Data compression sampling method and device

Abstract

The invention provides a data compression sampling method and device, and the method comprises the steps: obtaining an original collection signal through data collection equipment, and storing the original collection signal to a target region; processing the original acquisition signal in the target area by using a deterministic measurement matrix and a sparse transformation matrix in response to the condition that the data volume of the target area reaches a preset value to obtain a processing result; and transmitting the processing result to target equipment. Through the deterministic measurement matrix and the sparse transformation matrix, linear mapping and instant data stream compression are carried out on the sampled original acquisition signals, the data transmission amount is significantly reduced, real-time signal processing and interaction in a resource-constrained environment are supported, and the method has small limitation on a hardware environment, has good transportability and deployment flexibility, and is suitable for large-scale popularization and application. The method is suitable for various system application requirements, so that efficient data compression sampling is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information and communication technology, and in particular to a data compression sampling method and device. BACKGROUND

[0002] In application scenarios such as structural health monitoring and industrial equipment monitoring, the popularity of intelligent sensing devices leads to an exponential growth of the amount of monitoring data collected, which puts forward more stringent technical requirements for data transmission and storage systems.

[0003] The existing data acquisition device mainly adopts an equal-interval sampling method to collect and transmit analog signals point by point based on the Nyquist sampling theorem. Although such a device has the advantage of easy implementation, when dealing with high-frequency signals or high-precision monitoring requirements, it will produce obvious data redundancy, leading to a sharp consumption of storage resources and increasing the load of wireless communication bandwidth and system computing resources. For embedded monitoring systems with strict real-time requirements, low-power constraints or limited computing resources, this full-quantity acquisition and transmission mode has been difficult to meet the actual engineering requirements.

[0004] In summary, the prior art has the problem of heavy load of communication bandwidth and computing resources. SUMMARY

[0005] The present application provides a data compression sampling method and device to solve the defect of heavy load of communication bandwidth and computing resources in the prior art, reduce the load of communication bandwidth and computing resources, and realize high-efficiency data compression sampling.

[0006] The present application provides a data compression sampling method, comprising the following steps:

[0007] acquiring an original acquisition signal using a data acquisition device, and storing the original acquisition signal in a target area;

[0008] in response to the data amount of the target area reaching a preset value, processing the original acquisition signal in the target area using a deterministic measurement matrix and a sparse transformation matrix to obtain a processing result;

[0009] transmitting the processing result to a target device.

[0010] According to the data compression sampling method provided by the present application, the construction step of the deterministic measurement matrix comprises:

[0011] constructing a target finite field and a target polynomial function family, wherein the characteristic number of the target finite field is a target prime number, and the highest degree of the target polynomial function family is less than the target prime number;

[0012] Based on the target finite field, the target polynomial function family is subjected to coefficient value, and a plurality of target coefficient combinations are obtained;

[0013] The plurality of target coefficient combinations are substituted into the target polynomial function family, and a plurality of target mapping matrices are obtained, wherein the target mapping matrix is a mapping matrix of the independent variable to the polynomial value of the target polynomial function family;

[0014] The plurality of target mapping matrices are expanded to obtain a binary matrix;

[0015] The binary matrix is subjected to dimension expansion to obtain a deterministic measurement matrix.

[0016] According to the data compression sampling method provided by the present application, the sparse transform matrix is constructed according to the characteristics of the original acquisition signal, and the sparse transform matrix is used to map the original acquisition signal to a sparse representation domain.

[0017] According to the data compression sampling method provided by the present application, the original acquisition signal is obtained by using a data acquisition device, and the original acquisition signal is stored in a target area, comprising:

[0018] Timing starts ADC sampling, and the original acquisition signal is obtained by using a data acquisition device;

[0019] The original acquisition signal is preprocessed based on the pre-set preprocessing logic, and the preprocessing result is stored in the target area.

[0020] According to the data compression sampling method provided by the present application, in response to the data amount of the target area reaching a preset value, the original acquisition signal in the target area is processed by using a deterministic measurement matrix and a sparse transform matrix to obtain a processing result, comprising:

[0021] In response to the data amount of the target area reaching a preset value, the original acquisition signal in the target area is subjected to a sparse transform operation by using the sparse transform matrix to obtain a first operation result;

[0022] The first operation result is subjected to a linear mapping operation by using the deterministic measurement matrix to obtain a compressed low-dimensional observation vector as a processing result.

