Determination method and device for integrity of decentralized edge data

Abstract

The invention provides a method and device for determining the integrity of decentralized edge data, and the method comprises the steps: enabling an edge server to adaptively determine a target data item in a verifier according to the cache data states of other edge servers in a system of the edge server; the edge server performs batch verification on the target data item through the verifier; and the edge server determines the integrity of the target data item according to the batch verification result. According to the efficient decentralized edge data integrity scheme AccEDI provided by the invention, batch data verification can be realized by utilizing an RSA accumulator, the calculation, communication and storage costs in a mobile edge calculation environment are remarkably reduced, and meanwhile, the high integrity guarantee of data is ensured.

Description

Technical Field

[0001] The present invention relates to the field of computers, and in particular to a method and device for determining the integrity of decentralized edge data. Background Art

[0002] Mobile Edge Computing (MEC) has become an important extension of cloud computing and a key driver of 5G / 6G networks. MEC enables application vendors (AppVends) to provide low-latency services by deploying edge servers (ES) near users and caching data on them. However, this distributed data storage model brings challenges in maintaining edge data integrity (EDI). Specifically, the EDI problem refers to the problem that the data copy on the ES is inconsistent with the original data due to malicious events such as software failures, hardware failures, and network attacks. This jeopardizes AppVends service quality, data availability, and user security.

[0003] Existing solutions to the EDI problem can be roughly divided into two types: centralized and decentralized.

[0004] (1) Centralized solutions rely heavily on AppVends to perform EDI audits every time they interact with ES, which imposes a huge computing burden on AppVends and may lead to a single point of failure.

[0005] (2) Decentralized solutions mainly use the Merkle Hash Tree (MHT) method to leverage collaboration between ESs for EDI verification. In a decentralized MEC environment, each ES typically only caches a portion of AppVend's data set, resulting in different subsets of data being stored on various ESs. Therefore, when an ES wants to verify the integrity of its cached data, it must act as the initiator of the EDI and rely on the information of neighboring ESs to verify its data. This naturally requires an adaptive approach to adapt to the changing data distribution. However, since current decentralized solutions assume that each validator caches the same data items as the initiating ES, this makes the current solution lack the flexibility to adapt to the dynamic data distribution of the MEC environment. This conflicts with the adaptive requirements of the decentralized MEC environment, resulting in increased communication costs and longer verification times. Table 1 Decentralized EDI environment

[0006] To illustrate this problem, consider the decentralized scenario in Table 1. A set of edge servers S = {s 0 ,s 1 ,s2 ,s 3}Cache data set D = {d 1 ,d 2 ,d 3}. ESs work together to assist EDI initiators 0 Verify the integrity of its copy of the data without AppVend. In this decentralized MEC environment, each ES caches only a portion of the AppVend data D, with a different subset cached on each server. 1 Cache 1 ,d 2 ,d 3 , and s 2 Only cache d 1 ,d 2 ,s 3 Only cache d 3 .

[0007] Existing solutions rely on individual verification. After determining which data items are cached by other validators ES, the initiator s 0 Each data item must be validated separately. For example, it selects 1 and 2 To verify and 1 Then, it again chooses s 1 and 2 Right 2 Finally, it chooses s 1 and 3 Right 3 Vote. In total, s 0 Three voting events and six rounds of communication (each round consists of a request and a response) are required to verify all its data items. In contrast, the AccEDI proposed in this invention uses the aggregation characteristics of RSA accumulators to achieve efficient batch verification and adaptive data selection. In the same scenario, when the initiator s 0 When it wants to validate all its data items, it adaptively selects s 1 To batch verify and collect all its data items d 1 ,d 2 ,d 3 It also adaptively selects s 2 To batch verify 1 ,d 2 , finally select s 3 To verify 3 , and finally collect their votes on these data items. In total, s 0 Perform one voting event and three rounds of communication to verify all its data items d 1 ,d 2 ,d3 As shown in Table 2, this significant reduction in voting events and communication rounds demonstrates the efficiency of AccEDI in leveraging the decentralized nature of the MEC environment for EDI verification. Table 2 Decentralized EDI solution

[0008] In addition, while some MHT structures allow batch processing, they cannot precisely locate the location of corruption in the batch proof when corruption is detected. This makes it still a significant challenge to achieve corruption localization while effectively leveraging the decentralized nature of the MEC environment for efficient batch verification and adaptive data selection.

[0009] In general, existing decentralized EDI solutions face the following problems.

