Data aggregation method, system, device and storage medium based on a remainder system

Abstract

The application discloses a data aggregation method, system and device based on a remainder system and a storage medium, which is a data aggregation scheme satisfying energy saving, reliability, non-plaintext transmission and tolerance to measurement error (one unlimited size and any small error). In the scheme, each source node only transmits compressed data, so that the energy of the sensor node can be saved, and the data erasure problem caused by failure or wireless network unreliability can be tolerated, so that the reliability is relatively high. Meanwhile, the compression of the source node to the sensing data is essentially a lightweight encryption. If the corresponding modulus is unknown, the sensing data cannot be obtained even if the compressed data is obtained, so the data transmission belongs to non-plaintext data transmission. In addition, one unlimited size measurement error can be tolerated, which is more in line with the actual needs.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of information aggregation, and particularly relates to a data aggregation method and system based on a remainder system, a device and a storage medium. BACKGROUND

[0002] Wireless sensor networks (WSNs) are a kind of multi-hop self-organizing network system, which has been successfully applied in different fields such as data collection, security monitoring and smart home. In order to improve the sensing accuracy and coverage, a large number of sensor nodes are usually deployed in the monitoring area, and the high density feature makes the data measured by adjacent sensing nodes the same or similar, resulting in a large amount of redundant data in the network. The communication overhead in the network is the main energy consumption aspect, and the energy consumption of transmitting 1 bit of information 100 meters away is equivalent to executing 3000 computing instructions, so reducing the transmission of redundant data is the top priority.

[0003] In view of the problems caused by data redundancy and communication overhead, data aggregation technology emerges as the times require. The traditional data aggregation (in-network data aggregation) refers to that part of the sensing nodes in the network, after receiving the data of the predecessor node, perform some merging processing (such as mean value) on the received data and the data sensed by itself, and then forward to the successor node. In 2013, Campobello et al. proposed an in-network data aggregation scheme based on the Chinese remainder theorem, which achieved good results. Since the amount of data transmitted is greatly reduced, the transmission energy consumption is reduced, thereby effectively prolonging the life cycle of WSNs.

[0004] However, traditional data aggregation has some problems. For example: there is no reduction of redundant data transmission from source nodes to aggregation nodes; no consideration of reliability, aggregated data is easy to be unable to reach the sink node due to the unreliability of the wireless network; no consideration of security, if no additional encryption measures are taken, the perception data is transmitted in plaintext. In view of these defects, researchers have also proposed data aggregation based on compressed sensing on the basis of traditional data aggregation scheme, such as in 2018, Mehrjoo et al. aimed at the error of compressed sensing reconstruction phase and the improvement of reconstruction accuracy to realize adaptive processing, but the signal is required to be sparse; data aggregation based on distributed source coding, such as in 2021, Manikandan et al. proposed a method to provide simple routing services, which can reduce traffic delay and end-to-end delay, and use distributed source coding and high energy consumption method to achieve better network performance, but it needs an independent encoder with prior function; data aggregation based on robust Chinese remainder theorem, such as in 2022, Zhang et al. proposed a data aggregation scheme based on robust Chinese remainder theorem, which can meet the requirements of energy saving, reliability and non-plaintext transmission at the same time, and does not require the signal to be sparse nor an independent encoder with prior function, but requires the source node to have only small measurement error, which is difficult to guarantee in reality due to various faults and disturbances.

[0005] Therefore, it is necessary to design a data aggregation scheme that meets the requirements of energy saving, reliability and non-plaintext transmission and tolerates a measurement error of unlimited size. SUMMARY

[0006] The purpose of the present application is to provide a data aggregation method, system, device and storage medium based on the remainder system, which is a data aggregation scheme that meets the requirements of energy saving, reliability, non-plaintext transmission and tolerates measurement error (a measurement error of unlimited size and any small error).

[0007] The purpose of the present application is achieved by the following technical solutions:

[0008] A data aggregation method based on the remainder system, comprising:

[0009] Selecting a positive integer according to the number of source nodes to form a modulus corresponding to each source node;

[0010] Each source node perceives the measured physical data, compresses the perception data in combination with its own modulus, and then stores and transmits the compressed perception data to the sink node;

[0011] The sink node receives a part of compressed sensing data of the source nodes, wherein a sensing node has unlimited size measurement error, the sink node selects compressed sensing data of any source node as a reference to adjust the received compressed sensing data set, constructs a suspect set by combining intermediate information obtained in the adjustment process, and constructs a new compressed sensing data set based on the suspect set, selects compressed sensing data of any source node from the new compressed sensing data set as a reference to adjust the new compressed sensing data set, and reconstructs the sensing data by combining the Chinese remainder theorem if the adjusted compressed sensing data set meets the requirements.