[0023] According to the data compression sampling method provided by the present application, the processing result is transmitted to a target device, and then further comprising:

[0024] The processing result is reconstructed to obtain an approximate reconstruction result of the original acquisition signal.

[0025] According to the data compression sampling method provided in the application, the binary matrix is dimensionally expanded to obtain a deterministic measurement matrix, including:

[0026] The binary matrix is diagonally expanded as a sub-block matrix to obtain the deterministic measurement matrix.

[0027] The application further provides a data compression sampling device, including the following modules:

[0028] The sampling unit is configured to acquire an original acquisition signal by using a data acquisition device and store the original acquisition signal in a target area.

[0029] The compression unit is configured to, in response to the data amount of the target area reaching a preset value, process the original acquisition signal in the target area by using a deterministic measurement matrix and a sparse transformation matrix to obtain a processing result.

[0030] The transmission unit is configured to transmit the processing result to a target device.

[0031] The application further provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the data compression sampling method according to any one of the above when executing the program.

[0032] The application further provides a non-transitory computer-readable storage medium having a computer program stored thereon, and the computer program is executable by a processor to implement the data compression sampling method according to any one of the above.

[0033] The application further provides a computer program product including a computer program, and the computer program is executable by a processor to implement the data compression sampling method according to any one of the above.

[0034] The data compression sampling method and device provided in the application perform linear mapping on the original acquisition signal by using the deterministic measurement matrix and the sparse transformation matrix, compress the data stream in real time, significantly reduce the data transmission amount, support real-time signal processing and interaction in a resource-limited environment, and have small limitation on the hardware environment, good portability, and flexible deployment, and are suitable for various system application requirements, thereby realizing efficient data compression sampling. BRIEF DESCRIPTION OF DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are some embodiments of the application, and those skilled in the art can obtain other drawings according to these drawings without any creative effort.

[0036] Figure 1 is a flowchart of a data compression sampling method provided by the present application;

[0037] Figure 2 is a schematic diagram of a result of reconstructing a processing result in the data compression sampling method provided by the present application;

[0038] Figure 3 is a difference quantization diagram of a reconstructed signal and an original acquisition signal of the data compression sampling method provided by the present application;

[0039] Figure 4 is a structural schematic diagram of a data compression sampling device provided by the present application;

[0040] Figure 5 is a structural schematic diagram of an electronic device provided by the present application. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical scheme and advantages of the present application clearer, the technical scheme of the present application will be described clearly and completely below in combination with the drawings in the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0042] Most of the data processing using the compressed sensing theory is still in the algorithm stage, and the random measurement matrix is difficult to be efficiently deployed in the embedded platform, resulting in large system resource consumption, poor real-time performance, and difficulty in meeting the application requirements of low power consumption, miniaturization and strong real-time performance of the embedded system. The compressed sensing theory provides an innovative solution to the efficient data compression problem.

[0043] The core principle of the technology is that when the target signal meets the sparsity condition, by designing a specific linear measurement matrix, only a small amount of non-adaptive linear measurement values need to be obtained to retain the key feature information of the signal. This method can significantly reduce the actual sampling point number under the premise of ensuring the signal reconstructability, thereby effectively avoiding the generation of redundant data and significantly reducing the system burden of the data processing, transmission and storage links.

[0044] However, most of the existing compressed sensing technologies are focused on algorithm research and simulation verification level, and lack of hardware level implementation path. On the one hand, the widely used random measurement matrix (such as Gaussian matrix, Bernoulli matrix) has superior theoretical performance, but is difficult to implement in actual embedded platform deployment, facing problems such as large storage overhead and high computational complexity; on the other hand, the mainstream compressed sensing hardware implementation relies on high-cost platforms such as FPGA and DSP, which is difficult to popularize in power-constrained and small-size embedded scenarios.

[0045] On this basis, the application provides a data compression sampling method and device.