[0010] (1) There is a lack of a mechanism to adaptively select the data items stored by ES, which results in the inability to effectively utilize decentralized network resources and increases latency.

[0011] (2) It is unable to provide efficient batch processing capabilities and requires each data item to be verified individually, which significantly increases the computing resources and costs required for integrity verification.

[0012] (3) When using MHT batch processing proofs, it is impossible to accurately locate the location of damage. Summary of the invention

[0013] The present invention aims to provide a method for determining the integrity of decentralized edge data, which can solve the above technical problems. The AccEDI proposed in the present invention significantly reduces the computing, communication and storage costs while maintaining strong data integrity protection in the MEC environment.

[0014] According to one aspect of the present invention, a method for determining the integrity of decentralized edge data is provided, comprising: an edge server adaptively determining a target data item in a verifier based on the cache data status of other edge servers in its system; the edge server batch-verifying the target data item through the verifier; and the edge server determining the integrity of the target data item based on the result of the batch verification.

[0015] Preferably, the edge server adaptively determines the target data item in the verifier according to the cache data status of other edge servers in the system, including: the edge server broadcasts an invitation message in the system, wherein the invitation message includes the public key of the edge server, the identifier of the target data copy and the identifier of the target data item in the target data copy; the edge server receives a series of response messages from the other edge servers, wherein the response messages include the public keys of one or more edge servers in the other edge servers, the identifier of the target data copy and a shared data item sequence, wherein the shared data item sequence is calculated by the one or more edge servers based on the identifier of the target data item; and the edge server adaptively determines the verifier of the target data item according to the response message.

[0016] Preferably, before the edge server performs batch verification on the target data items through the verifier, it also includes: the edge server initializes a first RSA accumulator using a base g; the edge server identifies first shared data using the shared data item sequence; the edge server constructs the first RSA accumulator using a prime number mapped to the first shared data; and the edge server uses the RSA accumulator to calculate the verification proof of the verifier.

[0017] Preferably, the edge server performs batch verification on the target data item through the verifier, including: the edge server sends a sampling request to the verifier, wherein the sampling request includes the public key of the edge server, the identifier of the target data copy, the shared data item sequence and the verification proof; the verifier determines the second shared data according to the shared data item sequence; the verifier constructs a second RSA accumulator using the prime number mapped to the second shared data; the verifier calculates the proof using the second RSA accumulator; the verifier compares the calculated proof with the verification proof to obtain a voting result; and the verifier broadcasts the voting result in the system.

[0018] Preferably, the verifier uses the second RSA accumulator to calculate the proof according to the following formula: Wherein, A is the calculated proof, g is the base, H(di) is the hash value of the target data item di, n is the number of the target data items, and N is the RSA modulus.

[0019] Preferably, the edge server determines the integrity of the target data item according to the batch verification result, including: the edge server determines whether the number of true results among the multiple voting results exceeds half, and if the judgment result is yes, the edge server determines that the target data item is complete.

[0020] According to another aspect of the present invention, there is provided an apparatus for determining the integrity of decentralized edge data, comprising: an adaptive determination module for adaptively determining a target data item in a verifier according to the cache data status of other edge servers in the edge server system; a batch verification module for enabling the edge server to perform batch verification on the target data item through the verifier; and an integrity determination module for enabling the edge server to determine the integrity of the target data item according to the result of the batch verification.

[0021] Preferably, the adaptive determination module includes: an invitation message broadcasting unit, used to enable the edge server to broadcast an invitation message in the system, wherein the invitation message includes a public key of the edge server, an identifier of a target data copy, and an identifier of the target data item in the target data copy; a response message receiving unit, used to enable the edge server to receive a response message from the other edge servers, wherein the response message includes a public key of one or more edge servers among the other edge servers, an identifier of the target data copy, and a shared data item sequence, wherein the shared data item sequence is calculated by the initiator and the verifier based on the identifier of the target data item; and a verifier determination unit, used to enable the edge server to adaptively determine the verifier of the target data item according to the response message.

[0022] Preferably, the device also includes: an RSA accumulator initialization module, used to enable the edge server to initialize the first RSA accumulator using the base g; a shared data identification module, used to enable the edge server to identify the first shared data using the shared data item sequence; an RSA accumulator construction module, used to enable the edge server to construct the first RSA accumulator using a prime number mapped to the first shared data; and a verification proof calculation module, used to enable the edge server to calculate the verification proof of the verifier using the RSA accumulator.

