A privacy-protecting electronic data evidence storage method

Abstract

The present invention relates to the field of information technology, and more particularly to a privacy-preserving electronic data evidence storage method, comprising: assigning node serial numbers between nodes, dividing electronic data into a plurality of sub-segments, and assigning sub-segment serial numbers; extracting a sub-segment hash value of each sub-segment; the node establishing a fitting function through privacy-preserving calculations, and storing the electronic data in association with the fitting function; when verifying the electronic data, requesting a plurality of nodes to assist in verification, and sending the fitting function stored by the node to the nodes participating in the verification; the plurality of nodes participating in the verification each extract a hash value of each sub-segment to obtain a sub-segment hash value; and if the fitting errors are all less than a preset threshold, determining that the electronic data has not been modified. The substantial effect of the present invention is: establishing a unified fitting function between multiple nodes, enabling mutual verification of the authenticity of data, and establishing low-cost electronic data evidence storage.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a privacy-protecting electronic data evidence storage method. Background Art

[0002] The Internet of Things (IoT), or "Internet of Everything," is an extension and expansion of the Internet. It combines various information sensing devices with the internet to form a vast network, enabling the interconnection of people, machines, and objects at any time and anywhere. The IoT is a crucial component of next-generation information technology. The vast number of sensors and monitoring devices integrated into the IoT enables multi-dimensional monitoring of regional data, generating a wealth of data. This provides a foundation for big data applications. Big data applications often utilize monitoring data collected by multiple entities connected to the IoT, necessitating verification and validation of the data's authenticity to prevent errors and anomalies in big data applications caused by erroneous and falsified data. Since IoT sensors and monitoring devices continuously generate large amounts of monitoring data, the use of the currently widely used blockchain evidence storage technology would place significant pressure on the blockchain network. Furthermore, the capital required for blockchain evidence storage would place additional financial pressure on IoT entities. Therefore, it is necessary to develop an evidence storage technology suitable for IoT monitoring data.

[0003] For example, Chinese patent CN107659579A, published on February 2, 2018, discloses an on-site evidence storage method, device, and related evidence storage system. The method includes: receiving an evidence preservation application initiated by a client business system located at the same site; obtaining and storing the data to be preserved contained in the evidence preservation application; calculating a hash value for the data to be preserved, and sending the calculated hash value to an electronic data preservation center on the network side. Its technical solution provides evidence storage for business systems with high requirements for original data confidentiality or business systems with large original file volumes. However, the authenticity of its evidence storage is centralized at the electronic data preservation center, which poses a centralization risk and lacks high credibility. Summary of the Invention

[0004] The present invention aims to address the current lack of efficient and low-cost electronic data storage solutions. This paper proposes a privacy-preserving electronic data storage method that can establish electronic data storage between multiple data nodes, offering the advantages of low latency and low cost.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: a privacy-protecting electronic data notarization method, including: assigning node numbers between nodes, multiple nodes dividing their respective electronic data into several sub-segments and assigning sub-segment numbers; extracting the hash value of each sub-segment and recording it as the sub-segment hash value; the node establishes a fitting function through privacy-protecting calculation: sub-segment hash value = f(sub-segment number, node number), the fitting error of the fitting function is less than a preset threshold, and the electronic data is associated with the fitting function and stored; when verifying the electronic data, several nodes are requested to assist in verification, and the fitting function stored by the node is sent to the nodes participating in the verification; several nodes participating in the verification each extract the hash value of each sub-segment to obtain the sub-segment hash value; the nodes participating in the verification each calculate the fitting error. If the fitting errors are all less than the preset threshold, it is determined that the electronic data has not been modified. Otherwise, if the fitting errors are all greater than the preset threshold, it is determined that the electronic data has been modified.

[0006] Preferably, the fitting function is a polynomial, and the coefficients of the polynomial constitute a coefficient vector. After the fitting function is established, the node associates and stores the electronic data with the coefficient vector.