[0046] The data compression sampling method of the application is described below Figures 1-3 The data compression sampling method of the application is described below Figure 1 The flowchart of the data compression sampling method provided by the application is shown in Figure 1 The method comprises the following steps.

[0047] Step 110: acquiring an original acquisition signal by using a data acquisition device, and storing the original acquisition signal in a target area.

[0048] The data acquisition device is used to acquire the original acquisition signal. The application does not limit the position of the data acquisition device. The data acquisition device can be arranged outside the execution platform of the method provided by the application, such as an external intelligent sensor, or can be arranged inside the execution platform of the method provided by the application.

[0049] In some embodiments, the original acquisition signal comprises a physical signal (such as vibration, current, temperature, etc.).

[0050] After the original acquisition signal is acquired, the original acquisition signal is stored in a preset target area. The target area can be any pre-set area on the execution platform of the method provided by the application.

[0051] In one specific embodiment, the method provided by the application is executed on an embedded microcontroller (MCU) platform, and the target area comprises a buffer area of the embedded microcontroller (MCU).

[0052] Step 120: in response to the data amount of the target area reaching a preset value, processing the original acquisition signal in the target area by using a deterministic measurement matrix and a sparse transformation matrix to obtain a processing result.

[0053] The deterministic measurement matrix and the sparse transformation matrix are pre-constructed and stored in the target area, so as to avoid repeated loading each time the method is run, and effectively improve the access efficiency and real-time response capability.

[0054] The deterministic measurement matrix is used for compressive sensing; and the sparse transformation matrix is used for mapping the original signal to a sparse representation domain to meet the reconstruction premise of compressive sensing.

[0055] This step sets the trigger condition of the compression transmission task, i.e., the data amount of the target area reaches the preset value. After the trigger condition is met, a compression transmission task is automatically triggered to compress the original acquisition signal stored in the target area.

[0056] In the compression process, the original acquisition signal in the target region is compressed in sequence by using a sparse transformation matrix and a deterministic measurement matrix in the target region to obtain a processing result. The sparse transformation matrix is used for sparse transformation, and the deterministic measurement matrix is used for linear mapping.

[0057] In the specific implementation, based on the above embodiment, the method provided in the application is executed on an embedded microcontroller (MCU) platform, the target region includes a buffer area of the embedded microcontroller (MCU), and meanwhile, a fixed-capacity data buffering mechanism is used, when the data amount of the buffer area reaches a set trigger, the original signal is subjected to sparse linear mapping by using a preset sparse transformation matrix and a measurement matrix to generate a low-dimensional observation vector.

[0058] Further, in some embodiments, in the implementation of the application, the data acquisition and subsequent compression process can be synchronously run by using a pipeline mechanism, so that sampling, compression and transmission are processed in parallel. The original acquisition signal obtained by the current batch of sampling enters the compression calculation stage at the same time as the next batch of sampling operation is started synchronously, so that the continuous data processing process is highly overlapped, and the system throughput and real-time responsiveness are greatly improved.

[0059] Step 130: transmitting the processing result to a target device.

[0060] The target device includes an upper system and a remote terminal.

[0061] In the transmission process, the processing result can be transmitted to the target device through a communication module interface, or can be transmitted by other means, which is not limited in the application.

[0062] In some embodiments, the processing result is sent to the target device through a UART (Universal Asynchronous Receiver / Transmitter, universal asynchronous receiver / transmitter), SPI (Serial Peripheral Interface, serial peripheral interface) or CAN (Controller Area Network, controller area network) communication interface module, to complete a real-time low-bandwidth transmission task.

[0063] Further, in some embodiments, the processing result is sent to the target device in real time by using a serial communication mode.

[0064] Each step of the application can be implemented on a general MCU, has good portability and deployment flexibility, and is suitable for various system application requirements.

[0065] The data compression sampling method provided in the application solves the engineering adaptability problem of a random matrix by designing and constructing a deterministic measurement matrix to replace the random matrix in the prior art, and integrating the deterministic measurement matrix into a target region; by linear mapping of sensor output data, the data transmission amount is significantly reduced, and real-time signal processing and interaction in a resource-limited environment are supported.