[0023] Preferably, the batch verification module includes: a sampling request sending unit, used to enable the edge server to send a sampling request to the verifier, wherein the sampling request includes the public key of the edge server, the identifier of the target data copy, the shared data item sequence and the verification proof; a second shared data determination unit, used to enable the verifier to determine the second shared data according to the shared data item sequence; a second RSA accumulator construction unit, used to enable the verifier to construct a second RSA accumulator using a prime number mapped to the second shared data; a proof calculation unit, used to enable the verifier to calculate the proof using the second RSA accumulator; a voting result determination unit, used to enable the verifier to compare the calculated proof with the verification proof to obtain a voting result; and a voting result broadcasting unit, used to enable the verifier to broadcast the voting result in the system.

[0024] The present invention provides a method and device for determining the integrity of decentralized edge data, the determination method comprising: the edge server adaptively determines the target data item in the verifier according to the cache data status of other edge servers in its system; the edge server batch verifies the target data item through the verifier; and the edge server determines the integrity of the target data item according to the result of the batch verification. The present invention proposes an efficient decentralized edge data integrity scheme AccEDI, which can realize batch data verification by using RSA accumulators, significantly reduce the computing, communication and storage costs in the mobile edge computing environment, and ensure high integrity protection of data. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0026] Figure 1 A flow chart of a method for determining the integrity of decentralized edge data according to an embodiment of the present invention is shown;

[0027] Figure 2 The integrity verification process according to an embodiment of the present invention is shown, wherein (a) and (b) are VMHT and RSA accumulators, respectively;

[0028] Figure 3 The AccEDI process according to an embodiment of the present invention is shown;

[0029] Figure 4 The accuracy according to the embodiment of the present invention is shown, where (a) and (b) are the damage rate (cr) and the sampling size (ss), respectively;

[0030] Figure 5 1 shows the time consumption according to an embodiment of the present invention, wherein (a), (b), and (c) are the data block size (bz), the data scale (k), and the edge scale (m), respectively;

[0031] Figure 6 Graphs show the communication overhead according to an embodiment of the present invention, where (a), (b), and (c) are edge scale (m), data scale (k), and sampling scale (ss), respectively; and

[0032] Figure 7 The storage overhead according to an embodiment of the present invention is shown, where (a) and (b) are the data block size (bz) and the data scale (k), respectively. DETAILED DESCRIPTION

[0033] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0034] Reference will now be made in detail to various embodiments of the present invention, examples of which are shown in the accompanying drawings and described below. For ease of interpretation and precise definition in the appended claims, the terms "upper", "lower", "inner" and "outer" are used to describe features of the exemplary embodiments shown in the drawings with reference to their positions.

[0035] According to an embodiment of the present invention, a method for determining the integrity of decentralized edge data is provided, such as Figure 1 As shown, it includes: the edge server adaptively determines the target data item in the verifier according to the cache data status of other edge servers in its system; the edge server batch verifies the target data item through the verifier; and the edge server determines the integrity of the target data item according to the result of the batch verification.

[0036] In the related art, the batch data verification capability of decentralized edge data integrity is low, resulting in increased communication and computing overhead. In the embodiments of the present invention, through the synergy of adaptive selection, batch verification and integrity consensus, the communication and computing overhead is significantly reduced, which is particularly suitable for data integrity verification in a mobile edge computing environment.

[0037] According to an embodiment of the present invention, the edge server adaptively determines the target data item in the verifier according to the cache data status of other edge servers in the system, including: the edge server broadcasts an invitation message in the system, wherein the invitation message includes the public key of the edge server, the identifier of the target data copy and the identifier of the target data item in the target data copy; the edge server receives a response message from the other edge servers, wherein the response message includes the public key of one or more edge servers among the other edge servers, the identifier of the target data copy and a shared data item sequence, wherein the shared data item sequence is calculated by the one or more edge servers based on the identifier of the target data item; and the edge server adaptively determines the verifier of the target data item according to the response message.

[0038] In the embodiment of the present invention, by adaptively matching the target data items and their verifiers, unnecessary verification overhead is reduced, verification efficiency is improved, and the dynamic nature of data distribution in a decentralized environment is adapted.

[0039] According to an embodiment of the present invention, before the edge server performs batch verification on the target data items through the verifier, it also includes: the edge server initializes a first RSA accumulator using a base g; the edge server identifies first shared data using the shared data item sequence; the edge server constructs the first RSA accumulator using a prime number mapped to the first shared data; and the edge server uses the RSA accumulator to calculate the verification proof of the verifier.

