete Fourier transform is performed by the converter 10. The output shown in the following equation (6) is obtained by the inverse discrete Fourier transform.

  The spectrum of the equation (6) when the sampling frequency is fs / (2P) is equivalent to the spectrum of the equation (5) in the frequency range of 0 to fs / (2P) −fs / N. As shown in FIG. 4, the spectrum of equation (6) is a repetitive signal having a frequency of fs / (2P).

  The output of the inverse discrete Fourier transform 10 is input to the parallel-serial converter 11, converted into serial data by N / (2P) terms, and subjected to up-sampling after being interpolated by the interpolator 12 more than twice. The In this embodiment, twice the interpolation is performed. That is, as shown in FIG. 5, the sampling frequency at the output of the interpolator 12 is fs / P which is twice the original frequency. Here, the interpolation is not limited to twice, and it is possible to perform the interpolation at twice or more as appropriate depending on the situation.

  Next, the output of the interpolator 12 is input to the complex BPF 13. Thereby, as shown in FIG. 6, only the component of frequency 0-fs / (2P) -fs / N is taken out from the spectrum of a signal. At this time, the sampling frequency required for Equation (6) of the inverse discrete Fourier transform is 1 / (2P) times the sampling frequency required for Equation (5). The frequency is inversely proportional to the parallel number P, and as a result, the sampling frequency can be greatly reduced.

  Thereafter, only the real component (I component) is extracted from the signal band-limited by the complex BPF 13 by the IQ separator 14 and converted into an analog signal by the D / A converter 15. Further, the output of the D / A converter 15 is passed through the LPF 16 to extract a spectrum of frequencies 0 to fs / (2P) −fs / N from which spurious has been removed, as shown in FIG.

The above is the signal processing in the low frequency signal module SG ₁ corresponding to the first term of the equation (5), and the high frequency signal modules SG ₂ and SG ₃ corresponding to the second term and the following of the equation (5). , ..., parallel runs and the signal processing in the SG _P.

Signal Module SG ₂ for high-frequency, SG _3, ..., the signal processing in the SG _P]
As described above, the signal module for high frequency SG _2, SG _3, ..., the signal processing in the SG _P are those corresponding to the following second term of equation (5). Each module _{SG 2, SG 3, ...,} SG P , because the operating frequency band is the same as the basic operation on different only in the following, primarily the signal components {X (N / (2P) ), X ( N / (2P) +1), ..., will be described as a representative in the signal module SG ₂ which performs processing for X (2N / (2P) -1 )}.

Input signal components {X (N / (2P)), X (N / (2P) +1),..., X (2N / (2P) -1)} to the signal module SG ₂ The signal is input to the converter 10 and subjected to inverse discrete Fourier transform. The inverse discrete Fourier transformer 10 has signal components {X (N / (2P)), X (N / (2P) +1),..., X (2N / (2P) -1 at a sampling frequency fs / (2P). )} Is subjected to inverse discrete Fourier transform, and the transformation result shown in the following equation (7) is obtained.

  When the signal processing of the equation (7) is performed at the sampling frequency fs / (2P), the spectrum is in the range of the frequency 0 to fs / (2P) −fs / N in the second term of the equation (5). Equivalent to the complex spectrum in parentheses {}. As shown in FIG. 8, in the other frequency regions, the spectrum of frequencies 0 to fs / (2P) −fs / N is repeated.

The output of the inverse discrete Fourier transform 10 is input to the interpolator 12 via the parallel-serial converter 11 and subjected to interpolation twice or more as shown in FIG. Further, in order to remove spurious from the output of the interpolator 12 and extract a complex spectrum of frequency components 0 to fs / (2P) -fs / N, the output signal from the interpolator 12 is input to the complex BPS 13. The Thereby, as shown in FIG. 10, only the component of frequency 0-fs / (2P) -fs / N is taken out. Here, the interpolation is not limited to twice, and it is possible to perform the interpolation at twice or more as appropriate depending on the situation. Moreover, there is no need to make the same multiple as the processing performed in the signal module SG ₁ . You may go with different multiples.

  Next, the complex BPS 13 output is multiplied by a complex sine wave EXP (j · π · fs · t / P), and the real part of the output signal is extracted, thereby obtaining the second term of equation (5). . Where t is a continuous time variable. Actually, in order to obtain the component in square brackets [] in the second term of the equation (5), the output of the complex BPS 13 is converted into a real part (I component) and an imaginary part (Q component) by the IQ separator 14. To separate. Then, each is converted into an analog signal by the D / A converters 15a and 15b.

