JP2012165357A - Signal processing system and signal processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a signal processing system that increases a compression efficiency and decreases a processing load.SOLUTION: The signal processing system includes a controller that, in order to perform compression and decompression on a signal having a non-dense property in frequency domain or temporal domain, detects the presence or absence of the signal on the basis of: generation of parameters necessary for generating an averaged random measurement matrix, an averaged random restoration matrix, a non-averaged random measurement matrix, and a non-averaged random restoration matrix; and power of the averaged signal, and performs non-average interval calculation. The signal processing system averages a signal converted from analog to digital, compresses the signal by using an averaging random measurement matrix unit, and receives an input signal in a non-averaging interval. The signal processing system compresses the signal by using a non-averaging random measurement matrix unit, restores the compressed signal by using the averaged random restoration matrix, and notifies the controller of an output of a restoration result and the restoration result. The signal processing system restores the compressed signal by using the non-averaged random restoration matrix, and notifies the controller of an output of a restoration result and the restoration result.

Description

  The present invention relates to a signal processing system and a signal processing method for compressed sensing that AD-convert, average (or average after autocorrelation processing), compress, and decompress a signal.

  Conventionally, it is known that a network band can be effectively utilized by AD-converting, compressing, and transmitting a signal (data series). In this specification, a technique for AD-converting, compressing, and restoring a signal is referred to as a compressed sensing technique. In the conventional compressed sensing technology, the input data signal is multiplied by a random measurement matrix to generate compressed data. This random measurement matrix is composed of random variables, and the number of input data (signal) The number of rows is such that the number of output data (compressed data) is small.

  The concept of conventional compression sensing is shown in FIG. 30 (for example, see Non-Patent Documents 1 and 2). FIG. 30 assumes a sparse signal in the frequency domain or time domain. Since the signal is sparse in frequency or time, it is converted by a random measurement matrix composed of 40 × 100, and 40 compressed data are obtained from 100 input signals.

  FIG. 31 shows a system configuration for performing conventional compression sensing. In FIG. 31, an input signal is converted from an analog signal to a digital signal by the AD conversion unit 1002, and the conversion result is output to the compression sensing unit 1004. The random measurement matrix generation unit 1008 generates a random measurement matrix composed of random variables. The random restoration matrix generation unit 1010 generates a random restoration matrix by multiplying the random measurement matrix by the IFFT matrix. The compression sensing unit 1004 compresses the input digital signal using the random measurement matrix generated by the random measurement matrix generation unit 1008. The compression result is output to the signal restoration unit 1006. The signal restoration unit 1006 restores the input compressed data using a random measurement matrix and a restoration algorithm such as an L1-minimization restoration algorithm, and outputs the restored signal.

E. Candes et al., "An Introduction to Compressive Sampling," IEEE Signal Processing Magazine, pp. 21-30, March, 2008. E. Candes et al., "Near-Optimal Signal Recovery from RandomProjections: Universal Encoding Strategies?" IEEE Transaction on Information Theory, pp. 5406-5425, December, 2006.

  By the way, in the conventional compression sensing technique, since all the sampled samples are compressed and restored, there is a problem that the processing amount and the processing time increase.

  The present invention has been made in view of such circumstances, and an object of the present invention is to provide a signal processing system and a signal processing method that can improve the compression efficiency and reduce the processing load.

  The present invention is a signal processing system that compresses and decompresses a signal having a sparse property in the frequency domain or in the time domain, and converts an input analog signal into a digital signal and outputs the digital signal, Necessary for generating an averaging unit that averages and outputs the signal output from the AD conversion unit, an averaged random measurement matrix, an averaged random recovery matrix, a non-averaged random measurement matrix, and a non-averaged random recovery matrix A control unit that detects the presence or absence of a signal based on the generation of a parameter and the power of the averaged signal and calculates a non-average period; and the averaged random measurement matrix based on the parameter of the averaged random measurement matrix generation An averaged random measurement matrix generating unit that generates a signal, and an averaged compression cell that compresses a signal output from the averaged unit using the averaged random measurement matrix unit. A non-averaged random measurement matrix generating unit that generates the non-averaged random measurement matrix based on a parameter for generating the non-averaged random measurement matrix, and the digital signal output from the AD conversion unit. A non-averaged compression sensing unit that inputs a signal of an average interval and compresses the signal using the non-averaged random measurement matrix unit, and the averaged random recovery matrix based on the parameters of the averaged random recovery matrix generation An averaged random restoration matrix generation unit to be generated and a signal compressed by the averaged compression sensing unit are restored using the averaged random restoration matrix, and an output of the restoration result and a restoration result are notified to the control unit An averaged restoration unit and a non-averaged unit that generates the non-averaged random restoration matrix based on the parameters of the non-averaged random restoration matrix generation A non-average restoration matrix generating unit and a non-averaged signal that is restored by the non-averaged compressed sensing unit using the non-averaged random restoration matrix and notifies the control unit of the output of the restoration result and the restoration result And a reconstruction / restoration unit.

  The present invention is a signal processing system composed of a central station and a remote station for compressing and decompressing a signal having a sparse property in a frequency domain or a time domain, and the remote station receives an input analog signal An A / D conversion unit that converts the signal into a digital signal, an averaging unit that averages and outputs the signal output from the A / D conversion unit, and generates the averaged random measurement matrix based on the parameters for generating the averaged random measurement matrix Based on an averaged random measurement matrix generation unit, an averaged compression sensing unit that compresses a signal output from the average unit using the averaged random measurement matrix unit, and a parameter for generating the non-averaged random measurement matrix A non-averaged random measurement matrix generation unit for generating the non-averaged random measurement matrix; and a digital signal output from the AD conversion unit. A non-average compression sensing unit that inputs a signal of a non-average period and compresses the signal using the non-averaged random measurement matrix unit, and a communication unit that performs communication between the central station, The central station is based on the generation of averaged random measurement matrix, averaged random recovery matrix, non-averaged random measurement matrix, generation of parameters necessary for generation of non-averaged random recovery matrix and power of averaged signal A control unit that detects the presence or absence of a signal and calculates a non-average interval, an averaged random recovery matrix generation unit that generates the averaged random recovery matrix based on the parameters of the averaged random recovery matrix generation, and the average An average restoration unit that performs restoration using the averaged random restoration matrix, and notifies the control unit of the output of the restoration result and the restoration result; A non-averaged random restoration matrix generation unit that generates the non-averaged random restoration matrix based on the parameters of the non-averaged random restoration matrix generation; and a signal compressed by the non-averaged compressed sensing unit A non-averaging restoration unit that performs restoration using a restoration matrix and notifies the control unit of the output of the restoration result and the restoration result; and a communication unit for performing communication with the remote station. Features.

  The present invention is a signal processing system that compresses and decompresses a signal having a sparse property in the frequency domain or in the time domain, and converts an input analog signal into a digital signal and outputs the digital signal, An average unit that averages and outputs the signal output from the AD conversion unit, generation of parameters necessary for generation of an averaged random measurement matrix and an averaged random restoration matrix, and a signal based on the power of the averaged signal A control unit that detects the presence / absence of a non-compressed section, an output unit that outputs a signal of an uncompressed section among the digital signals output from the AD converter, and generation of the averaged random measurement matrix An averaged random measurement matrix generation unit that generates the averaged random measurement matrix based on a parameter, and the averaged random measurement matrix unit is used to output the averaged random measurement matrix An averaged compression sensing unit that compresses a signal, an averaged random restoration matrix generation unit that generates the averaged random restoration matrix based on the parameters of the averaged random restoration matrix generation, and the averaged compression sensing unit And an average restoration unit that performs restoration using the averaged random restoration matrix and notifies the control unit of the output of the restoration result and the restoration result.