[0066] The application samples original signals and compresses data streams in real time, solves the hardware deployment problem of compressive sensing from the structural design level, thereby reducing the transmission and storage burden, improving the overall operation efficiency of the system, realizing an integrated closed-loop process of data acquisition, compression and transmission, and improving the data processing efficiency and operation stability of the system. The application is applicable to any embedded scene requiring low-complexity compression sampling, including but not limited to device and structural health monitoring, etc.

[0067] The deterministic measurement matrix is further described below. In some embodiments, the construction step of the deterministic measurement matrix comprises:

[0068] Step 210: constructing a target finite field and a target polynomial function family, wherein the characteristic number of the target finite field is a target prime number, and the highest degree of the target polynomial function family is less than the target prime number.

[0069] This step first needs to construct a finite field F as a target finite field, and the characteristic number of which is a target prime number p. At the same time, define a polynomial function family P r as a target polynomial function family, where 0 < r < p:

[0070] P r ={Q(t)=a0+a1t+…+a n t r |a n ∈F};

[0071] wherein a0, a1, …, a n are element values in the finite field F, and the mapping matrix of t→Q(t) is denoted as E.

[0072] Step 220: based on the target finite field, performing coefficient value taking on the target polynomial function family to obtain a plurality of target coefficient combinations.

[0073] Specifically, based on the finite field F, a plurality of different polynomial coefficient value combinations are obtained as a plurality of target coefficient combinations by performing coefficient value taking.

[0074] Step 230: substituting the plurality of target coefficient combinations into the target polynomial function family to obtain a plurality of target mapping matrices, wherein the target mapping matrix is a mapping matrix of the independent variable to the polynomial value of the target polynomial function family.

[0075] According to the different combinations of polynomial coefficient values obtained in step 220, the target polynomial function family is substituted to construct a plurality of mapping matrices E of t→Q(t), denoted as target mapping matrices.

[0076] Step 240: The plurality of target mapping matrices is unfolded to obtain a binary matrix.

[0077] For each target mapping matrix, it is unfolded into a first-dimensional vector, and the first-dimensional vectors of the plurality of target mapping matrices are stacked to obtain a binary matrix.

[0078] It should be noted that in the unfolding process of the target mapping matrix, the dimension of the unfolded vector is not limited by the present application.

[0079] In one specific embodiment, the target mapping matrix is unfolded in sequence to form a p 2 ×1 vector, thereby constructing a binary matrix Φ0with a dimension of p 2 ×p r+1 .

[0080] Step 250: Dimension expansion is performed on the binary matrix to obtain a deterministic measurement matrix.

[0081] The binary matrix is dynamically adjusted and expanded to obtain a measurement matrix that meets the requirements of specific dimensions and sparse reconstruction, as a deterministic measurement matrix.

[0082] It should be noted that the dimension expansion method of the binary matrix is not limited by the present application, and can be flexibly selected according to different scenarios and data characteristics, including but not limited to matrix splicing, cyclic arrangement, block combination, and other expansion strategies.

[0083] It should be noted that the dimension of the deterministic measurement matrix can be set according to actual conditions, and the present application does not make specific limitations.

[0084] On this basis, further, in the dimension expansion process, the compression ratio can be flexibly adjusted according to the specific application scenario to adapt to different scales of data characteristics.

[0085] The sparse transform matrix is further described below. In some embodiments, the sparse transform matrix is constructed according to the characteristics of the original acquisition signal, and the sparse transform matrix is used to map the original acquisition signal to a sparse representation domain.

[0086] The sparse transform matrix is constructed according to the characteristics of the original acquisition signal. The present application does not limit the specific type of the sparse transform matrix, which can be flexibly selected according to actual applications, including but not limited to a discrete cosine transform (DCT) matrix, a discrete wavelet transform (DWT) matrix, or a discrete Fourier transform (DFT) matrix, etc.

[0087] Further, in some embodiments, acquiring the original acquisition signal by using the data acquisition device and storing the original acquisition signal to the target area, comprising:

[0088] Step 310: timing starting ADC sampling, acquiring the original acquisition signal by using the data acquisition device.