[0012] A data aggregation system based on a remainder system, comprising:

[0013] An integer selection part selects positive integers according to the number of source nodes to form moduli corresponding to the source nodes;

[0014] A sensing data compression storage part senses measured physical data by each source node, compresses the sensing data in combination with the modulus of the source node, and stores and transmits the compressed sensing data to the sink node;

[0015] A sensing data reconstruction part, wherein the sink node receives a part of compressed sensing data of the source nodes, wherein a sensing node has unlimited size measurement error, the sink node selects compressed sensing data of any source node as a reference to adjust the received compressed sensing data set, constructs a suspect set by combining intermediate information obtained in the adjustment process, and constructs a new compressed sensing data set based on the suspect set, selects compressed sensing data of any source node from the new compressed sensing data set as a reference to adjust the new compressed sensing data set, and reconstructs the sensing data by combining the Chinese remainder theorem if the adjusted compressed sensing data set meets the requirements.

[0016] A processing device, comprising: one or more processors; a memory for storing one or more programs;

[0017] When the one or more programs are executed by the one or more processors, the one or more processors implement the foregoing method.

[0018] A readable storage medium storing a computer program, when the computer program is executed by a processor, the foregoing method is implemented.

[0019] From the above technical solutions provided by the present application, 1) energy of source nodes is saved: each source node only transmits compressed data; 2) high reliability: data erasure problems caused by faults or unreliability of wireless networks can be tolerated; 3) non-plaintext data transmission: compression of the source node on the perception data is essentially a lightweight encryption, and even if the compressed data is obtained, the perception data cannot be obtained without knowing the corresponding modulus; 4) can tolerate a measurement error of unlimited size, and is more in line with the actual needs. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0021] Figure 1 A flowchart of a data aggregation method based on a remainder system provided by the embodiment of the present application;

[0022] Figure 2 A schematic diagram of a data aggregation system based on a remainder system provided by the embodiment of the present application;

[0023] Figure 3 A schematic diagram of a processing device provided by the embodiment of the present application. DETAILED DESCRIPTION

[0024] The technical solutions in the embodiments of the present application will be described clearly and completely in the following with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0025] First, the terms that can be used in this paper are described as follows:

[0026] The terms "include", "contain", "have", "possess" or other similar semantic descriptions should be interpreted as non-exclusive inclusion. For example: including a technical feature element (such as raw materials, components, ingredients, carriers, dosage forms, materials, sizes, parts, components, mechanisms, devices, steps, processes, methods, reaction conditions, processing conditions, parameters, algorithms, signals, data, products or articles, etc.), should be interpreted as not only including the explicitly listed technical feature element, but also including other technical feature elements not explicitly listed in the art.

[0027] A data aggregation method, system, device and storage medium based on a remainder system are described in detail below. The content not described in detail in the embodiments of the present application belongs to the prior art known to those skilled in the art. If no specific conditions are indicated in the embodiments of the present application, the conventional conditions or the conditions recommended by the manufacturer are used.

[0028] Embodiment one

[0029] The embodiments of the present application provide a data aggregation method based on a remainder system, which is a data aggregation scheme that meets energy saving, reliability, non-plaintext transmission and tolerance to measurement errors (an unlimited size and any small error) at the same time. The scheme can still ensure to obtain a reasonable approximation value of the true value of the measured physical quantity in the case of an unlimited size measurement error caused by failure or interference. As shown in the following formula (1), it mainly includes the following three steps: Figure 1

[0030] 1. Selection of moduli.

[0031] In the embodiments of the present application, positive integers are selected according to the number of source nodes to form the moduli corresponding to each source node. Specifically, let the number of source nodes be n, select n-1 positive integers that are pairwise co-prime, and let their set be , satisfy , , represents the function of taking the minimum value, is a known positive integer, and the modulus of the i-th source node is , , is a positive integer selected pairwise co-prime, specifically, is the greatest common divisor of the modulus and the modulus .