  The present invention is a signal processing system composed of a central station and a remote station for compressing and decompressing a signal having a sparse property in a frequency domain or a time domain, and the remote station receives an input analog signal An A / D conversion unit that converts the signal into a digital signal, an averaging unit that averages and outputs the signal output from the A / D conversion unit, and generates the averaged random measurement matrix based on the parameters for generating the averaged random measurement matrix Among the digital signals output from the averaged random measurement matrix generation unit, the averaged compression sensing unit that compresses the signal output from the average unit using the averaged random measurement matrix unit, and the AD conversion unit An output unit that outputs a signal in an uncompressed section and a communication unit that communicates with the central station, the central station including an averaged random measurement matrix and A control unit that detects the presence / absence of a signal based on generation of parameters necessary for generation of an averaged random recovery matrix and the power of the averaged signal, and calculates an uncompressed section; and generation of the averaged random recovery matrix An averaged random restoration matrix generation unit that generates the averaged random restoration matrix based on a parameter, and restores the signal compressed by the averaged compression sensing unit using the averaged random restoration matrix, and outputs a restoration result And an averaging restoration unit for notifying the control unit of the restoration result, and a communication unit for communicating with the remote station.

  In the present invention, the averaging unit includes an autocorrelation processing unit that performs autocorrelation calculation processing and an average processing unit that performs average calculation processing.

  The present invention further includes an FFT absolute value calculation unit that outputs an absolute value calculation result signal after FFT conversion of the signal output from the AD conversion unit, and the averaging step is output from the FFT absolute value calculation unit The signal is averaged and output.

  The present invention provides an AD conversion unit, an average unit, a control unit, an averaged random measurement matrix generation unit, a non-conversion unit, a compression unit and a decompression unit for compressing and decompressing a signal having a sparse property in a frequency domain or a time domain Averaged random measurement matrix generation unit, averaged compressed sensing unit, non-averaged compressed sensing unit, averaged random restoration matrix generation unit, non-averaged random restoration matrix generation unit, averaged restoration unit, non- A signal processing method in a signal processing system including an averaging restoration unit, wherein the AD conversion unit converts an input analog signal into a digital signal and outputs the digital signal, and the average unit includes the AD conversion. Averaging step for averaging and outputting the signal output from the unit, and the control unit includes an averaged random measurement matrix, an averaged random restoration matrix, a non-averaged random measurement matrix, and a non-average A control step for detecting the presence or absence of a signal based on generation of parameters necessary for generation of a random restoration matrix and the power of the averaged signal and calculating a non-average period, and the averaged random measurement matrix generation unit, An averaged random measurement matrix generating step for generating the averaged random measurement matrix based on the averaged random measurement matrix generation parameter, and the averaged compressed sensing unit uses the averaged random measurement matrix unit to calculate the average unit. A non-averaged random measurement matrix generating unit for generating the non-averaged random measurement matrix based on the non-averaged random measurement matrix generation parameter; Whether the averaged random measurement matrix generation step and the non-averaged compressed sensing unit are the AD conversion step. A non-averaged compression sensing step of inputting a signal in a non-average period from the digital signal to be output, and compressing the signal using the non-averaged random measurement matrix unit, and the averaged random restoration matrix generation unit An averaged random recovery matrix generation step for generating the averaged random recovery matrix based on the averaged random recovery matrix generation parameter, and the averaged recovery unit outputs the signal compressed in the averaged compression sensing step. An averaged restoration unit that performs restoration using the averaged random restoration matrix and notifies the control step of a restoration result output and a restoration result; and the non-averaged random restoration matrix generation unit, the non-averaged random restoration A non-averaged random restoration matrix generation step for generating the non-averaged random restoration matrix based on matrix generation parameters; The averaging restoration unit restores the signal compressed in the non-average compression sensing step using the non-averaged random restoration matrix, and notifies the control unit of the output of the restoration result and the restoration result And a restoration step.

  The present invention provides an AD conversion unit, an average unit, a control unit, an averaged random measurement matrix generation unit, an average for compressing and decompressing a signal having a sparse property in the frequency domain or the time domain A signal processing method in a signal processing system comprising an averaged compression sensing unit, an averaged random restoration matrix generation unit, and an averaging restoration unit, wherein the AD conversion unit converts an input analog signal into a digital signal An AD conversion step to output; an averaging step in which the averaging unit averages and outputs the signal output from the AD conversion unit; and the control unit generates an averaged random measurement matrix and an averaged random restoration matrix A control step for detecting the presence or absence of a signal based on the generation of parameters necessary for averaging and the power of the averaged signal and calculating an uncompressed section; and the averaged random measurement A column generation unit generates an averaged random measurement matrix based on the averaged random measurement matrix generation parameter, and an averaged compressed sensing unit includes the averaged random measurement matrix unit. An averaged compression sensing step for compressing a signal output from the averaging unit, and the averaged random recovery matrix generation unit generates the averaged random recovery matrix based on parameters of the averaged random recovery matrix generation An averaged random restoration matrix generation step, and the averaging restoration unit restores the signal compressed in the averaged compression sensing step using the averaged random restoration matrix, and outputs a restoration result and a restoration result. And an averaging step for notifying the control unit.

  The present invention is characterized in that the averaging step includes an autocorrelation processing step for performing autocorrelation calculation processing and an average processing step for performing average calculation processing.

  In the present invention, the signal processing system further includes an FFT absolute value calculation unit, and the FFT absolute value calculation unit outputs an FFT-calculated absolute value calculation result signal of the signal output from the AD conversion unit. An absolute value calculating step, wherein the averaging step averages and outputs the signal output from the FFT absolute value calculating unit;

  According to the present invention, it is possible to maintain the quality of a detected signal while reducing the processing amount and processing time required for signal detection in compressed sensing. Thereby, the effect that the compression efficiency can be improved and the processing load can be reduced is obtained.

It is a block diagram which shows the structure of the 1st Embodiment of this invention. 2 is a flowchart showing operations for performing data averaging, data compression, and decompression in the signal processing system shown in FIG. 1. 2 is a flowchart illustrating an operation of the AD conversion unit 102 illustrated in FIG. 1. It is a flowchart which shows operation | movement of the average part 104 shown in FIG. 3 is a flowchart illustrating operations of an averaged compressed sensing unit 106 and an averaged random measurement matrix generation unit 110 illustrated in FIG. 1. 3 is a flowchart illustrating operations of an average restoration unit 108 and an average random restoration matrix generation unit 112 illustrated in FIG. 1. 2 is a flowchart showing operations of a non-averaged random measurement matrix generation unit 116 and a non-averaged random compressed sensing unit 120 shown in FIG. 3 is a flowchart showing operations of a non-averaged restoration unit 122 and an averaged random restoration matrix generation unit 118 shown in FIG. It is a flowchart which shows operation | movement of the control part 114 shown in FIG. It is a block diagram which shows the structure of a remote station and a central station. It is a block diagram which shows the structure of the 2nd Embodiment of this invention. It is explanatory drawing which shows the compression sensing method for an average. It is explanatory drawing which shows the compression sensing method for non-average. It is explanatory drawing which shows the example of the random decompression | restoration matrix production | generation for an average. It is explanatory drawing which shows the example of the random restoration | regeneration matrix for non-means. It is a block diagram which shows the structure of the 3rd Embodiment of this invention. 17 is a flowchart showing operations for performing data averaging, data compression, decompression, and uncompressed output in the signal processing system shown in FIG. 16. 17 is a flowchart illustrating an operation of the AD conversion unit 302 illustrated in FIG. 16. It is a flowchart which shows operation | movement of the control part 314 shown in FIG. It is a block diagram which shows the structure of the 4th Embodiment of this invention. It is a block diagram which shows the structure of the 5th Embodiment of this invention. It is a flowchart which shows the operation | movement which performs a data average, data compression, and decompression | restoration in the signal processing system shown in FIG. 22 is a flowchart showing an operation of an AD conversion unit 102 in the signal processing system shown in FIG. It is a block diagram which shows the structure of the 6th Embodiment of this invention. It is explanatory drawing which shows the averaged compression sensing method of this invention. It is a block diagram which shows the structure of the 7th Embodiment of this invention. 27 is a flowchart showing an operation of outputting data average, data compression, decompression, and uncompressed data in the signal processing system shown in FIG. 26. 27 is a flowchart showing an operation of outputting data average, data compression, decompression, and uncompressed data in the signal processing system shown in FIG. 26. It is a block diagram which shows the structure of the 8th Embodiment of this invention. It is explanatory drawing which shows the concept of the conventional compression sensing. It is a block diagram which shows the system configuration | structure which performs the conventional compression sensing.