[0089] In this embodiment, the original acquisition signal acquired by the data acquisition device is digitized by the ADC module and then enters the target area, realizing non-blocking data acquisition and storage transfer.

[0090] Step 320: pre-processing the original acquisition signal based on the pre-set pre-processing logic, and storing the pre-processing result to the target area.

[0091] In this embodiment, the front-end signal quality enhancement mechanism is set by the hardware pre-processing logic, which can suppress noise of the original signal during the sampling process and enhance stability and anti-interference ability.

[0092] In one specific embodiment, the above pre-processing logic is realized by ADC and hardware configuration.

[0093] Among them, the pre-processing method can be selected according to the characteristics of the original acquisition signal, including but not limited to baseline drift correction, mean filtering, feature enhancement, etc.

[0094] Further, in some embodiments, in response to the data amount of the target area reaching a preset value, the original acquisition signal in the target area is processed by using a deterministic measurement matrix and a sparse transformation matrix to obtain a processing result, comprising:

[0095] Step 121: in response to the data amount of the target area reaching a preset value, performing a sparse transformation operation on the original acquisition signal in the target area by using the sparse transformation matrix to obtain a first operation result;

[0096] Step 122: performing a linear mapping operation on the first operation result by using the deterministic measurement matrix to obtain a compressed low-dimensional observation vector as a processing result.

[0097] Specifically, in the process of processing the original acquisition signal, the sparse transformation matrix and the deterministic measurement matrix are used to perform sparse transformation and linear mapping operations on the original acquisition signal in the target area in sequence to obtain a compressed low-dimensional observation vector. Specifically, first, the sparse transformation matrix is used to perform a sparse transformation operation on the original acquisition signal to obtain a first operation result, and then the deterministic measurement matrix is used to perform a linear mapping operation on the first operation result to obtain a processing result.

[0098] In some embodiments, the processing result is transmitted to the target device, and then the method further comprises:

[0099] Step 140: reconstructing the processing result to obtain an approximate reconstruction result of the original acquisition signal.

[0100] Specifically, the target device executes an optimized reconstruction algorithm on the compressed processing result, so as to obtain an approximate reconstruction result of the original acquisition signal.

[0101] In the implementation of the present application, the target device continuously receives the compressed processing result until the data acquisition device stops collecting data. A sparse reconstruction algorithm is executed on the received processing result to recover the original high-dimensional signal data, which is used for subsequent tasks such as device and structure operation state monitoring.

[0102] In some embodiments, the binary matrix is dimensionally extended to obtain a deterministic measurement matrix, comprising:

[0103] Step 251: diagonally extending the binary matrix as a sub-block matrix to obtain a deterministic measurement matrix.

[0104] In the present embodiment, the binary matrix is diagonally extended as a sub-block matrix to obtain a high-dimensional deterministic measurement matrix.

[0105] The present application does not elaborate on the specific extension steps of diagonal extension, which can be implemented according to actual conditions.

[0106] The data compression sampling method provided by the present application significantly reduces the data transmission amount by linearly mapping the obtained original acquisition signal through the deterministic measurement matrix and the sparse transformation matrix, supports real-time signal processing and interaction in a resource-limited environment, has small hardware environment limitation, has good portability and deployment flexibility, is suitable for various system application requirements, and thus realizes high running efficiency of data compression sampling.

[0107] Further, in order to more specifically describe the data compression sampling method provided by the present application, a specific embodiment is given.

[0108] The present application implements the data compression sampling method provided by the present application based on a control moment gyro (CMG) experimental table, the original acquisition signal is an actual vibration acceleration signal, the data acquisition device is a sensor, and the target device is an upper computer.

[0109] In actual implementation, the embodiment is designed and deployed based on an STM32H7 series embedded platform, a deterministic measurement matrix, a sparse transformation matrix and a compression encoding module are integrated in a target area (i.e., a buffer), and the original data output by a sensor can be compressed in real time. Whenever the original collected signal reaches a set trigger condition, the compression sensing processing flow is started, and the processed result obtained after compression is sent to the upper computer through a serial communication interface.