[0032] It is considered that some source nodes may fail or the unreliability of the wireless network in the transmission process may cause the sink node to be unable to access some perception data, which is considered to be erased. It is assumed that at most f perception data are erased, i.e., f source nodes work abnormally, and at most one unlimited size error exists. In order to meet the reliability and tolerance to unlimited size measurement errors, the true value of the measured physical quantity (measured physical data) needs to satisfy , satisfy , where , is a known positive integer, and the parameter quantity , represents the function of the least common multiple, is a set composed of the moduli corresponding to all source nodes, , ​the number of source nodes, represents a subset of bases selected from the set of bases, represents the number of source nodes that are working abnormally, represents a set of moduli of the source nodes, represents the number of elements in the set

[0033] As can be understood by those skilled in the art, according to the residue system, to tolerate t unrestricted size measurement errors, there are 2t, and here only one unrestricted size measurement error is considered, so there are 2, and essentially -2 can ensure that the approximate solution obtained is unique.

[0034] 2, compression of the sensed data.

[0035] In this part, each source node senses the measured physical data, and compresses the sensed data in combination with its own modulus, and then stores and transmits the compressed sensed data to the sink node.

[0036] In the embodiment of the application, in order to save storage resources, the sensed data will be compressed by the source node. Because of the existence of measurement errors, for the first source node, its sensed data is , is the measurement error of the first source node; the first source node compresses the sensed data in combination with its own modulus: , wherein, represents the compressed sensed data of the first source node, is a modulo function, and the symbols and combined together represent the meaning of congruence, is the modulus corresponding to the first source node.

[0037] 3, reconstruction of the sensed data.

[0038] ​​​​In the embodiment of the present application, after receiving the compressed sensing data of a part of the source nodes, the sink node selects the compressed sensing data of any source node as a reference to adjust the received compressed sensing data set, constructs a suspect set combining the intermediate information obtained in the adjustment process, and constructs a new compressed sensing data set based on the suspect set. The sink node selects the compressed sensing data of any source node as a reference to adjust the new compressed sensing data set, and reconstructs the sensing data combining the Chinese remainder theorem if the adjusted compressed sensing data set meets the requirements.

[0039] In this part, it is assumed that f source nodes are abnormal, and the sink node receives the compressed sensing data of k source nodes, wherein, , Without loss of generality, it is assumed that the first f remainders are abnormal, and the compressed sensing data of the k source nodes are denoted as a set S = {s1, s2, …, sk}, and the moduli of the k source nodes are denoted as a set M = {m1, m2, …, mk}. , wherein, , , , , .

[0040] In the embodiment of the present application, there is an unlimited size measurement error in a source node in the k source nodes, and the rest are small measurement errors. The definition of the size of the measurement error is as follows: is a small measurement error, otherwise it is an unlimited size measurement error, wherein represents taking an absolute value.

[0041] The reconstruction process of the sensing data can be described as follows:

[0042] 1) The sink node first randomly selects a remainder s as a reference to adjust, and obtains . According to the adjustment process, a suspect set S is constructed.

[0043] 2) The sink node executes a While loop:

[0044] While (S ){

[0045] i. The sink node randomly selects ;

[0046] ii. If , the sink node selects a new ;

[0047] Otherwise, the sink node selects a new ;

[0048] ​​iii. Let ; ;

[0049] iv. The aggregation node is based on As a reference Adjustments were made to obtain ;

[0050] v. if It is consistent, without loss of generality, and is assumed to be ,according to The following system of congruence equations can be obtained:

[0051]

[0052] Solving using the generalized Chinese Remainder Theorem yields the following result. ,in , , ( ), , , A function representing the least common multiple. .if If the loop ends, proceed to vi; otherwise, proceed to vi.

[0053] vi. ;

[0054] }

[0055] 3) The aggregation node obtains the sensing data. ,so ,Right now for A reasonable approximation.

[0056] The specific implementation method is as follows:

[0057] 1) Select the aggregation node. Compressed sensing data from each source node As a reference set Make adjustments to obtain the adjusted data. Based on the intermediate information obtained during the adjustment process, using sets Constructing a set of doubts from compressed sensing data and using a set of doubts Constructing a new compressed sensing dataset (i.e., from) Subtract Get an element From the new compressed sensing dataset Compressed sensing data from any source node in the data. As a reference, the new compressed sensing dataset Make adjustments to obtain the adjusted data. ,

[0058] Aggregation node selection Compressed sensing data from each source node As a reference set Make adjustments to obtain the adjusted data. The steps are as follows:

[0059] calculate:

[0060]

[0061] in, Given a positive integer, For the remainder function, Indicates the first Compressed sensing data from each source node.

[0062] Then, calculate:

[0063]

[0064] ,

[0065] in, Indicates rounding up. , , These are all intermediate parameters.

[0066] Recalculate:

[0067]

[0068] in, For the adjusted number Compressed sensing data from each source node.

[0069] The final adjusted Compressed sensing data from each source node To obtain the adjusted data .