  Then, in order to remove spurious from the analog signal, the output signals of the D / A converters 15a and 15b are input to the LPFs 16a and 16b, respectively, so that only the frequency components of frequencies 0 to fs / (2P) −fs / N are obtained. Limit bandwidth. The I component signal band-limited by one LPF 16 a is input to the in-phase side of the quadrature modulator 17, and the Q component signal band-limited by the other LPF 16 b is input to the quadrature input side of the quadrature modulator 17.

  The quadrature modulator 17 performs quadrature modulation on the I and Q signals from the LPFs 16 a and 16 b with a local oscillation frequency of fs / (2P), up-converts them, and outputs them to the mixer 4. The output spectrum of the quadrature modulator 17 is a spectrum as shown in FIG. 11, which is equivalent to the second term of the equation (5).

Signal module SG ₃ for other high frequency, ..., even in SG _P, similar signal processing is performed. That is, the signal module SG _3, the signal component {X (2N / (2P) ), X (2N / (2P) +1), ..., X (3N / (2P) -1)} are processed. In the same manner, the last signal component {X ((P-1) N / (2P)), X ((P-1) N / (2P) +1), ..., X (PN / (2P)] -1)} it is processed in the signal module SG _P. However, the local oscillator frequency of the quadrature modulator 17 in the signal module SG ₃ is (2fs) / (2P), and the local oscillator frequency of the quadrature modulator 17 in the last signal module SG _P is (P-1) · fs. / (2P).

More signal module SG ₁ for the low-frequency subjected to signal processing and signal module SG ₂ for high-frequency, SG _3, ..., the outputs of SG _P is synthesized in the mixer 4. Thereby, the target signal xout having the output spectrum shown in FIG. 12 is obtained. This target signal xout can be regarded as equivalent to a signal obtained by continuation of equation (5).

Above in the present embodiment as described above, a plurality of signal module SG ₁ operating at a low sampling frequency of the extent feasible the inverse discrete Fourier transform, SG _2, SG _3, ..., dividing treated with SG _P. Subsequently, the output of the inverse discrete Fourier transformer 10 operating at a low sampling frequency that can be realized is up-converted by the orthogonal modulator 17. Finally, each signal module output is synthesized by the mixer 4. Thereby, an arbitrary signal can be generated.

  In addition, a signal module is configured by combining high-precision digital signal processing and high-speed analog signal processing, and a plurality of signal modules are operated in parallel. As a result, it is possible to reduce performance changes due to environmental changes such as temperature and humidity, and performance changes due to secular changes in the characteristics of elements and materials. Furthermore, the bandwidth of the target signal can be increased.

  Moreover, since modulation is performed using inverse discrete Fourier transform, the amplitude and phase of the spectrum of each signal module can be directly controlled. Furthermore, the frequency and spectrum of the target signal can be set freely.

  In the present embodiment, the example in which the complex spectrum of the target signal is calculated by the discrete Fourier transformer 3 and the target signal is generated via the inverse discrete Fourier transformers 10, 10,. Here, since the discrete Fourier transform and the inverse discrete Fourier transform are a transform pair, it is also possible to obtain a complex spectrum using the inverse discrete Fourier transform and generate the target signal using the discrete Fourier transform.

  Next, a second embodiment of the present invention will be described. FIG. 13 is a block diagram of an arbitrary signal generator according to a second embodiment of the present invention.

The second mode enables generation of target signals and data transmission with one signal generator, and a plurality of signal modules SG ₁ , SG ₂ , SG ₃ ,. One signal module of _P is configured as a signal module for data transmission by an orthogonal frequency division multiplexing system. Hereinafter, a description will be given centering on portions that are different from the first embodiment.

FIG. 13 shows an example in which the signal module SG ₁ in the first mode is assigned as the signal module SGd for data transmission. Then, the inverse discrete Fourier transformer 10 is changed to a discrete Fourier transformer 10A. A complex or real data signal xd is input to the signal module SGd via the serial-parallel converter 20. The input signal xd is OFDM-modulated via the discrete Fourier transformer 10 </ b> A and output to the mixer 4.

Other signal module _{SG 2, SG 3, ...,} SG P is the same as in the first embodiment. The time domain signal xin is input to the discrete Fourier transformer 3 via the serial-parallel converter 2. Based on the input signal xin, a complex spectrum is calculated by the discrete Fourier transformer 3. The obtained complex spectrum is subjected to dispersion processing by the inverse discrete Fourier transformers 10, 10,... Of each module, and a signal having a predetermined frequency is generated and output to the mixer 4 as described in the first embodiment.