[0007] Preferably, the method for a node to establish a fitting function through privacy-preserving calculation includes: selecting any node as a service node, the service node randomly generating the highest-order term m of a polynomial, where the value of m is less than or equal to the number of nodes; randomly generating the term coefficient aij of the polynomial, where the expression of the polynomial is: sub-segment hash value = ∑∑aij*sub-segment sequence number ^i* Node serial number ^j, i, j∈[0, m]; establish a coefficient vector and send the coefficient vector to all nodes; the node calculates the difference between the polynomial fitting node serial number under the coefficient vector and its own serial number. If the difference is less than the preset second threshold, the square of the difference is split into several addends, and the addends are randomly sent to several nodes; the node sums up all the addends received and sends them to the service node; the service node sums up the sums sent by the node again to obtain the fitting error of the fitting function; adjust the item coefficient, send the coefficient vector to all nodes again, and the node feeds back the square of the difference to the service node again to obtain the fitting error; if the fitting error decreases, continue to adjust the item coefficients of the polynomial in the direction of the last adjustment. If the fitting error increases, continue to adjust the item coefficients of the polynomial in the opposite direction of the last adjustment until the fitting error is less than the preset threshold; send the final coefficient vector to all nodes.

[0008] Preferably, the fitting function is: the last N bits of the sub-segment hash value = f(sub-segment number, node number).

[0009] Preferably, the nodes divide their respective electronic data into several sub-segments and assign original sub-segment serial numbers in sequence; extract the sub-segment hash values, arrange the sub-segment hash values in ascending order, assign evidence sub-segment serial numbers in the ascending order, associate the original sub-segment serial numbers with the corresponding evidence sub-segment serial numbers and store them; use the evidence sub-segment serial numbers as sub-segment serial numbers to participate in the establishment and verification of the fitting function.

[0010] The substantial effects of the present invention are: establishing a unified fitting function between multiple nodes, multiple nodes have relative independence, can mutually verify the authenticity of the data, and at the same time form constraints on data tampering with each other, thereby establishing low-cost electronic data evidence; a large amount of monitoring data generated in the Internet of Things, the loss or error of a small amount of data in these monitoring data will not have a significant impact on the application of the data, so this solution is suitable for the evidence of Internet of Things monitoring data; the fitting function is established through privacy-preserving calculations, and there is no need to expose the sub-segment hash value of the node, so that the fitting function must be established with the participation of all nodes, and it is difficult for a single node to tamper with the fitting function without causing the fitting error of other nodes to exceed the fitting error threshold, forming a more effective evidence. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 This is a flow chart of the electronic data evidence storage method according to Example 1.

[0012] Figure 2 This is a flow chart of the method for establishing a fitting function in Example 1.

[0013] Figure 3 This is a schematic diagram of electronic data evidence storage in Example 1.

[0014] Figure 4 This is a schematic diagram of adjusting the sub-segment sequence number in Example 1.

[0015] Where: 10, node, 20, addend. DETAILED DESCRIPTION

[0016] The specific implementation of the present invention will be further described below with reference to specific embodiments and in conjunction with the accompanying drawings.

[0017] Example 1:

[0018] A privacy-protecting electronic data evidence storage method, please see the attached Figure 1, including: step A01) assigning node numbers between nodes 1010, and multiple nodes 1010 divide their respective electronic data into several sub-segments and assign sub-segment numbers; step A02) extracting the hash value of each sub-segment and recording it as the sub-segment hash value; step A03) node 1010 establishes a fitting function through privacy protection calculation: sub-segment hash value = f(sub-segment number, node number), the fitting error of the fitting function is less than a preset threshold, and the electronic data is associated with the fitting function and stored; step A04) when verifying the electronic data, requesting several nodes 10 to assist in the verification, and sending the fitting function stored by the node 10 to the nodes 10 participating in the verification; step A05) several nodes 10 participating in the verification each extract the hash value of each sub-segment to obtain the sub-segment hash value; step A06) the nodes 10 participating in the verification each calculate the fitting error. If the fitting errors are all less than the preset threshold, it is determined that the electronic data has not been modified. On the contrary, if the fitting errors are all greater than the preset threshold, it is determined that the electronic data has been modified. The fitting function is a polynomial, and the coefficients of the polynomial form a coefficient vector. After the fitting function is established, the node 10 associates and stores the electronic data with the coefficient vector. In theory, a polynomial can fit any function.