[0070] According to the adjustment process, there are: sets , represents the set of source node pairs that do not meet the consistency requirement; set This indicates dissatisfaction. A set of node pairs; where, For the adjusted number Compressed sensing data from each source node ; For the first Modulus of each source node and the Modulus of each source node The greatest common factor;

[0071] set up: , Represents the set of source node pairs that do not meet the consistency requirement. The obtained set of doubts, Indicates that the condition is not satisfied. The set of node pairs The obtained set of doubts; if the set If it is not empty, then the set is suspected. If set empty and set If it is not empty, then the set is suspected. ;

[0072] Compressed sensing data from any source node selected from the set of suspected nodes. ,if Then select a new source node for compressed sensing data. Otherwise, select compressed sensing data from a different source node. ;

[0073] Constructing a new compressed sensing dataset , ,in, Represents the new set of moduli. For the first The modulus of each source node.

[0074] The principle is as follows: the doubt set contains perceived data with unlimited measurement errors, so multiple attempts are needed, i.e., a While loop. If the selected data happens to be perceived data with unlimited measurement errors, then the remaining data are all perceived data with small errors, naturally yielding an approximate result. If not, it means this is perceived data with small errors, so it needs to be removed from the doubt set. Simultaneously, a new data point (if not empty) is selected from the doubt set that might contain perceived data with unlimited measurement errors. As the number of loops increases, the doubt set becomes smaller, so it's necessary to determine the relationship between A and the doubt set to further decide how to select a new data point.

[0075] Based on the compressed sensing data of the newly selected source nodes As a reference, the new compressed sensing dataset Make adjustments to obtain the adjusted data. This part also uses the four formulas provided above for calculation, resulting in adjusted data. wherein, represents the compressed sensing data of the adjusted th source node, , represents the number of compressed sensing data in the compressed sensing data set 2) if the data satisfies the consistency requirement, a congruence equation set is constructed using the data and the solution of the congruence equation set is solved using the generalized Chinese remainder theorem , that is, the reconstructed sensing data, specifically:

[0076] The congruence equation set is constructed as follows:

[0077]

[0078] wherein, represents the modulus of the th source node.

[0079] According to the generalized Chinese remainder theorem, the solution is obtained, if belongs to the set range, that is, , , that is, is a reasonable approximation of , and is the reconstructed sensing data. If does not belong to the set range, it enters the loop stage, that is, a new compressed sensing data set is constructed again using the suspect set (that is, from subtracts another element in to obtain ), and the compressed sensing data of any source node is selected as a reference to adjust the new compressed sensing data set to obtain the adjusted data , if the data satisfies the consistency requirement, a congruence equation set is constructed using the data and the solution of the congruence equation set is solved using the generalized Chinese remainder theorem; if the solution belongs to the set range, the solution is the reconstructed sensing data, otherwise, it continues to enter the loop stage until the obtained solution belongs to the set range.

[0080] The above part of the reconstruction of the sensing data of the sink node uses a residue system to realize, and the principle of a residue number system with nonpairwise-prime moduli (RNS-NPM) is introduced below.

[0081] For the following congruence system:

[0082] (1)

[0083] Let where all elements are pairwise coprime and greater than 1, and let Let modulus so , . Where represents the function of the least common multiple, represents the function of the greatest common divisor.

[0084] Let R=( , ,…, ), according to the generalized Chinese remainder theorem, if R is consistent, that is that is divides , then the above equation system has a unique solution .

[0085] When each remainder (in (1) is the measurement value of each source node in the WSN) has no error, the original value can be recovered; however, since the source nodes are independently measured, independent measurement errors will be introduced. At this time, the remainder (compressed sensing data) is . Let ( , ,…, ), and the positive integer satisfies , represents the function of taking the minimum value. Call a small error, and vice versa, where represents the absolute value. Assuming that there is an unlimited size measurement error and any number of small errors, the recovered value may have a large error compared to the original value or the equation system may have no solution at all, so a remainder is selected as a reference to adjust ( , ,…, ), and the adjusted value is denoted as , the process is shown as follows:

[0086] Without loss of generality, the first remainder, that is , is taken as the reference for calculation:

[0087] Calculation:

[0088] (3)

[0089] Calculation according to formula (3):

[0090] Indicates rounding up

[0091] , (4)

[0092] Calculate the new remainder according to formula (4):

[0093] (5)

[0094] Get the new congruence equation system according to formula (5):

[0095] (6)

[0096] According to the adjustment process, there are , , Let .