[0110] In the experiment, a large amount of high-frequency data collected by the sensor is compressed, which effectively reduces the data transmission amount and greatly reduces the communication bandwidth and the computing pressure of the upper computer. After receiving the signal, the upper computer uses a reconstruction algorithm to complete data recovery, the reconstruction signal error is controlled within a reasonable range, the key features and dynamic change trend of the target signal are successfully preserved, and the deployability and effectiveness of the data compression sampling method provided by the application in the actual embedded environment and the engineering test platform are verified.

[0111] The running steps of the embodiment include:

[0112] S101: Based on the algebraic structure of a finite field and the polynomial mapping rule, a deterministic measurement matrix with good mutual irrelevance is designed and constructed, and a high-dimensional matrix is generated through a four-fold diagonal block expansion method and is preloaded to the cache of an MCU;

[0113] S102: A sparse transformation module composed of a sparse transformation matrix is pre-set in the MCU to realize sparse expression of the input signal and meet the reconstruction condition of the subsequent compression sensing theory;

[0114] S103: The vibration acceleration signal of an integrated circuit piezoelectric (IEPE) sensor is collected by a timer driving ADC, real-time sampling and preliminary noise filtering are completed without adding additional delay, and the original collected signal is stored in the buffer;

[0115] S104: When the original collected signal in the buffer reaches a set value, the system automatically triggers a compression transmission task, first maps the original time domain signal to a target sparse domain through a pre-set sparse transformation matrix to generate a sparse coefficient vector;

[0116] S105: The pre-set deterministic measurement matrix is called to perform a multiplication and addition operation in a finite field on the sparse vector to obtain a low-dimensional observation vector as a processing result;

[0117] S106: The compressed observation values (i.e., the processed result obtained after compression) are sent to the upper computer one by one through a UART communication interface module to complete a real-time low-bandwidth transmission task;

[0118] S107: The host computer end continues to receive the compressed processing results, and pauses the data sampling of the front-end sensor after two minutes. The host computer receives a total of 4000 compressed data points, with a compression ratio of 5:1. The sparse reconstruction algorithm is executed to reconstruct an approximate expression of the original signal containing 20000 data points in the time domain, and the reconstruction error is calculated.

[0119] The comparison of the data compression sampling method provided by the present application with the traditional data acquisition method and the data compression sampling method of the random measurement matrix is shown in Table 1.

[0120] Table 1 Comparison of the data compression sampling method provided by the present application with the traditional data acquisition method and the data compression sampling method of the random measurement matrix

[0121]

[0122]

[0123] As shown in Table 1, the deterministic measurement matrix is deployed in the embedded system, which significantly optimizes the performance bottleneck of the traditional data sampling method of "full sampling + full transmission". Compared with the traditional full sampling method, the total transmission data volume, memory occupation and running time of the present application are greatly reduced, providing better guarantee for the real-time response of the system.

[0124] The deterministic measurement matrix used in the present application does not need to be randomly generated at runtime, and has a clear structure and reusability. Although the reconstruction error is slightly higher, the time consumed for reconstruction of the compressed processing results collected by the present application in the host computer is shortened by nearly one time compared with the random measurement matrix, significantly reducing the storage demand and computational complexity of the embedded system.

[0125] In addition, the compressed processing results generated by the data compression sampling method provided by the present application are transmitted to the host computer in real time through a serial port with low bandwidth. As shown in Figure 2 , the reconstructed waveform is highly consistent with the main structure of the original signal, and the noise and other interference factors are reduced, and the reconstruction error is stably controlled within an acceptable range. The difference between the reconstructed signal and the original time domain signal is shown in Figure 3 , the signal difference is slightly fluctuating around zero, which verifies the good balance between the compression performance and the recovery accuracy of the data compression sampling method provided by the present application.

[0126] The data compression sampling device provided by the present application is described below. The data compression sampling device described below can be referred to in conjunction with the data compression sampling method described above. Figure 4 is a structural schematic diagram of the data compression sampling device provided by the present application, as shown in Figure 4 , the device includes the following modules:

[0127] The sampling unit 410 is configured to acquire an original acquisition signal by using a data acquisition device and store the original acquisition signal in a target area.