Hereinafter, a signal processing system according to a first embodiment of the present invention will be described with reference to the drawings.
First, the averaged compressed sensing method of the present invention will be described with reference to FIG. The input data is compressed by averaging the signal for the sparse signal in the frequency domain or the time domain and then multiplying the averaged random measurement matrix. In FIG. 12, data obtained by averaging 10,000 data every 100 data is processed by an averaged random measurement matrix to be compressed into 40 data. The averaged random measurement matrix is generated from a random measurement matrix generated with random variables.

  Next, the non-averaged compressed sensing method of the present invention will be described with reference to FIG. The input data is compressed by multiplying the interval in which the presence of the processing signal of the average compression sensing is detected by a non-averaged random measurement matrix. In FIG. 13, only 300 pieces of data in a section in which a signal is detected among 10000 pieces of data are compressed to 120 pieces without being averaged. The non-averaged random measurement matrix is generated from a random measurement matrix generated with random variables.

<First Embodiment>
Next, the configuration of the signal processing system according to the first embodiment of the present invention will be described with reference to FIG. FIG. 1 is a block diagram showing the configuration of the embodiment. In FIG. 1, an AD conversion unit 102 receives a received signal, performs analog-digital conversion, and outputs digital data to an averaging unit 104. In addition, the signal of the non-averaged section received from the control unit 114 is output to the non-averaged compressed sensing unit 120. Note that, as will be described later, whether the averaged section or the non-averaged section is determined is determined by whether or not the absolute value of a certain frequency component exceeds a threshold value. That is, the presence or absence of a signal is determined based on the averaged power of the signal.

  The averaging unit 104 receives the “presence / absence” parameter of the autocorrelation calculation process output from the control unit 114, and when the autocorrelation calculation process is “none”, the averaging unit 104 converts the signal of the averaging interval to each averaged compressed sensing frame And output the average result. Here, the autocorrelation calculation process is performed when the signals to be averaged (sample values) have positive and negative values, and when they are averaged, they cancel each other and signals are detected. This is to avoid that it may be difficult. In other words, by performing the autocorrelation calculation process before the averaging process, it is easy to detect a certain section of the signal on the receiving side.

  On the other hand, if the autocorrelation detection process is not performed, the calculation amount is reduced and the circuit scale is reduced. When the autocorrelation calculation process is “present”, the signals in the average interval are averaged after performing the autocorrelation calculation process for each averaged compressed sensing frame, and the averaged result of the autocorrelation is output. The averaged random measurement matrix generation unit 110 receives parameters necessary for generating the averaged random measurement matrix from the control unit 114, and generates an averaged random measurement matrix. The averaged compression sensing unit 106 compresses the digital data using the averaged random measurement matrix generated at 110 for the digital signal input from the average unit 104, and outputs the compressed data.

  The averaged random restoration matrix generation unit 112 receives parameters necessary for generation of the averaged random restoration matrix from the control unit 114 and generates an averaged random restoration matrix. The averaging restoration unit 108 receives the compressed data output from the averaged compression sensing unit 106, performs signal restoration using the averaged random restoration matrix generated by the averaged random restoration matrix generation unit 112, and performs restoration. Output the data. At this time, a restoration algorithm such as L1-minimization operates. At the same time, the averaged restoration data is output to the control unit 114 as “known information”.

  The non-averaged random measurement matrix generation unit 116 receives parameters necessary for generating the non-averaged random measurement matrix from the control unit 114 and generates a non-averaged random measurement matrix. The non-average compression sensing unit 120 compresses the digital data using the non-average random measurement matrix generated at 116 for the non-average period digital signal input from the AD conversion unit 102, and outputs the compressed data. To do. The non-averaged random restoration matrix generation unit 118 receives parameters necessary for the non-averaged random restoration matrix generation unit from the control unit 114 and generates a non-averaged random restoration matrix. The non-average restoration unit 122 receives the compressed data output from the non-average compression sensing unit 120 as an input, restores the signal using the non-average random restoration matrix generated at 118, and outputs the restored data. Output. At this time, a restoration algorithm such as L1-minimization operates. In addition, the non-averaged restoration data is output to the control unit 114 as known information.

  Note that the averaged random measurement matrix and the non-averaged random measurement matrix may be essentially the same matrix. However, by making each system a separate system, for example, in the case of a sparse signal in the frequency domain or time domain, for example, a non-average digital signal when the signal is detected, It is also possible to perform compression processing by matrix calculation that performs weighting that increases in the frequency band or time band. At this time, as the averaged random measurement matrix, a matrix having no bias in the frequency domain and the time domain can be used.

  With reference to the autocorrelation calculation parameters output by the averaging unit 104 and the control unit 114, the averaged compressed sensing frame, and the average period, and when there is no autocorrelation, the data input from the AD conversion unit 102 is The data is divided for each average period, and the data in the average period is further divided for each averaged compressed sensing frame and averaged to output data for the compressed sensing frame. If there is autocorrelation, the data input from the AD conversion unit 102 is divided for each average period, and the data within the average period is further divided for each averaged compressed sensing frame before performing autocorrelation processing. The correlation processing results are averaged, and data for the compressed sensing frame is output.

  Next, operations for performing data averaging, data compression, and decompression in the signal processing system shown in FIG. 1 will be described with reference to FIG. FIG. 2 is a flowchart showing operations for performing data averaging, data compression, and decompression in the signal processing system shown in FIG. First, the AD conversion unit 102 converts the input analog signal into a digital signal, and outputs all converted data to the averaging unit 104. At the same time, the data in the non-average section is output to the non-average compressed sensing 120 (step S1). When there is no autocorrelation, the averaging unit 104 divides the data input from the AD conversion unit 102 for each average period, and further averages the data within the average period for each averaged compressed sensing frame, Output data for compressed sensing frames. If there is autocorrelation, the data input from the AD conversion unit 102 is divided for each average period, and the data within the average period is further divided for each averaged compressed sensing frame before performing autocorrelation processing. The correlation processing results are averaged, and data for the compressed sensing frame is output (step S2).

  The averaged compression sensing unit 106 compresses the digital signal output from the average unit 104 using the averaged random measurement matrix (step S3). The averaging restoration unit 108 performs restoration processing on the data output from the averaged compressed sensing unit 106 using the averaged random restoration matrix (step S4), and outputs the restored signal (step S5).

  On the other hand, the non-average compression sensing unit 120 compresses the non-average interval data output from the AD conversion unit using the non-average random measurement matrix (step S6). The non-average restoration unit 122 performs a restoration process on the data output from the non-average compression sensing unit 120 using a non-average random restoration matrix (step S7), and outputs a restored signal (step S8).