[0097] Let the suspect set be S, and the default suspect set S= If is not empty, then S= ; if is empty and is not empty, set S= . According to S, combined with the adjustment process (the same as the way introduced above) and the generalized Chinese remainder theorem, can be uniquely solved such that . We call that satisfies the reasonable approximation of

[0098] The following provides three examples based on the above scheme of the present application.

[0099] Example 1.

[0100] In this example, assume that the suspect set , and the main process is as follows:

[0101] 1) Selection of modulus.

[0102] Assume that the measured physical quantity is , and there are source nodes, , . According to and , take , then , 799459, , , , so .

[0103] 2) Compression of the perception data.

[0104] Without loss of generality, assume that the measurement error of the 4th source node is an unbounded large measurement error, the measurement error is , , , , so 1886256, .

[0105] 3) Reconstruction of the perception data.

[0106] Without loss of generality, assume that the data of the 5th source node is erased, so the sink node receives ( ), assume that the sink node chooses as the reference to adjust, and obtains

[0107]

[0108]

[0109]

[0110] According to the above information, the sink node calculates to obtain:

[0111]

[0112]

[0113]

[0114]

[0115] After that, the sink node calculates to obtain:

[0116]

[0117]

[0118]

[0119]

[0120] Finally, the sink node computes ( ).

[0121] Since the suspect set in this example, if the errors are all small errors, then is always true, so according to the backstepping method, if is satisfied, it is possible that there is an unlimited size error. According to the foregoing calculation process, we know that , , , , so the sink node constructs the suspect set as .

[0122] The sink node constructs , . Then, take , and suppose is , that is, adjust with as a reference to obtain

[0123]

[0124]

[0125] According to the foregoing information, the sink node computes:

[0126]

[0127]

[0128]

[0129] Then, the sink node computes:

[0130]

[0131]

[0132]

[0133] Finally, the sink node computes ( ).

[0134] It is easy to know that is consistent, so according to the generalized Chinese remainder theorem, the solution is = . Therefore, the sink node obtains the perception data Obviously, is a reasonable approximation of .

[0135] In addition, the effect of data aggregation can also be judged by calculating the energy saving factor. The energy saving factor ERF calculation formula is:

[0136]

[0137] wherein, is the total number of source nodes, is the number of bits of the true value of the measured physical quantity, is the total number of bits sent by the source node after data aggregation. The number of bits is the number of binary bits of the value, for example, the number of bits of 8 is 4, because 8 is written in binary as 1000, which has 4 bits.

[0138] Example 2.

[0139] In this example, it is assumed that the suspect set The main process is as follows:

[0140] 1) Selection of modulus.

[0141] It is assumed that the measured physical quantity is , and there are source nodes, , According to and , take , then , 799459, , , , So .

[0142] 2) Compression of perception data.

[0143] Without loss of generality, it is assumed that the measurement error of the 4th source node is an unlimited size measurement error, and the measurement error is , , , , So 1913282, .

[0144] 3) Reconstruction of perception data.

[0145] Without loss of generality, it is assumed that the data of the 5th source node is erased, so the sink node receives ​( ), assuming the sink node picks as the reference to adjust, we have

[0146]

[0147]

[0148]

[0149] According to the above information, the sink node calculates

[0150]

[0151]

[0152]

[0153]

[0154] After that, the sink node calculates

[0155]

[0156]

[0157]

[0158]

[0159] Finally, the sink node calculates ( ).

[0160] In this example, the suspect set is If all the errors are small errors, then is always true. Therefore, according to the backstepping method, if is satisfied, it is possible that there is an unlimited size error. According to the aforementioned calculation process, we have , , , , so the sink node constructs the suspect set as .

[0161] Without loss of generality, assume that the sink node constructs , . Then, take , and without loss of generality, assume that is , that is, adjust with as the reference, we have

[0162]

[0163]

[0164] According to the above information, the sink node calculates:

[0165]

[0166]

[0167]

[0168] After that, the sink node calculates:

[0169]

[0170]

[0171]

[0172] Finally, the sink node calculates ( ).

[0173] It is easy to know that is consistent, according to the generalized Chinese remainder theorem, the solution is = . So , continue to calculate.

[0174] The sink node constructs , . After that, take , suppose is , that is, adjust with as the reference, get:

[0175]

[0176]

[0177] According to the above information, the sink node calculates:

[0178]

[0179]

[0180]

[0181] After that, the sink node calculates:

[0182]

[0183]

[0184]

[0185] Finally, the sink node computes ( ).

[0186] It is easy to see that are consistent, so according to the generalized Chinese remainder theorem, we have = . So the sink node gets the sensing data = . Obviously, is a reasonable approximation of .