[0128] The compression unit 420 is configured to, in response to the data amount of the target area reaching a preset value, process the original acquisition signal in the target area by using a deterministic measurement matrix and a sparse transformation matrix to obtain a processing result.

[0129] The transmission unit 430 is configured to transmit the processing result to a target device.

[0130] According to the data compression sampling device provided in the present application, the construction of the deterministic measurement matrix comprises:

[0131] constructing a target finite field and a target polynomial function family, wherein the characteristic number of the target finite field is a target prime number, and the highest degree of the target polynomial function family is less than the target prime number;

[0132] based on the target finite field, performing coefficient value setting on the target polynomial function family to obtain a plurality of target coefficient combinations;

[0133] substituting the plurality of target coefficient combinations into the target polynomial function family to obtain a plurality of target mapping matrices, wherein the target mapping matrix is a mapping matrix of the independent variable to the polynomial value of the target polynomial function family;

[0134] expanding the plurality of target mapping matrices to obtain a binary matrix;

[0135] performing dimension extension on the binary matrix to obtain a deterministic measurement matrix.

[0136] According to the data compression sampling device provided in the present application, the sparse transformation matrix is constructed according to the characteristics of the original acquisition signal, and the sparse transformation matrix is used to map the original acquisition signal to a sparse representation domain.

[0137] According to the data compression sampling device provided in the present application, the original acquisition signal is acquired by using a data acquisition device and stored in a target area, comprising:

[0138] starting ADC sampling at a fixed time, and acquiring the original acquisition signal by using a data acquisition device;

[0139] preprocessing the original acquisition signal based on a pre-set preprocessing logic, and storing the preprocessing result in a target area.

[0140] According to the data compression sampling device provided in this application, in response to the data volume of the target area reaching a preset value, the original acquired signal in the target area is processed using a deterministic measurement matrix and a sparse transformation matrix to obtain a processing result, including:

[0141] In response to the data volume in the target area reaching a preset value, the original acquired signal in the target area is subjected to a sparse transformation operation using the sparse transformation matrix to obtain a first operation result;

[0142] The first operation result is linearly mapped using the deterministic measurement matrix to obtain a compressed low-dimensional observation vector as the processing result.

[0143] According to the data compression sampling device provided in this application, the processing result is transmitted to the target device, and then the device further includes:

[0144] The processing result is reconstructed to obtain an approximate reconstruction result of the original acquired signal.

[0145] According to a data compression sampling device provided in this application, the binarized matrix is ​​dimensionally expanded to obtain a deterministic measurement matrix, including:

[0146] The binarized matrix is ​​diagonally expanded as a sub-block matrix to obtain the deterministic measurement matrix.

[0147] Figure 5 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 5 As shown, the electronic device may include a processor 510, a communication interface 520, a memory 530, and a communication bus 540, wherein the processor 510, the communication interface 520, and the memory 530 communicate with each other via the communication bus 540. The processor 510 can call logical instructions in the memory 530 to execute a data compression sampling method, which includes: acquiring a raw acquisition signal using a data acquisition device and storing the raw acquisition signal in a target area; in response to the data volume in the target area reaching a preset value, processing the raw acquisition signal in the target area using a deterministic measurement matrix and a sparse transformation matrix to obtain a processing result; and transmitting the processing result to a target device.

[0148] Further, the logic instructions in the memory 530 described above can be implemented in the form of software functional units and sold or used as independent products, and can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the parts that make contributions to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0149] In another aspect, the present application also provides a computer program product, which comprises a computer program, the computer program can be stored on a non-transitory computer readable storage medium, and the computer program is executed by a processor, so that the computer can execute the data compression sampling method provided by the above-mentioned methods. The method comprises: acquiring an original acquisition signal by using a data acquisition device, and storing the original acquisition signal to a target area; in response to the data volume of the target area reaching a preset value, processing the original acquisition signal in the target area by using a deterministic measurement matrix and a sparse transformation matrix to obtain a processing result; and transmitting the processing result to a target device.

[0150] In another aspect, the present application also provides a non-transitory computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the data compression sampling method provided by the above-mentioned methods. The method comprises: acquiring an original acquisition signal by using a data acquisition device, and storing the original acquisition signal to a target area; in response to the data volume of the target area reaching a preset value, processing the original acquisition signal in the target area by using a deterministic measurement matrix and a sparse transformation matrix to obtain a processing result; and transmitting the processing result to a target device.