[0040] In the embodiment of the present invention, the computational overhead is significantly reduced by utilizing the characteristics of the RSA accumulator that it can be dynamically updated and has a fixed size.

[0041] According to an embodiment of the present invention, the edge server performs batch verification on the target data item through the verifier, including: the edge server sends a sampling request to the verifier, wherein the sampling request includes the public key of the edge server, the identifier of the target data copy, the shared data item sequence and the verification proof; the verifier determines the second shared data according to the shared data item sequence; the verifier constructs a second RSA accumulator using the prime number mapped to the second shared data; the verifier calculates the proof using the second RSA accumulator; the verifier compares the calculated proof with the verification proof to obtain a voting result; and the verifier broadcasts the voting result in the system.

[0042] In the embodiment of the present invention, the batch verification feature of the RSA accumulator is utilized to achieve efficient verification of multiple data items, and to quickly locate damaged data items when verification fails, thereby improving the overall verification efficiency and accuracy.

[0043] According to an embodiment of the present invention, the verifier uses the second RSA accumulator to calculate the proof according to the following formula: Wherein, A is the calculated proof, g is the base, H(di) is the hash value of the target data item di, n is the number of the target data items, and N is the RSA modulus.

[0044] In the embodiment of the present invention, the batch verification formula is used to significantly improve the verification efficiency of multiple data items and achieve accurate positioning of damaged data items.

[0045] According to an embodiment of the present invention, the edge server determines the integrity of the target data item based on the result of the batch verification, including: the edge server determines whether the number of true results in one or more of the voting results exceeds half, and if the judgment result is yes, the edge server determines that the target data item is complete.

[0046] In the embodiment of the present invention, through statistical analysis of consistency ratios, the method can accurately determine the integrity status of data items and repair the data in a timely manner when the data is damaged, thereby enhancing the reliability of data verification.

[0047] According to another embodiment of the present invention, there is provided an apparatus for determining the integrity of decentralized edge data, comprising: an adaptive determination module, for adaptively determining a target data item in a verifier according to the cache data status of other edge servers in an edge server system; a batch verification module, for enabling the edge server to perform batch verification on the target data item through the verifier; and an integrity determination module, for enabling the edge server to determine the integrity of the target data item according to the result of the batch verification.

[0048] In the related art, the batch data verification capability of decentralized edge data integrity is low, resulting in increased communication and computing overhead. In the embodiments of the present invention, through the synergy of adaptive selection, batch verification and integrity consensus, the communication and computing overhead is significantly reduced, which is particularly suitable for data integrity verification in a mobile edge computing environment.

[0049] According to an embodiment of the present invention, the adaptive determination module includes: an invitation message broadcasting unit, used to enable the edge server to broadcast an invitation message in the system, wherein the invitation message includes a public key of the edge server, an identifier of a target data copy, and an identifier of the target data item in the target data copy; a response message receiving unit, used to enable the edge server to receive a response message from the other edge servers, wherein the response message includes a public key of one or more edge servers among the other edge servers, an identifier of the target data copy, and a shared data item sequence, wherein the shared data item sequence is calculated by the one or more edge servers based on the identifier of the target data item; and a verifier determination unit, used to enable the edge server to adaptively determine the verifier of the target data item according to the response message.

[0050] In the embodiment of the present invention, by adaptively matching the target data items and their verifiers, unnecessary verification overhead is reduced, verification efficiency is improved, and the dynamic nature of data distribution in a decentralized environment is adapted.

[0051] According to an embodiment of the present invention, the device also includes: an RSA accumulator initialization module, used to enable the edge server to initialize a first RSA accumulator using a base g; a shared data identification module, used to enable the edge server to identify first shared data using the shared data item sequence; an RSA accumulator construction module, used to enable the edge server to construct the first RSA accumulator using a prime number mapped to the first shared data; and a verification proof calculation module, used to enable the edge server to use the RSA accumulator to calculate the verification proof of the verifier.

[0052] In the embodiment of the present invention, the computational overhead is significantly reduced by initializing and constructing the RSA accumulator.