[0187] Similarly, according to the energy saving factor ERF, we have

[0188]

[0189] Example Three.

[0190] In this example, assume that the suspect set , the main process is as follows:

[0191] 1) The selection of modulus.

[0192] Assume that the measured physical quantity is , there are source nodes, , . According to and , take , then , 799459, , , , , so .

[0193] 2) Compression of sensing data.

[0194] Without loss of generality, assume that the measurement error of the fourth source node is an unlimited size measurement error, and the measurement error is , , , , , so 1886245, .

[0195] 3) Reconstruction of the sensing data.

[0196] Without loss of generality, assume that the data of the 5th source node is erased, thus the sink node receives ( ), assume that the sink node chooses as the reference to adjust, we get

[0197]

[0198]

[0199]

[0200] According to the above information, the sink node calculates

[0201]

[0202]

[0203]

[0204]

[0205] After that, the sink node calculates

[0206]

[0207]

[0208]

[0209]

[0210] Finally, the sink node calculates ( ).

[0211] It is easy to know that the sink node constructs the suspicion set as .

[0212] Without loss of generality, the sink node constructs , . After that, take , and suppose is , that is, adjust with as the reference, we get

[0213]

[0214]

[0215] According to the above information, the sink node calculates:

[0216]

[0217]

[0218]

[0219] Then, the sink node calculates:

[0220]

[0221]

[0222]

[0223] Finally, the sink node calculates ( ).

[0224] It is known that is consistent, so according to the generalized Chinese remainder theorem, the solution is = . Therefore, the sink node obtains the perception data = . Obviously, is a reasonable approximation of .

[0225] Similarly, according to the energy saving coefficient ERF, we have:

[0226]

[0227] The above scheme provided by the embodiment of the application mainly obtains the following beneficial effects:

[0228] 1) Energy saving: each source node only transmits compressed data.

[0229] 2) Reliability: can tolerate data erasure problems caused by faults or wireless network unreliability.

[0230] 3) Non-plaintext data transmission: the compression of the source node on the perception data is essentially a lightweight encryption, and even if the compressed data is obtained, the perception data cannot be obtained without knowing the corresponding modulus.

[0231] 4) Can tolerate a measurement error of unlimited size, more in line with the needs of reality.

[0232] Those skilled in the art can clearly understand that the above-mentioned embodiments can be implemented by software, or can be implemented by means of software and necessary universal hardware platforms based on the description of the above embodiments. Based on such understanding, the technical solutions of the above-mentioned embodiments can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a U disk, a mobile hard disk, etc.), and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments of the present application.

[0233] Embodiment two

[0234] The present application also provides a data aggregation system based on a remainder system, which is mainly realized based on the method provided in the above-mentioned embodiments, as shown in the following figure, which mainly includes: Figure 2

[0235] A modulus selection part selects positive integers according to the number of source nodes to form the modulus corresponding to each source node.

[0236] A compressed sensing data storage part senses the measured physical data by each source node, compresses the sensing data in combination with the modulus of the source node, and then stores and transmits the compressed sensing data to the sink node.

[0237] A compressed sensing data reconstruction part receives the compressed sensing data of a part of the source nodes by the sink node, wherein there is an unlimited size measurement error in a certain sensing node, the sink node selects the compressed sensing data of any source node as a reference to adjust the received compressed sensing data set, constructs a suspect set in combination with the intermediate information obtained in the adjustment process, and constructs a new compressed sensing data set based on the suspect set, selects the compressed sensing data of any source node in the new compressed sensing data set as a reference to adjust the new compressed sensing data set, and reconstructs the sensing data in combination with the Chinese remainder theorem if the adjusted compressed sensing data set meets the requirements.

[0238] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional modules is exemplified, and in actual application, the above-mentioned functions can be completed by different functional modules according to needs, that is, the internal structure of the system is divided into different functional modules to complete all or part of the functions described above.

[0239] Embodiment three

[0240] The present application also provides a processing device, as shown in the following figure: Figure 3 ​As shown, it mainly comprises: one or more processors; a memory for storing one or more programs; wherein when the one or more programs are executed by the one or more processors, the one or more processors realize the method provided by the foregoing embodiments.

[0241] Further, the processing device further comprises at least one input device and at least one output device; in the processing device, the processor, the memory, the input device and the output device are connected through a bus.

[0242] In the embodiments of the present application, the specific types of the memory, the input device and the output device are not limited; for example:

[0243] The input device can be a touch screen, an image acquisition device, a physical key or a mouse, etc.