  Next, the operation of the AD conversion unit 102 in the signal processing system shown in FIG. 1 will be described with reference to FIG. FIG. 3 is a flowchart showing the operation of the AD conversion unit 102 in the signal processing system shown in FIG. First, the input analog signal is sampled into a digital signal (step S11). All the sampled data is input to the averaged compression sensing unit 106 (step S12). And the presence or absence of the non-average compression process input from the control part 114 is referred (step S13), and when there exists, the data of a non-average area are input into the non-average compression sensing part 120 (step S14).

  Next, the operation of the averaging unit 104 in the signal processing system shown in FIG. 1 will be described with reference to FIG. FIG. 4 is a flowchart showing the operation of the averaging unit in the signal processing system shown in FIG. First, the presence / absence of autocorrelation processing is set with reference to the parameter of presence / absence of autocorrelation processing input from the control unit 114 (step S21). If there is autocorrelation processing, the autocorrelation of the data in the averaged compressed sensing frame is calculated for each averaged compressed sensing frame and the result is output (step S22). Then, the result of the autocorrelation process in step S22 is averaged for each averaged compressed sensing frame (step S23). If there is no autocorrelation, the data input from the AD converter 102 is averaged for each averaged compressed sensing frame (step S23). Then, the result is output to the averaged compression sensing unit 106 (step S24).

  Next, operations of the averaged compressed sensing unit 106 and the averaged random measurement matrix generation unit 110 in the signal processing system shown in FIG. 1 will be described with reference to FIG. The averaged random measurement matrix generation unit 110 refers to the number of rows and the number of columns of the averaged random measurement matrix, which is an averaged random measurement matrix generation parameter input from the control unit 114, and A random matrix having the size of the number of rows and the number of columns is generated (step S31). Each component of the averaged random measurement matrix is generated by a random number generator. The number of columns of the averaged random measurement matrix is the same value as the averaged compressed sensing frame. The averaged compressed sensing unit 106 divides input data for each averaged compressed sensing frame and multiplies the averaged random measurement matrix generated by the averaged random measurement matrix generation unit 110 (step S32). Data corresponding to the number of rows of the averaged random measurement matrix obtained as a result of multiplication for each averaged compressed sensing frame is output (step S33).

  Next, operations of the averaging restoration unit 108 and the averaged random restoration matrix generation unit 112 in the signal processing system shown in FIG. 1 will be described with reference to FIG. The averaged random restoration matrix generation unit 112 multiplies the averaged random measurement matrix generated by the averaged random measurement matrix generation unit 110 by an IFFT matrix having the number of columns of the averaged random measurement as the size of rows and columns, and averages them. A random restoration matrix is generated (step S41). FIG. 14 shows an example of generation of an averaged random restoration matrix when the number of rows of the averaged random measurement matrix is 40 and the number of columns is 100. The averaging restoration unit 108 performs signal restoration processing using a restoration algorithm such as L1-minimization using the averaged random restoration matrix generated by the averaged random restoration matrix generation unit 112 (step S42). Then, the restored signal is output to the control unit 114 and the outside (step S43).

  Next, operations of the non-averaged random measurement matrix generation unit 116 and the non-averaged random compressed sensing unit 120 in the signal processing system shown in FIG. 1 will be described with reference to FIG. The non-averaged random measurement matrix generation unit 116 refers to the number of rows and the number of columns of the non-averaged random measurement matrix, which is the non-averaged random measurement matrix generation parameter input from the control unit 114, and determines the non-averaged random measurement matrix A random matrix having the size of the number of rows and the number of columns of the non-averaged random measurement matrix is generated (step S61). Each component of the non-averaged random measurement matrix is generated by a random number generator. The number of columns of the non-averaged random measurement matrix is the same value as the non-averaged compressed sensing frame. The non-average compressed sensing unit 120 divides the input data for each non-average compressed sensing frame and multiplies it by the non-averaged random measurement matrix generated by the non-averaged random measurement matrix generation unit 116 (step S62). Data corresponding to the number of rows of the non-averaged random measurement matrix obtained as a result of multiplication for each non-averaged compressed sensing frame is output (step S63).

  Next, with reference to FIG. 8, the operations of the non-average restoration unit 122 and the averaged random restoration matrix generation unit 118 in the signal processing system shown in FIG. 1 will be described. The non-averaged random restoration matrix generation unit 118 multiplies the non-averaged random measurement matrix generated by the non-averaged random measurement matrix generation unit 116 by an IFFT matrix having the number of columns of the non-averaged random measurement as the size of rows and columns. Thus, a non-averaged random restoration matrix is generated (step S71). FIG. 15 shows an example of non-averaged random restoration matrix generation when the number of rows of the non-averaged random measurement matrix is 40 and the number of columns is 100. The non-average restoration unit 122 performs signal restoration processing using a restoration algorithm such as L1-minimization using the non-average random restoration matrix generated by the non-average random restoration matrix generation unit 118 (step S72). Then, the restored signal is output to the control unit 114 and the outside (step S73).

  Next, the operation of the control unit 114 in the signal processing system shown in FIG. 1 will be described with reference to FIG. The control unit 114 outputs the average interval and the size of the averaged compressed sensing frame to the average unit 104 (step S91). In addition, the control unit 114 outputs the number of rows and columns of the averaged random measurement matrix and the size of the averaged compressed sensing frame to the averaged random measurement matrix generation unit 110 and the averaged random restoration matrix generation unit 112 (Step S92). ). Then, the absolute value of the frequency component of the signal input from the averaging restoration unit 108 is analyzed for each average interval (step S93), and the presence / absence of a non-average processing target signal is determined (step S94). The presence / absence of this signal is determined by whether or not the absolute value of the frequency component exceeds a preset threshold value.

  If there is a non-average processing target signal, information on the average interval in which the signal is detected is output to the AD conversion unit 102 (step S95). Subsequently, the number of rows and columns of the non-averaged random measurement matrix and the size of the non-averaged compressed sensing frame are output to the non-averaged random measurement matrix generation unit 116 and the non-averaged random restoration matrix generation unit 118 (step). S96). Then, the absolute value of the frequency component of the signal input from the non-averaging restoration unit 122 is analyzed (step S97), and the presence / absence of a non-average processing target signal is determined (step S94). The presence / absence of this signal is determined by whether or not the absolute value of the frequency component exceeds a preset threshold value.

<Second Embodiment>
Next, the configuration of the signal processing system according to the second embodiment of the present invention will be described. FIG. 10 illustrates a configuration of a communication system that includes one or more remote stations 1 to N and a central station 10, and the remote stations 1 to N and the central station 10 are connected via a transmission path. In the second embodiment, an example in which the signal processing system of the present invention is applied to the communication system shown in FIG. 10 will be described.

  FIG. 11 is a block diagram showing configurations of the remote station 1 and the central station 10 shown in FIG. The AD conversion unit 202 receives the received signal, performs analog-digital conversion, and outputs digital data to the averaging unit 204. In addition, the non-average period signal received from the control unit 114 via the transmission path is output to the non-averaged compressed sensing unit 220. The averaging unit 204 refers to the parameter of the presence / absence of the autocorrelation calculation process received from the control unit 214 via the transmission path. If the autocorrelation calculation process is not performed, the averaging unit 204 calculates the average interval signal for each averaged compressed sensing frame. Average and output the average result. If there is autocorrelation calculation processing, the signals in the average interval are averaged after performing autocorrelation calculation processing for each averaged compressed sensing frame, and the averaged autocorrelation result is output.