[0053] According to an embodiment of the present invention, the batch verification module includes: a sampling request sending unit, used to enable the edge server to send a sampling request to the verifier, wherein the sampling request includes the public key of the edge server, the identifier of the target data copy, the shared data item sequence and the verification proof; a second shared data determination unit, used to enable the verifier to determine the second shared data according to the shared data item sequence; a second RSA accumulator construction unit, used to enable the verifier to construct a second RSA accumulator using a prime number mapped to the second shared data; a proof calculation unit, used to enable the verifier to calculate the proof using the second RSA accumulator; a voting result determination unit, used to enable the verifier to compare the calculated proof with the verification proof to obtain a voting result; and a voting result broadcasting unit, used to enable the verifier to broadcast the voting result in the system.

[0054] In the embodiment of the present invention, the batch verification feature of the RSA accumulator is utilized to achieve efficient verification of multiple data items, and to quickly locate damaged data items when verification fails, thereby improving the overall verification efficiency and accuracy.

[0055] 1. The AccEDI system model and investigation process are described in detail below.

[0056] 1. System model

[0057] In the proposed scheme, the edge cache system consists of n ESs deployed in a specific area, represented by S = {s i |0≤i≤n}. Each ESs i ∈S has a public / private key pair and pass its public key The functions required by AccEDI are initialized at the beginning, including the cryptographic hash function H with a prime field prime () and the random function rand(). The symbols used in the present invention are summarized in Table 3.

[0058] Table 3: Key symbols

[0059] (1) Network Model

[0060] AccEDI follows the partially synchronous network assumption, where a message sent by an ES may be delayed but can eventually reach its target ES within a fixed and known time threshold Δ. This is practical in areas where ESs are connected by high-speed links.

[0061] (2) Fault model

[0062] This study is inspired by the robustness of distributed consensus protocols such as Paxos, Raft, and CooperEDI, which can effectively manage consensus among network participants even in the presence of potentially faulty processors. The basic assumption shared by these studies is that no more than half of the participants are simultaneously faulty. Applied to the EDI challenge, this assumption ensures that despite the possibility of data item corruption by some participants, a majority can still reach consensus. This assumption is adopted in the study where h data items from D are cached on n ESs. It is assumed that at any given time, only a minority of the data D are corrupted simultaneously - ensuring that at least Data copies remain intact. This situation is reasonable in a realistic distributed MEC environment because large-scale simultaneous corruption due to hardware failure, software problems, or network attacks is unlikely. In addition, widespread simultaneous corruption may also occur if many ESs fail to update their data copies in a timely manner, resulting in outdated data being considered corrupted. In this case, updating more than half of the ESs is sufficient to maintain data integrity.

[0063] (3) Security Model

[0064] AccEDI achieves deterministic integrity verification by thoroughly checking all data items stored on ESs. The protocol optimizes the integrity checking process by inviting verifiers and adaptively selecting ESs to verify different data items. It is assumed that at least three ESs are required to verify a data item to meet the consensus requirements of the failure model. Although operational limitations may occur when fewer than three ESs are available to verify a specific data item, this situation is considered unlikely. Therefore, the assumptions of the model are considered reasonable in ensuring effective data integrity under typical operating conditions. For simplicity, similar to other studies, the present invention assumes that there is no Byzantine attack.

[0065] 2. Investigation process

[0066] During the execution of AccEDI, only necessary parameters, voting information and data summaries are passed between ESs, avoiding the transmission of the entire data block. The communication between ESs follows the request-response protocol.

[0067] AccEDI's verification process is divided into two distinct phases: adaptive selection and batch verification, involving a total of four specific steps.

[0068] First, in the adaptive selection phase, the EDI initiator sends invitations to potential validators to form the integrity consensus committee.

[0069] The initiator is then responsible for adaptively selecting the data items to inspect.

[0070] Subsequently, during the bulk verification phase, the composed committee conducts bulk verification to assess the integrity of the selected data items and detect any possible corruption.

[0071] The specific process of AccEDI is as follows Figure 3 shown.

[0072] Phase 1: Adaptive Selection

[0073] Invite: After the system is set up, the s holding copy D 0 Send survey invitations to possible verifiers in the system. This invitation is mainly for S D ESs within the organization were consulted to solicit their willingness to participate in the D integrity assessment.

[0074] Answer: After receiving the survey invitation, each ES v ∈S D You can accept the invitation and become a validator. Each validator will submit 0 Send a reply, telling s 0 The sequence of data items they share is represented by Therefore, all the ESs that responded officially become validators and communicate with s 0 Together they formed the Integrity Consensus Committee.