[0151] The device embodiments described above are only schematic, wherein the units described as separate components can or can not be physically separate, and the components displayed as units can or can not be physical units, that is, they can be located in one place, or distributed on multiple network units. Part or all of the modules can be selected to achieve the purpose of the present embodiment scheme according to actual needs. Those skilled in the art can understand and implement without creative labor.

[0152] Those skilled in the art can clearly understand the implementation of the various embodiments by means of software and necessary general hardware platforms through the description of the above embodiments, and of course, the implementation can also be through hardware. Based on such understanding, the above technical solutions, essentially or in other words, the part of the prior art that makes a contribution, can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in the various embodiments or some parts of the embodiments.

[0153] Finally, it should be noted that: the above examples are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing examples, or make equivalent replacement for some technical features thereof; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A data compression sampling method, characterized by, The method comprises the following steps: acquiring original acquisition signals by using a data acquisition device, and storing the original acquisition signals to a target area; in response to the data volume of the target area reaching a preset value, processing the original acquisition signals in the target area by using a deterministic measurement matrix and a sparse transform matrix to obtain a processing result; transmitting the processing result to a target device.

2. The data compression sampling method of claim 1, wherein, The construction of the deterministic measurement matrix comprises the following steps: constructing a target finite field and a target polynomial function family, wherein the characteristic number of the target finite field is a target prime number, and the highest degree of the target polynomial function family is less than the target prime number; based on the target finite field, performing coefficient value setting on the target polynomial function family to obtain a plurality of target coefficient combinations; substituting the plurality of target coefficient combinations into the target polynomial function family to obtain a plurality of target mapping matrices, wherein the target mapping matrix is a mapping matrix from the independent variable of the target polynomial function family to the polynomial value; expanding the plurality of target mapping matrices to obtain a binary matrix; dimensionally extending the binary matrix to obtain a deterministic measurement matrix.

3. The data compression sampling method of claim 1, wherein, The sparse transform matrix is constructed according to the characteristics of the original acquisition signals, and is used to map the original acquisition signals to a sparse representation domain.

4. The data compressive sampling method of claim 1, wherein, Acquiring original acquisition signals by using a data acquisition device, and storing the original acquisition signals to a target area, comprises: starting ADC sampling at a fixed time, and acquiring original acquisition signals by using a data acquisition device; preprocessing the original acquisition signals based on preprocessing logic, and storing the preprocessing result to a target area.

5. The data compressive sampling method of claim 1, wherein, In response to the data volume of the target area reaching a preset value, processing the original acquisition signals in the target area by using a deterministic measurement matrix and a sparse transform matrix to obtain a processing result, comprises: in response to the data volume of the target area reaching a preset value, performing a sparse transform operation on the original acquisition signals in the target area by using the sparse transform matrix to obtain a first operation result; performing a linear mapping operation on the first operation result by using the deterministic measurement matrix to obtain a compressed low-dimensional observation vector as a processing result.

6. The data compressive sampling method of claim 1, wherein, Transmitting the processing result to a target device, and then further comprising: reconstructing the processing result to obtain an approximate reconstruction result of the original acquisition signals.

7. The data compressive sampling method of claim 2, wherein, Dimensionally extending the binary matrix to obtain a deterministic measurement matrix, comprises: diagonally extending the binary matrix as a sub-block matrix to obtain a deterministic measurement matrix.

8. A data compression sampling device, characterized by, The method comprises the following steps: a sampling unit is configured to acquire original acquisition signals by using a data acquisition device, and store the original acquisition signals to a target area; a compression unit is configured to, in response to the data volume of the target area reaching a preset value, process the original acquisition signals in the target area by using a deterministic measurement matrix and a sparse transform matrix to obtain a processing result; a transmission unit is configured to transmit the processing result to a target device.

9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the data compression sampling method according to any one of claims 1 to 7. 10.A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by a processor, implements the data compression sampling method according to any one of claims 1 to 7.

Images