  The averaged random measurement matrix generation unit 210 refers to parameters necessary for the averaged random measurement matrix generation unit received from the control unit 214 via the transmission path, and generates an averaged random measurement matrix. The averaged compression sensing unit 206 compresses the digital data using the averaged random measurement matrix generated in 210 for the digital signal input from the average unit 204, and outputs the compressed data. The averaged random recovery matrix generation unit 212 refers to parameters necessary for the averaged random recovery matrix generation unit received from the control unit 214 via the transmission path, and generates an averaged random recovery matrix. The output of the averaged compressed sensing unit 206 is transmitted to the central station 10 via the transmission path.

  The averaging restoration unit 208 receives the compressed data received from the averaged compression sensing unit 206 via the transmission path, and uses the averaged random restoration matrix generated by the averaged random restoration matrix generation unit 212 to generate a signal. Restore and output the restored data. At this time, a restoration algorithm such as L1-minimization operates. At the same time, the averaged restoration data is output to the control unit 214 as known information. The non-averaged random measurement matrix generation unit 216 refers to parameters necessary for the non-averaged random measurement matrix generation unit received from the control unit 214 via the transmission path, and generates a non-averaged random measurement matrix.

  The non-average compression sensing unit 220 compresses the digital data using the non-average random measurement matrix generated in 216 for the non-average period digital signal input from the AD conversion unit 202, and outputs the compressed data. To do. The output of the non-averaged compressed sensing unit 220 is transmitted to the central station 10 through the transmission path. The non-averaged random restoration matrix generation unit 218 receives parameters necessary for the non-averaged random restoration matrix generation unit from the control unit 214 and generates a non-averaged random restoration matrix. The non-average restoration unit 222 receives the compressed data received from the non-average compression sensing unit 220 via the transmission path, and uses the non-average random restoration matrix generated by the non-average random restoration matrix generation unit 218. Then, the signal is restored and the restored data is output. At this time, a restoration algorithm such as L1-minimization operates. In addition, the non-averaged restoration data is output to the control unit 214 as known information.

  AD conversion unit 202, averaging unit 204, averaged compressed sensing unit 206, averaging restoration unit 208, averaged random measurement matrix generation unit 210, averaged random restoration matrix generation unit 212, control unit 214, in the second embodiment The processing operations of the non-averaged random measurement matrix generation unit 216, the non-averaged random restoration matrix generation unit 218, the non-averaged compressed sensing unit 220, and the non-averaged restoration unit 222 are respectively the AD conversion unit 102 of the first embodiment. , Averaging unit 104, averaged compressed sensing unit 106, averaging restoration unit 108, averaged random measurement matrix generation unit 110, averaged random restoration matrix generation unit 112, control unit 114, non-averaged random measurement matrix generation unit 116, The same as the non-averaged random restoration matrix generation unit 118, the non-averaged compressed sensing unit 120, and the non-averaged restoration unit 122 So here is a detailed description thereof will be omitted.

<Third Embodiment>
Next, the configuration of a signal processing system according to the third embodiment of the present invention will be described with reference to FIG. FIG. 16 is a block diagram showing the configuration of the embodiment. In FIG. 16, the AD conversion unit 302 receives the received signal, performs analog-digital conversion, and outputs digital data to the averaging unit 304. In addition, the signal of the uncompressed section received from the control unit 314 is output to the outside as it is. Note that whether the compression period or the non-compression period is set is determined by whether or not the absolute value of a certain frequency component exceeds a threshold value. That is, the presence or absence of a signal is determined based on the averaged power of the signal.

  The average unit 304, the average compressed sensing unit 306, the average restoration unit 308, the averaged random compression measurement matrix generation unit 310, and the averaged random restoration matrix generation unit 312 are respectively the average unit 104 of the first embodiment. The average compressed sensing unit 106, the average restoration unit 108, the averaged random compression measurement matrix generation unit 110, and the averaged random restoration matrix generation unit 112 perform the same operations.

  Next, with reference to FIG. 17, the operation of performing data averaging, data compression, decompression, and output of uncompressed data in the signal processing system shown in FIG. 16 will be described. FIG. 17 is a flowchart showing operations for outputting data average, data compression, decompression, and uncompressed data in the signal processing system shown in FIG. First, the AD conversion unit 302 converts the input analog signal into a digital signal, and outputs all the converted data to the averaging unit 304 (step S3011). At the same time, the data in the uncompressed section is output as it is (step S308). Step 302, Step 303, Step 304, and Step 305 perform the same operations as Step S2, Step S3, Step S4, and Step S5 of the first embodiment shown in FIG.

  Next, the operation of the AD conversion unit 302 in the signal processing system shown in FIG. 17 will be described with reference to FIG. FIG. 18 is a flowchart showing the operation of the AD conversion unit 302 in the signal processing system shown in FIG. First, the input analog signal is sampled into a digital signal (step S3011). All the sampled data is input to the averaged compression sensing unit 106 (step S3012). Then, the presence / absence of uncompressed processing input from the control unit 314 is referred to (step S3013), and if present, the data of the uncompressed section is output to the outside (step S3014).

  Next, the operation of the control unit 314 in the signal processing system shown in FIG. 17 will be described with reference to FIG. The control unit 314 outputs the average interval and the size of the averaged compressed sensing frame to the average unit 304 (step S3091). Also, the control unit 314 outputs the number of rows and columns of the averaged random measurement matrix and the size of the averaged compressed sensing frame to the averaged random measurement matrix generation unit 310 and the averaged random restoration matrix generation unit 312 (step S3092). ). Then, the absolute value of the frequency component of the signal input from the averaging restoration unit 308 is analyzed for each average interval (step S3093), and the presence / absence of a signal to be uncompressed is determined (step S3094). The presence / absence of this signal is determined by whether or not the absolute value of the frequency component exceeds a preset threshold value. If there is a signal to be uncompressed, information on a section in which the signal is detected is output to the AD conversion unit 302 (step S3095).

<Fourth Embodiment>
Next, the configuration of the signal processing system according to the fourth embodiment of the present invention will be described. The overall configuration of the signal processing system according to the fourth embodiment is the same as the configuration shown in FIG. 10, and includes one or more remote stations 1 to N and the central station 10, and the remote stations 1 to N and the central station 10. Are connected by a transmission line. FIG. 20 is a block diagram showing the configuration of the remote station 1 and the central station 10 in the fourth embodiment. The AD conversion unit 402 receives the received signal, performs analog-digital conversion, and outputs digital data to the averaging unit 404. At the same time, the signal in the uncompressed section received from the control unit 414 via the transmission path is transmitted as it is to the central station 10 via the transmission path.

  The averaging unit 404, the averaged compressed sensing unit 406, the averaged random measurement matrix generation unit 410, the average restoration unit 408, and the averaged random restoration matrix generation unit 412 in the fourth embodiment are respectively shown in FIG. The same operation as the averaging unit 204, the averaged compressed sensing unit 206, the averaged random measurement matrix generation unit 210, the average restoration unit 208, and the averaged random restoration matrix generation unit 212 shown in FIG.

  Next, the operation of the control unit 414 in the signal processing system shown in FIG. 20 will be described. The control unit 414 outputs the average interval and the size of the averaged compressed sensing frame to the average unit 404 via the transmission path. In addition, the control unit 414 outputs the number of rows and columns of the averaged random measurement matrix and the size of the averaged compressed sensing frame to the averaged random measurement matrix generation unit 410 via a transmission path. In addition, the control unit 414 outputs the number of rows and columns of the averaged random restoration matrix and the size of the averaged compressed sensing frame to the averaged random restoration matrix generation unit 412. Then, the absolute value of the frequency component of the signal input from the averaging restoration unit 408 is analyzed for each average section, the presence / absence of a signal to be uncompressed is determined, and the uncompressed process is determined. Further, the information of the non-compression processing section is output to the AD conversion unit 402 via the transmission path. The presence / absence of this signal is determined by whether or not the absolute value of the frequency component exceeds a preset threshold value.