  Thereby, the output of a target signal and data transmission can be performed from the mixer 4. For example, when used for pulse radar, radar pulse generation and data transmission can be performed by one signal generator. Note that data transmission and generation of a target signal may be performed simultaneously. Alternatively, some or all of the subcarriers may be used alternately in a time division manner.

  When generating a target signal, as described above, a complex spectrum obtained by performing discrete Fourier transform on the target signal in the time domain is stored in a memory in advance, and is read out from the memory to the inverse discrete Fourier transformer when necessary. A signal may be generated.

In addition, as the signal module SGd for data transmission, it is possible to assign another signal module such as SG ₂ instead of the signal module SG ₁ . Thereby, the frequency which performs data transmission can be changed.

  In the second embodiment, as in the first embodiment, not only the frequency and spectrum of the target signal can be set freely, but also the signal generation and data transmission used for pulse radar and the like are realized by one device. Therefore, it is possible to provide a device with a wider application range while suppressing an increase in cost.

The block diagram of the arbitrary signal generator according to the first embodiment of the present invention. As above, an explanatory diagram showing a basic algorithm for generating an arbitrary signal by division processing Same as above, explanatory diagram showing signal spectrum in low frequency signal module Same as above, explanatory diagram showing spectrum after inverse discrete Fourier transform in low frequency signal module Same as above, explanatory diagram showing the spectrum after interpolation in the low-frequency signal module As above, an explanatory diagram showing the spectrum after filtering by the complex bandpass filter in the low frequency signal module As above, an explanatory diagram showing a signal extracted by the low frequency signal module Same as above, explanatory diagram showing spectrum after inverse discrete Fourier transform in high-frequency signal module Same as above, explanatory diagram showing the spectrum after interpolation in the high-frequency signal module As above, an explanatory diagram showing a spectrum after filtering by a complex bandpass filter in a high-frequency signal module Same as above, explanatory diagram showing output spectrum of quadrature modulator Same as above, explanatory diagram showing the output spectrum of the mixer The block diagram of the arbitrary signal generator concerning 2nd Embodiment of this invention.

Explanation of symbols

1 Arbitrary Signal Generator 3 Discrete Fourier Transformer 4 Mixer 10 Inverse Discrete Fourier Transformer 12 Interpolator 17 Quadrature Modulator SG ₁ ,..., SG _P Signal Module xin Input Signal xout Target Signal

Claims (12)

A signal distributor that divides and distributes a frequency domain spectrum calculated from a time domain signal into a plurality of signal components;
A signal converter that converts each of the plurality of signal components distributed from the signal distributor into a signal in the time domain;
A frequency converter that up-converts the frequency of at least one of the plurality of signals converted by the signal converter;
An arbitrary signal generator comprising: a signal synthesis unit that synthesizes a plurality of signals including the signal up-converted by the frequency conversion unit and outputs a target signal.   The arbitrary signal generator according to claim 1, wherein the processing by the signal conversion unit is performed within a predetermined frequency band narrower than a frequency band of the target signal.   3. The arbitrary signal generator according to claim 2, wherein the narrow predetermined frequency band is a value obtained by dividing the frequency bandwidth of the target signal by the number of divisions into the signal components by the signal distribution unit.   2. The arbitrary signal generator according to claim 1, wherein the processing by the signal conversion unit is digital signal processing, and the processing by the frequency conversion unit is analog signal processing.   2. The arbitrary signal generator according to claim 1, wherein the processing by the signal distributor is digital signal processing.   2. The arbitrary signal generator according to claim 1, wherein the processing by the signal synthesizer is analog signal processing.   2. The arbitrary signal generation according to claim 1, wherein the plurality of signals converted by the signal conversion unit are converted into a plurality of signals in a frequency band including regions overlapping each other through up-conversion by the frequency conversion unit. apparatus.   2. The arbitrary signal generator according to claim 1, wherein the plurality of signals converted by the signal conversion unit are converted into a plurality of signals in frequency bands that do not overlap each other through up-conversion by the frequency conversion unit.   9. The arbitrary signal generator according to claim 1, wherein data transmission is performed by modulating a part of a plurality of carriers constituting an output spectrum of the target signal with a data signal.   A process of performing data transmission by modulating part or all of a plurality of carriers constituting the output spectrum of the target signal with a data signal, and generating the target signal by part or all of the plurality of carriers The arbitrary signal generator according to any one of claims 1 to 9, wherein the processing is alternately performed in a time division manner.   The arbitrary signal generator according to claim 1, wherein the up-conversion is performed by acquiring a passband signal by orthogonal modulation.   The arbitrary signal generator according to any one of claims 1 to 11, wherein the target signal is a transmission signal of a pulse radar.

Images