[0244] The output device can be a display terminal.

[0245] The memory can be a random access memory (RAM), and can also be a non-volatile memory, for example, a disk memory.

[0246] Embodiment four

[0247] The present application also provides a readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the method provided by the foregoing embodiments is realized.

[0248] In the embodiments of the present application, the readable storage medium as the computer readable storage medium can be arranged in the foregoing processing device, for example, as the memory in the processing device. In addition, the readable storage medium can also be a U disk, a mobile hard disk, a read-only memory (ROM), a magnetic disk or an optical disk and various media which can store program codes.

[0249] The above is only the preferred specific implementation of the present application, but the protection scope of the present application is not limited to this, any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method of data aggregation based on a residue system, characterized in that, include: Select positive integers based on the number of source nodes to form the modulus corresponding to each source node; Each source node senses the measured physical data, compresses the sensed data in combination with its own modulus, and then stores and transmits the compressed sensed data to the aggregation node. The sink node receives a portion of the compressed sensing data from the source nodes. Among them, one of the sensing nodes has an unlimited measurement error. The sink node selects the compressed sensing data of any source node as a reference to adjust the received compressed sensing data set. It constructs a doubt set based on the intermediate information obtained in the adjustment process, and uses this to construct a new compressed sensing data set. From the new compressed sensing data set, it selects the compressed sensing data of any source node as a reference to adjust the new compressed sensing data set. If the adjusted compressed sensing data set meets the requirements, the sensing data is reconstructed using the Chinese Remainder Theorem.

2. The method of claim 1, wherein, The step of selecting positive integers based on the number of source nodes to form the modulus corresponding to each source node includes: Let the number of source nodes be n, select two relatively prime positive integers, let their set be , satisfy , is a known positive integer, the source node corresponding to the modulus , is selected two relatively prime positive integers.

3. The method of claim 1, wherein, The measured physical data is denoted as , satisfying , wherein, is a known positive integer, and the parameter quantity , represents a function of the least common multiple, is a set composed of moduli corresponding to all source nodes, , is the number of source nodes, represents the number of source nodes working abnormally, represents a set composed of moduli of source nodes, represents the number of elements in the set .

4. The data aggregation method based on the residue system according to claim 1 or 3, characterized in that, Each source node senses the measured physical data and compresses the sensed data by combining its own modulus, including: Let the measured physical data be For the first source node, its perception data is , is the measurement error of the first source node; The first source node compresses the perception data in combination with its own modality: ； wherein, represents the compressed sensing data of the th source node, is a modulo function, and the symbols and together represent the meaning of congruence, is the modulus corresponding to the th source node, is the measurement value of the th source node.

5. The method of claim 1, wherein, After receiving compressed sensing data from a portion of the source nodes, the aggregation node selects compressed sensing data from any one source node as a reference to adjust the received compressed sensing data set. It then constructs a doubt set based on the intermediate information obtained during the adjustment process, and uses this to construct a new compressed sensing data set. From this new set, it selects compressed sensing data from any one source node as a reference to adjust the new set. If the adjusted compressed sensing data set meets the requirements, the sensing data is reconstructed using the Chinese Remainder Theorem, including: The number of source nodes is n, and when f source nodes are abnormal, the sink node receives compressed sensing data of k source nodes, wherein, , ; Let the compressed sensing data from k source nodes be denoted as set. The modulus of the k source nodes is denoted as the set. ,in, , They represent the first Compressed sensing data and modulus of each source node ; The sink node selects compressed sensing data of the first source node Adjust the reference set of data , obtain adjusted data , according to the intermediate information obtained in the adjustment process, use the compressed sensing data in the set to construct a suspect set , and use the suspect set to construct a new compressed sensing data set ; select the compressed sensing data of any source node from the new compressed sensing data set as the reference set of the new compressed sensing data set Adjust the data , obtain adjusted data , if the data satisfy the consistency requirement, use the data to construct a congruence equation set, and use the Chinese remainder theorem to solve the solution of the congruence equation set , if belongs to the set range, then the solution is the reconstructed sensing data.