<Fifth Embodiment>
Next, a signal processing system according to a fifth embodiment of the present invention will be described. First, the averaged compressed sensing method of the present invention will be described with reference to FIG. Performs FFT (Fast Fourier Transform) conversion on signals that are sparse in the frequency domain or time domain, calculates the absolute value of the result, averages the calculation result, and then performs an averaged random measurement The input data is compressed by multiplying the matrix. In FIG. 25, data obtained by averaging 10,000 data every 100 data is processed by an averaged random measurement matrix to be compressed into 40 data. The averaged random measurement matrix is generated from a random measurement matrix generated with random variables. By performing FFT conversion on the signal before performing the averaging process, a compression effect by averaging can be obtained even for a signal sparse in the frequency domain.

  Next, the configuration of a signal processing system according to the fifth embodiment will be described with reference to FIG. FIG. 21 is a block diagram showing the configuration of the embodiment. In this figure, the same parts as those of the system of the other embodiments are denoted by the same reference numerals, and the description thereof is omitted. The system shown in this figure is different from the other embodiments in that an FFT absolute value calculation unit 103 is provided.

  In FIG. 21, the AD conversion unit 102 receives the received signal, performs analog-digital conversion, and outputs digital data to the FFT absolute value calculation unit 103. The FFT absolute value calculation unit 103 performs FFT conversion of the input digital data, calculates the absolute value of the coefficient, and outputs it to the averaging unit 104. In addition, the signal of the non-averaged section received from the control unit 114 is output to the non-averaged compressed sensing unit 120. Note that, as will be described later, whether the averaged section or the non-averaged section is determined is determined by whether or not the absolute value of a certain frequency component exceeds a threshold value. That is, the presence or absence of a signal is determined based on the averaged power of the signal.

  The FFT absolute value calculation unit 103 receives the averaged compressed sensing frame output from the control unit 114 and performs FFT conversion for each averaged compressed sensing frame. Then, the absolute value of the coefficient after FFT conversion is calculated and output to the averaging unit 104. The averaging unit 104 averages the signals in the averaging period for each averaged compressed sensing frame and outputs an average result.

  Next, operations for performing data averaging, data compression, and decompression in the signal processing system shown in FIG. 21 will be described with reference to FIG. FIG. 22 is a flowchart showing operations for performing data averaging, data compression, and decompression in the signal processing system shown in FIG. In this figure, the same reference numerals are given to the same parts as the operation of the system of the other embodiment, and the description thereof is omitted.

  First, the AD conversion unit 102 converts the input analog signal into a digital signal, and outputs all converted data to the FFT absolute value calculation unit 103. At the same time, the data in the non-average section is output to the non-average compressed sensing 120 (step S1). The FFT absolute value calculation unit 103 performs FFT conversion on the data input from the AD conversion unit 102 for each averaged compressed sensing frame, calculates the absolute value of the result, and outputs the result to the average unit 104 (step S1A). The averaging unit 104 divides the data input from the FFT absolute value calculation unit 103 for each average period, further divides the data within the average period for each averaged compressed sensing frame, and averages the data for the compressed sensing frame. Data is output (step S2).

  Next, the operation of the AD conversion unit 102 in the signal processing system shown in FIG. 21 will be described with reference to FIG. FIG. 23 is a flowchart showing the operation of the AD conversion unit 102 in the signal processing system shown in FIG. In this figure, the same reference numerals are given to the same parts as the operation of the system of the other embodiment, and the description thereof is omitted.

  First, the input analog signal is sampled into a digital signal (step S11). All the sampled data is input to the FFT absolute value calculation unit 103 (step S12A). And the presence or absence of the non-average compression process input from the control part 114 is referred (step S13), and when there exists, the data of a non-average area are input into the non-average compression sensing part 120 (step S14).

<Sixth Embodiment>
Next, the configuration of a signal processing system according to the sixth embodiment of the present invention will be described. In the sixth embodiment, an example in which the signal processing system of the present invention is applied to the communication system shown in FIG. 10 will be described. 24 is a block diagram showing the configuration of the remote station 1 and the central station 10 shown in FIG. In this figure, the same parts as those of the system of the other embodiments are denoted by the same reference numerals and the description thereof is omitted. The system shown in this figure is different from the other embodiments in that an FFT absolute value calculation unit 203 is provided. The AD conversion unit 202 receives the received signal, performs analog-digital conversion, and outputs digital data to the FFT absolute value calculation unit 203. In addition, the non-average period signal received from the control unit 114 via the transmission path is output to the non-averaged compressed sensing unit 220. The FFT absolute value calculation unit 203 divides the digital data input from the AD conversion unit 202 for each averaged compressed sensing frame received from the control unit 214 via the transmission path, and performs an FFT conversion process. The value is calculated and output to the averaging unit 204. The averaging unit 204 averages the signals in the average interval for each averaged compressed sensing frame and outputs an average result.

<Seventh Embodiment>
Next, with reference to FIG. 26, the structure of the signal processing system by the 7th Embodiment of this invention is demonstrated. FIG. 26 is a block diagram showing the configuration of the embodiment. In this figure, the same parts as those of the system of the other embodiments are denoted by the same reference numerals, and the description thereof is omitted. The system shown in this figure is different from the other embodiments in that an FFT absolute value calculation unit 303 is provided. In FIG. 26, the AD conversion unit 302 receives the received signal, performs analog-digital conversion, and outputs digital data to the FFT absolute value calculation unit 303. In addition, the signal of the uncompressed section received from the control unit 314 is output to the outside as it is. Note that whether the averaging period or the uncompressed period is set is determined by whether or not the absolute value of a certain frequency component exceeds a threshold value. That is, the presence or absence of a signal is determined based on the averaged power of the signal.

  The FFT absolute value calculation unit 303 receives the averaged compressed sensing frame output from the control unit 114 and performs FFT conversion for each averaged compressed sensing frame. Then, the absolute value of the coefficient after FFT conversion is calculated and output to the averaging unit 304. The average unit 304, the average compressed sensing unit 306, the average restoration unit 308, the average random compression measurement matrix generation unit 310, and the average random restoration matrix generation unit 312 are respectively the average unit 104 of the sixth embodiment. The average compression sensing unit 106, the average restoration unit 108, the average random compression measurement matrix generation unit 110, and the average random restoration matrix generation unit 112 perform the same operations.

  Next, with reference to FIG. 27, the operation of performing data averaging, data compression, decompression, and output of uncompressed data in the signal processing system shown in FIG. 26 will be described. FIG. 27 is a flowchart showing operations for outputting data average, data compression, decompression, and uncompressed data in the signal processing system shown in FIG. In this figure, the same reference numerals are given to the same parts as the operation of the system of the other embodiment, and the description thereof is omitted.

  First, the AD conversion unit 302 converts the input analog signal into a digital signal, and outputs all converted data to the FFT absolute value calculation unit 103 (step S301). At the same time, the data in the non-average section is output to the non-average compressed sensing 120 (step S308). The FFT absolute value calculation unit 103 performs FFT conversion on the data input from the AD conversion unit 302 for each averaged compressed sensing frame, calculates the absolute value of the result, and outputs the result to the average unit 304 (step S301A). The averaging unit 304 divides the data input from the FFT absolute value calculation unit 103 for each average period, further divides the data within the average period for each averaged compressed sensing frame, and averages the data for the compressed sensing frame. Data is output (step S302). Step 303, step 304, and step 305 perform the same operations as step 3, step 4, and step 5 of the first embodiment shown in FIG.