[0019] For the method of establishing the fitting function by using privacy-preserving calculations in node 10, please refer to the attached Figure 2, including: step B01) selecting any node 10 as a service node, the service node randomly generates the highest order term m of the polynomial, the value of m is less than or equal to the number of nodes 10; step B02) randomly generates the term coefficient aij of the polynomial, the expression of the polynomial is: sub-segment hash value = ∑∑aij*sub-segment sequence number ^i*node sequence number ^j, i, j∈[0,m]; step B03) establishing a coefficient vector, and sending the coefficient vector to all nodes 10; step B04) node 10 calculates the difference between the sequence number of the polynomial fitting node under the coefficient vector and its own sequence number. If the difference is less than a preset second threshold, the square of the difference is split into several addends 20, and the addends 20 are randomly sent to several nodes 10; step B05) node 10 sums up all the addends 20 received and sends them to the service node; step B06) the service node re-sums the sum sent by node 10 The fitting error of the fitting function is obtained by summing the terms. Step B07) The term coefficients are adjusted and the coefficient vector is sent to all nodes 10 again. Node 10 again feeds back the square of the difference to the service node to obtain the fitting error. Step B08) If the fitting error decreases, the term coefficients of the polynomial are adjusted in the direction of the previous adjustment. Step B09) If the fitting error increases, the term coefficients of the polynomial are adjusted in the opposite direction of the previous adjustment until the fitting error is less than a preset threshold. Step B10) The final coefficient vector is sent to all nodes 10. The exponent value of the highest-order term of the polynomial minus 1 represents the number of changes in the fitting function attribute. In this embodiment, the subsegment hash value = -5*(subsegment number + node number)^6 + 260*(subsegment number + node number)^5 - 12*(subsegment number + node number)^4 + 61*(subsegment number + node number)^2 - 161*(subsegment number + node number) + 154. If the sum of the subsegment and node numbers is considered the independent variable, the fitting function changes five times. That is, the fitting function first increases monotonically, then switches to decreasing monotonically, repeating this five times before finally returning to a monotonically decreasing state. There are five extreme points, the locations of which are determined by the term coefficients. Expanding the fitting function and combining similar terms yields the final fitting function.

[0020] Please see the attached Figure 3, node 10 calculates the difference between the polynomial fitting node number under the coefficient vector and its own number. If the difference is less than the preset second threshold, the square of the difference is split into several addends 20. Node 10 will also receive the addends 20 sent by other nodes 10, and after summing the received addends 20, it will achieve mixing. This ensures that the fitting error of each node 10 is kept confidential, and further ensures that the hash value of each node 10 sub-segment is kept confidential. If the difference is greater than the preset second threshold, the difference is set to a preset larger value, and the larger value is split into several addends 20 and sent to several nodes 10. As an alternative, if the difference is greater than the preset second threshold, a preset message is broadcast, the calculation of the current round of fitting error is skipped, and the service node immediately adjusts the item coefficient.

[0021] In the improved implementation of this embodiment, the fitting function is: the last N bits of the subsegment hash value = f(subsegment number, node number). Using the last N bits simplifies the form of the fitting function. The last N bits of the subsegment hash value are included in the calculation of the fitting function. The gap between the node number, subsegment number, and the last N bits of the subsegment hash value is smaller, facilitating the establishment of the fitting function. When N is 10, the probability that the last N bits of the subsegment hash values are the same is 1 / 16^10, which is an extremely small probability.

[0022] Please see the attached Figure 4 The method for adjusting sub-segment numbers includes: Step C01) Node 10 divides its respective electronic data into several sub-segments and sequentially assigns original sub-segment numbers; Step C02) extracts sub-segment hash values, arranges the sub-segment hash values in ascending order, assigns evidence sub-segment numbers in ascending order, and associates the original sub-segment numbers with the corresponding evidence sub-segment numbers for storage; Step C03) uses the evidence sub-segment numbers as sub-segment numbers for establishing and verifying the fitting function. Arranging the sub-segment hash values in ascending order can reduce the number of attribute changes required by the fitting function, thereby simplifying the form of the fitting function.