6. The method of claim 5, wherein, The sink node selects compressed sensing data of the first source node The sink node selects compressed sensing data of the first source node Adjust the reference set according to the adjustment process to obtain adjusted data Adjust the reference set according to the adjustment process to obtain adjusted data According to the intermediate information obtained in the adjustment process, the compressed sensing data in the reference set is used to construct a suspect set According to the intermediate information obtained in the adjustment process, the compressed sensing data in the reference set is used to construct a suspect set According to the intermediate information obtained in the adjustment process, the compressed sensing data in the reference set is used to construct a suspect set According to the intermediate information obtained in the adjustment process, the compressed sensing data in the reference set is used to construct a suspect set According to the intermediate information obtained in the adjustment process, the compressed sensing data in the reference set is used to construct a suspect set According to the intermediate information obtained in the adjustment process, the compressed sensing data in the reference set is used to construct a suspect set According to the intermediate information obtained in the adjustment process, the compressed sensing data in the reference set is used to construct a suspect set According to the intermediate information obtained in the adjustment process, the compressed sensing data in the reference set is used to construct a suspect set According to the intermediate information obtained in the adjustment process, the compressed sensing data in the reference set is used to The sink node selects the compressed sensing data of the first source node The reference set is adjusted to obtain adjusted data The adjustment is performed as follows: ​ calculate: ； wherein, is a known positive integer, is a remainder function, denotes the compressed sensing data of the th source node; Then, calculate: ； ， ； wherein denotes the ceiling function, , , are all intermediate parameters; Recalculate: ； wherein, the adjusted compressed sensing data of the i th source node; The adjusted compressed sensing data of the source node , obtaining the adjusted data ; According to the adjustment process: set , representing a set of source node pairs that do not meet the consistency requirement; set , representing a set of node pairs that do not meet , represents taking the absolute value; wherein, is the compressed sensing data of the adjusted th source node, , is the modulus of the th source node and the modulus of the th source node; and the greatest common divisor of the modulus​ Set: , denotes the set of node pairs according to source nodes not satisfying the consistency requirement obtained suspicion set, denotes the set of node pairs according to source nodes not satisfying the consistency requirement obtained suspicion set; if the set is not empty, the suspicion set is if the set is empty and the set is not empty, the suspicion set is selecting a compressed sensing data of any one source node from the suspicion set if , reselecting a compressed sensing data of one source node , else, reselecting a compressed sensing data of one source node ; Constructing new compressed sensing data sets , , wherein, denotes a new set of moduli, is a moduli of the th source node; compressed sensing data of the selected new source node as a reference to the new compressed sensing data set adjustment, obtaining adjusted data .

7. The method of claim 6, wherein, Utilizing data constructing a system of congruences, and solving the system of congruences using the Chinese remainder theorem That is, the reconstructed perception data includes: adjusted data wherein, denotes the adjusted data of the i-th source node, , denotes the number of compressed sensing data in the set of compressed sensing data ; Construct a system of congruence equations: ； wherein, denotes the modulo of the th source node; According to the Chinese remainder theorem, the solution is obtained as where , , , , , denotes the function of the least common multiple, ; If belongs to the set range, the solution is the reconstructed sensing data; if does not belong to the set range, it enters a loop phase, i.e. a new compressed sensing data set is constructed again using the suspicious set , and the compressed sensing data of any source node is selected from the new compressed sensing data set as a reference to adjust the new compressed sensing data set , to obtain adjusted data . If the data satisfies the consistency requirement, the data is used to construct a congruence equation set, and the Chinese remainder theorem is used to solve the solution of the congruence equation set; if the solution belongs to the set range, the solution is the reconstructed sensing data, otherwise, it continues to enter the loop phase until the solution belongs to the set range.

8. A data aggregation system based on a residue system, characterized in that Based on the method described in any one of claims 1 to 7, the system comprises: In the modulus selection part, positive integers are selected based on the number of source nodes to form the modulus corresponding to each source node; In the sensing data compression and storage section, each source node senses the measured physical data, compresses the sensed data in combination with its own modulus, and then stores and transmits the compressed sensed data to the aggregation node. In the sensing data reconstruction section, the aggregation node receives a portion of compressed sensing data from the source nodes. Among these, one sensing node has an unlimited measurement error. The aggregation node selects compressed sensing data from any source node as a reference to adjust the received compressed sensing data set. It constructs a doubt set based on the intermediate information obtained during the adjustment process, and uses this set to construct a new compressed sensing data set. From the new compressed sensing data set, it selects compressed sensing data from any source node as a reference to adjust the new compressed sensing data set. If the adjusted compressed sensing data set meets the requirements, the sensing data is reconstructed using the Chinese Remainder Theorem.

9. A processing device, characterized by include: One or more processors; Memory, used to store one or more programs; Wherein, when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method as described in any one of claims 1 to 7.

10. A readable storage medium, storing a computer program, characterized in that, When a computer program is executed by a processor, it implements the method as described in any one of claims 1 to 7.

Images