  Next, the operation of the AD conversion unit 302 in the signal processing system shown in FIG. 27 will be described with reference to FIG. FIG. 28 is a flowchart showing operations for outputting data average, data compression, decompression, and uncompressed data in the signal processing system shown in FIG. In this figure, the same reference numerals are given to the same parts as the operation of the system of the other embodiment, and the description thereof is omitted. First, the input analog signal is sampled into a digital signal (step S3011). All the sampled data is input to the FFT absolute value calculation unit 103 (step S3012A). Then, the presence / absence of uncompressed processing input from the control unit 314 is referred to (step S3013), and if present, the data of the uncompressed section is output to the outside (step S3014).

<Eighth Embodiment>
Next, the configuration of a signal processing system according to the eighth embodiment of the present invention will be described. FIG. 29 shows a configuration of a communication system that includes one or more remote stations 1 to N and the central station 10, and the remote stations 1 to N and the central station 10 are connected by a transmission path. In the eighth embodiment, an example in which the signal processing system of the present invention is applied to the communication system shown in FIG. 10 will be described.

  FIG. 29 is a block diagram showing configurations of the remote station 1 and the central station 10 shown in FIG. The AD conversion unit 402 receives the received signal, performs analog-digital conversion, and outputs digital data to the FFT absolute value calculation unit 403. In addition, the signal in the uncompressed section received from the control unit 414 via the transmission path is transmitted as it is to the central station via the transmission path.

  In this way, when the frequency or time distribution is stable in a data series for which compressed sensing is performed, compressed sensing using a random measurement matrix with weighting is applied to a frequency region with a high frequency density in the signal (data series). In addition, by performing compressed sensing with a random measurement matrix without weighting the frequency domain with a low frequency density, or performing compression sensing with a random measurement matrix with weighting for the time domain with a high time density, Further, by performing compressed sensing using a random measurement matrix without weighting a time region having a low time density, the number of columns of the random measurement matrix can be reduced as compared with a case where no weight is given. The random measurement matrix is weighted by extracting the reception signal frequency or time component from the signal reception cycle or reception result to obtain the reception signal reception timing, reception signal frequency component, or time component. . By using these pieces of information, it is possible to predict the position where the frequency is not 0 in the signal, or to predict the position where the signal is on the time axis, and thus it is possible to weight the random measurement matrix. This makes it possible to reduce the amount of calculation when performing compression sampling. In addition, since the number of rows of the random measurement matrix is proportional to the compression rate of the signal, the compression rate can be reduced (high compression) compared to the case where the random measurement matrix of the related art is not weighted.

  As described above, in order to reduce the amount of transmission data in the random compression technique, after the signal is AD converted, the average or the average after the autocorrelation processing is performed so as to reduce the processing amount and processing time in the compressed sensing and restoration processing. Thus, it is possible to perform compression and restoration by compression sensing that does not average only the section in which the signal is detected, thereby reducing the processing amount and processing time required for signal detection and maintaining the quality of the detected signal.

  In conventional random compression, all signals are subjected to matrix calculation as they are using a random measurement matrix, etc., but if the signals are sparse in the frequency domain or time domain, In the same way, there is a problem that the amount of transmission data increases. In the present invention, the sampling values of the transmission waveform are divided into several groups, averaged for each group, and then subjected to matrix calculation using a random measurement matrix or the like. By this averaging, it is possible to obtain the effect of reducing the transmission data amount. On the other hand, when a signal is included in the averaged sampling value, the information is fed back to the transmission side (the side that performs random compression), and the portion including the signal is not averaged. Compressed to send. Whether or not a signal is included in the averaged sampling value can be determined by measuring the power of the averaged value. Note that autocorrelation processing can also be performed before averaging in order to facilitate signal detection.

Note that a program for realizing the functions of the processing units shown in FIGS. 1 and 11 is recorded on a computer-readable recording medium, and the program recorded on the recording medium is read into a computer system and executed. The signal processing may be performed by Here, the “computer system” includes an OS and hardware such as peripheral devices.
The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. Further, the “computer-readable recording medium” refers to a volatile memory (RAM) in a computer system that becomes a server or a client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. In addition, those holding programs for a certain period of time are also included.

  The program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. Here, the “transmission medium” for transmitting the program refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line. The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, what is called a difference file (difference program) may be sufficient.

  In compression sensing for AD conversion, compression, and restoration of signals, the present invention can be applied to signal processing applications where it is essential to generate a weighted measurement matrix.

  1, 2, N: Remote station, 10: Central station, 102, 202, 302, 402 ... AD converter, 103, 203, 303, 403 ... FFT absolute value calculator, 104, 204, 304, 404 ... average part, 106, 206, 306, 406 ... averaged compression sensing part, 108, 208, 308, 408 ... average restoration part, 110, 210, 310, 410 ..Averaged random measurement matrix generation unit, 112, 212, 312, 412 ... Averaged random restoration matrix generation unit, 114, 214, 314, 414 ... Control unit, 116, 216 ... Non-averaged Random measurement matrix generation unit, 118, 218 ... non-averaged random restoration matrix generation unit, 120, 220 ... non-averaged compressed sensing unit, 122, 222 ... non-averaged restoration unit

Claims (10)