[0023] The substantial effects of this embodiment are: establishing a unified fitting function between multiple nodes 10, multiple nodes 10 have relative independence, can mutually verify the authenticity of data, and at the same time form constraints on data tampering with each other, thereby establishing low-cost electronic data evidence; a large amount of monitoring data is generated in the Internet of Things, and the loss or error of a small amount of data in these monitoring data will not have a significant impact on the application of the data, so this solution is suitable for the evidence of Internet of Things monitoring data; the fitting function is established through privacy-preserving calculations, and there is no need to expose the sub-segment hash value of the node 10, so that the fitting function must be established with the participation of all nodes 10, and it is difficult for a single node 10 to tamper with the fitting function without causing the fitting error of other nodes 10 to exceed the fitting error threshold, forming a more effective evidence.

[0024] The embodiment described above is only a preferred solution of the present invention and does not limit the present invention in any form. Other variations and modifications are possible without exceeding the technical solution described in the claims.

Claims

1. A privacy-protecting electronic data evidence storage method, characterized in that: include: Node numbers are assigned between nodes, and multiple nodes divide their electronic data into several sub-segments and assign sub-segment numbers; Extract the hash value of each sub-segment and record it as the sub-segment hash value; The node establishes a fitting function through privacy-preserving calculations: subsegment hash value = f(subsegment number, node number), and the fitting error of the fitting function is less than a preset threshold. The electronic data is then associated with the fitting function and stored. When verifying electronic data, several nodes are requested to assist in the verification, and the fitting function stored in the node is sent to the nodes participating in the verification; Several nodes participating in the verification extract the hash value of each sub-segment to obtain the sub-segment hash value; The nodes participating in the verification each calculate the fitting error. If the fitting errors are all less than the preset threshold, it is determined that the electronic data has not been modified. Conversely, if the fitting errors are all greater than the preset threshold, it is determined that the electronic data has been modified. The fitting function is a polynomial, and the coefficients of the polynomial constitute a coefficient vector. After the fitting function is established, the node associates and stores the electronic data with the coefficient vector; Methods for nodes to establish fitting functions through privacy-preserving calculations include: Select any node as a service node, and the service node randomly generates the highest-order term m of a polynomial, where the value of m is less than or equal to the number of nodes; Randomly generate the coefficients aij of the polynomial. The expression of the polynomial is: sub-segment hash value = ∑∑aij*sub-segment sequence number ^i*node sequence number ^j, i, j∈[0,m]; Create a coefficient vector and send it to all nodes; The node calculates the difference between the polynomial fitting node number under the coefficient vector and its own number. If the difference is less than a preset second threshold, the square of the difference is split into several addends, and the addends are randomly sent to several nodes. The node sums up all the received addends and sends it to the service node; The service node sums the sums sent by the nodes again to obtain the fitting error of the fitting function; Adjust the item coefficients and send the coefficient vector to all nodes again. The nodes will feed back the square of the difference to the service node to obtain the fitting error. If the fitting error decreases, the polynomial coefficients are adjusted in the direction of the last adjustment. If the fitting error increases, the polynomial coefficients are adjusted in the opposite direction of the last adjustment until the fitting error is less than the preset threshold. Send the final coefficient vector to all nodes.

2. A privacy-protecting electronic data evidence storage method according to claim 1, characterized in that: The fitting function is: the last N bits of the sub-segment hash value = f (sub-segment number, node number).

3. The privacy-protecting electronic data evidence storage method according to claim 1, characterized in that: The nodes divide their electronic data into several sub-segments and assign original sub-segment numbers in sequence; Extract the sub-segment hash values, sort the sub-segment hash values in ascending order, assign evidence sub-segment serial numbers in ascending order, and associate the original sub-segment serial numbers with the corresponding evidence sub-segment serial numbers for storage; The stored sub-segment number is used as the sub-segment number to participate in the establishment and verification of the fitting function.

Images