A signal processing system that compresses and decompresses signals having a sparse property in the frequency domain or in the time domain,
An AD converter that converts an input analog signal into a digital signal and outputs the digital signal;
An averaging unit that averages and outputs the signal output from the AD conversion unit;
Averaging random measurement matrix, averaged random restoration matrix, non-averaged random measurement matrix, generation of parameters required for generation of non-averaged random restoration matrix and presence / absence of signal based on power of averaged signal A control unit for detecting and calculating a non-average interval;
An averaged random measurement matrix generating unit that generates the averaged random measurement matrix based on the parameters of the averaged random measurement matrix generation;
An averaged compressed sensing unit that compresses a signal output from the average unit using the averaged random measurement matrix unit;
A non-averaged random measurement matrix generating unit that generates the non-averaged random measurement matrix based on the parameters of the non-averaged random measurement matrix generation;
Among the digital signals output from the AD conversion unit, a non-averaged compression sensing unit that inputs a signal of a non-averaged interval and compresses the signal using the non-averaged random measurement matrix unit;
An averaged random recovery matrix generation unit that generates the averaged random recovery matrix based on the parameters of the averaged random recovery matrix generation;
An average restoration unit that restores the signal compressed by the averaged compression sensing unit using the averaged random restoration matrix, and notifies the control unit of the output of the restoration result and the restoration result;
A non-averaged random recovery matrix generation unit that generates the non-averaged random recovery matrix based on the parameters of the non-averaged random recovery matrix generation;
A non-averaged restoration unit that restores the signal compressed by the non-averaged compression sensing unit using the non-averaged random restoration matrix, and outputs a restoration result and a restoration result to the control unit. A signal processing system. A signal processing system composed of a central station and a remote station that compresses and decompresses signals having a sparse property in the frequency domain or in the time domain,
The remote station is
An AD converter that converts an input analog signal into a digital signal;
An averaging unit that averages and outputs the signal output from the AD conversion unit;
An averaged random measurement matrix generating unit that generates the averaged random measurement matrix based on the parameters of the averaged random measurement matrix generation;
An averaged compressed sensing unit that compresses a signal output from the average unit using the averaged random measurement matrix unit;
A non-averaged random measurement matrix generating unit that generates the non-averaged random measurement matrix based on the parameters of the non-averaged random measurement matrix generation;
Among the digital signals output from the AD conversion unit, a non-averaged compression sensing unit that inputs a signal of a non-averaged interval and compresses the signal using the non-averaged random measurement matrix unit;
Comprising a communication unit for communicating with the central office;
The central office is
Averaging random measurement matrix, averaged random restoration matrix, non-averaged random measurement matrix, generation of parameters required for generation of non-averaged random restoration matrix and presence / absence of signal based on power of averaged signal A control unit for detecting and calculating a non-average interval;
An averaged random recovery matrix generation unit that generates the averaged random recovery matrix based on the parameters of the averaged random recovery matrix generation;
An average restoration unit that restores the signal compressed by the averaged compression sensing unit using the averaged random restoration matrix, and notifies the control unit of the output of the restoration result and the restoration result;
A non-averaged random recovery matrix generation unit that generates the non-averaged random recovery matrix based on the parameters of the non-averaged random recovery matrix generation;
A non-averaged restoration unit that performs restoration using the non-averaged random restoration matrix, and outputs a restoration result and a restoration result to the control unit;
A signal processing system comprising: a communication unit for communicating with the remote station. A signal processing system that compresses and decompresses signals having a sparse property in the frequency domain or in the time domain,
An AD converter that converts an input analog signal into a digital signal and outputs the digital signal;
An averaging unit that averages and outputs the signal output from the AD conversion unit;
A control unit that detects the presence or absence of a signal based on generation of parameters necessary for generation of an averaged random measurement matrix and an averaged random restoration matrix and the power of the averaged signal, and calculates an uncompressed section;
Among the digital signals output from the AD converter, an output unit that outputs a signal in an uncompressed section;
An averaged random measurement matrix generating unit that generates the averaged random measurement matrix based on the parameters of the averaged random measurement matrix generation;
An averaged compressed sensing unit that compresses a signal output from the average unit using the averaged random measurement matrix unit;
An averaged random recovery matrix generation unit that generates the averaged random recovery matrix based on the parameters of the averaged random recovery matrix generation;
An average restoration unit that restores the signal compressed by the averaged compressed sensing unit using the averaged random restoration matrix, and outputs the restoration result and the restoration result to the control unit. Signal processing system. A signal processing system composed of a central station and a remote station that compresses and decompresses signals having a sparse property in the frequency domain or in the time domain,
The remote station is
An AD converter that converts an input analog signal into a digital signal;
An averaging unit that averages and outputs the signal output from the AD conversion unit;
An averaged random measurement matrix generating unit that generates the averaged random measurement matrix based on the parameters of the averaged random measurement matrix generation;
An averaged compressed sensing unit that compresses a signal output from the average unit using the averaged random measurement matrix unit;
Among the digital signals output from the AD converter, an output unit that outputs a signal in an uncompressed section;
Comprising a communication unit for communicating with the central office;
The central office is
A control unit that detects the presence or absence of a signal based on generation of parameters necessary for generation of an averaged random measurement matrix and an averaged random restoration matrix and the power of the averaged signal, and calculates an uncompressed section;
An averaged random recovery matrix generation unit that generates the averaged random recovery matrix based on the parameters of the averaged random recovery matrix generation;
An average restoration unit that restores the signal compressed by the averaged compression sensing unit using the averaged random restoration matrix, and notifies the control unit of the output of the restoration result and the restoration result;
A signal processing system comprising: a communication unit for communicating with the remote station. The average part is
An autocorrelation processing unit for performing autocorrelation calculation processing;
The signal processing system according to claim 1, further comprising: an average processing unit that performs an average calculation process. An FFT absolute value calculation unit for outputting a signal of an absolute value calculation result after FFT conversion of the signal output from the AD conversion unit;
The signal processing system according to claim 1, wherein the averaging unit averages and outputs the signal output from the FFT absolute value calculation unit. AD converter, averaging unit, control unit, averaged random measurement matrix generation unit, and non-averaged random measurement to compress and decompress signals with sparse characteristics in the frequency domain or time domain Matrix generation unit, averaged compressed sensing unit, non-averaged compressed sensing unit, averaged random recovery matrix generation unit, non-averaged random recovery matrix generation unit, averaged recovery unit, and non-averaged recovery unit A signal processing method in a signal processing system comprising:
An AD conversion step in which the AD conversion unit converts an input analog signal into a digital signal and outputs the digital signal;
An averaging step in which the averaging unit averages and outputs the signal output from the AD conversion unit;
The control unit generates an averaged random measurement matrix, an averaged random recovery matrix, a non-averaged random measurement matrix, a parameter necessary for generating the non-averaged random recovery matrix, and an averaged signal power A control step for detecting the presence or absence of a signal and calculating a non-average interval;
The averaged random measurement matrix generation unit generates the averaged random measurement matrix based on the averaged random measurement matrix generation parameter, and
The averaged compressed sensing unit compresses a signal output from the average unit using the averaged random measurement matrix unit, and an averaged compressed sensing step;
The non-averaged random measurement matrix generation unit generates the non-averaged random measurement matrix based on the parameters of the non-averaged random measurement matrix generation; and
The non-averaged compression sensing unit inputs a non-average period signal among the digital signals output from the AD conversion step, and performs non-averaged signal compression using the non-averaged random measurement matrix unit A compressed sensing step;
The averaged random recovery matrix generation unit generates the averaged random recovery matrix based on the averaged random recovery matrix generation parameter; and
The averaging restoration unit restores the signal compressed in the averaging compression sensing step using the averaged random restoration matrix, and notifies the control step of the output of the restoration result and the restoration result. When,
The non-averaged random restoration matrix generation unit generates the non-averaged random restoration matrix based on the parameters of the non-averaged random restoration matrix generation; and
The non-average restoration unit restores the signal compressed in the non-average compression sensing step using the non-average random restoration matrix, and notifies the control unit of the output of the restoration result and the restoration result. A signal processing method comprising: an averaging restoration step. AD converter, averaging unit, control unit, averaged random measurement matrix generating unit, and averaged compressed sensing unit for compressing and decompressing signals having sparse properties in the frequency domain or time domain And a signal processing method in a signal processing system comprising an averaged random restoration matrix generation unit and an averaging restoration unit,
An AD conversion step in which the AD conversion unit converts an input analog signal into a digital signal and outputs the digital signal;
An averaging step in which the averaging unit averages and outputs the signal output from the AD conversion unit;
Control in which the control unit detects the presence or absence of a signal based on the generation of parameters necessary for generation of an averaged random measurement matrix and an averaged random restoration matrix, and calculates the uncompressed interval based on the power of the averaged signal Steps,
The averaged random measurement matrix generation unit generates the averaged random measurement matrix based on the averaged random measurement matrix generation parameter, and
The averaged compressed sensing unit compresses a signal output from the average unit using the averaged random measurement matrix unit, and an averaged compressed sensing step;
The averaged random recovery matrix generation unit generates the averaged random recovery matrix based on the averaged random recovery matrix generation parameter; and
An averaging step in which the averaging restoration unit restores the signal compressed in the averaged compression sensing step using the averaged random restoration matrix, and notifies the control unit of the output of the restoration result and the restoration result; ,
A signal processing method characterized by comprising: The averaging step includes
An autocorrelation processing step for performing autocorrelation calculation processing;
The signal processing method according to claim 7, further comprising: an average processing step for performing an average calculation process. The signal processing system further includes an FFT absolute value calculation unit,
The FFT absolute value calculation unit includes an FFT absolute value calculation step of outputting a signal of an absolute value calculation result after FFT conversion of the signal output from the AD conversion unit,
The signal processing method according to claim 7 or 8, wherein the averaging step averages and outputs the signal output from the FFT absolute value calculation unit.

Images