[0075] Phase 2: Batch Verification

[0076] Request: Received from s v After the response of 0 Reason The marked data items construct the RSA accumulator proof and send it to s v Initiate a verification request.

[0077] Vote: Receive votes from s 0 After the request, the verifier s v According to s 0The proof verifies the data item and votes on the data item. In addition, if the overall verification result is false, it means that s 0 and v There are differences in the data, s v The different data items in the batch are further identified and then voted on. Once consistent results are received from all validators, the EDI survey reaches a consensus on integrity.

[0078] The following introduces the RSA accumulator used in the present invention and discusses its main advantages over VMHT.

[0079] Structure of an RSA Accumulator

[0080] Cryptographic accumulators were first proposed by Benaloh et al. as one-way hash functions with quasi-commutative properties that can compactly represent sets of a large number of elements as a single, fixed-size value (accumulator). This property allows efficient verification of whether an element belongs to a set. Accumulators make commitments to sets, ensuring that it is computationally infeasible to map different sets to the same accumulator value. The main security guarantee of accumulators is collision resistance, which makes it computationally difficult to generate false membership proofs for elements that are not in the set. Accumulators can generate concise proofs for members and, in some cases, proofs for non-members. A well-known one-way accumulator implementation is the RSA accumulator, which is particularly suitable for multi-set operations. Specifically, the RSA multi-set accumulator can represent multi-set A using a compact summary:

[0081] Where g is the RSA quotient group A permanent member of H prime () is a hard-to-divisible hash function with prime domain. Importantly, in order to satisfy Lemma 2 of the correctness section in Section 5, the value of g and H prime The results of () must be different. The formal definition of the RSA accumulator value calculation is as follows:

[0082] Given an element set S = {x 1 ,x 2 ,...,x n}, the RSA accumulator computes a single value A using a one-way function f defined over a finite field, typically involving modular exponentiation with a large prime modulus. Formally, the accumulator value A is defined as:

[0083] Here, g is a randomly selected base, N is the product of two large prime numbers, and the accumulator is unknown to the user.

[0084] Specifically, if Figure 2 As shown, the basic RSA accumulator is based on modular exponentiation under the strong RSA assumption. Given two strong prime numbers p and q are t-bit integers such that N = p*q, and given a base and N are relatively prime, the set X={x 1 ,…,x 9}(x i =H prime (d i The cumulative value of )) is calculated as follows:

[0085] The number of bits of the modulus N should be at least k bits, where k is the number of bits required for the largest number in the set X. Typically, the accumulator of the present invention defaults to the RSA quotient group members of , so unless otherwise stated, all accumulator formulas in the present invention are modulo N.

[0086] Core Operations of AccEDI

[0087] The following introduces the core algorithms used by the RSA accumulator in the present invention: aggregate batch processing and simple member verification.

[0088] Aggregate batch processing

[0089] This operation uses addition operations to aggregate multiple data items into a single accumulator without changing the size of the accumulator. This approach is significantly different from the approach taken by Boneh et al., who used batch processing to aggregate witnesses. The method of the present invention uses only addition operations to keep the data integration simple. For details on this operation, please see the following formula. Unlike MHT, the present invention only needs to calculate It is expressed as:

[0090] Rapid damage location

[0091] Using Lemma 2, the present invention implements a verification mechanism using the delete operation in the RSA accumulator to determine whether a data item is corrupted. In the RSA accumulator, the delete function has obvious advantages over the MHT. Unlike the MHT, which requires a complete recalculation of the tree after deleting a data item, the RSA accumulator allows efficient deletion of data elements directly using the prescribed formula:

[0092] It should be emphasized that, unlike the method adopted by Boneh et al., the solution of the present invention requires a trusted setup to calculate This formula is used to remove an element x from the accumulator A. According to the properties of integer exponents and Lemma 2, in a non-modulo N environment, if g and x are relatively prime, and x is indeed an element of the original accumulator A, then the resulting A′ will be an integer. Conversely, if x is not in A, A′ will be an irrational number. This distinction is crucial to determine whether x belongs to A.

[0093] Advantages of VMHT in EDI

[0094] In the MEC environment, EDI usually generates signatures based on MHT to prove the integrity of data copies. Li et al. proposed a series of solutions based on MHT and its variants. Take VMHT as an example (distributed EDI solutions such as CooperEDI also use similar verification mechanisms). Figure 2 As shown in the dashed box in the figure, AppVend and the verifier need to compare the sampled subtree root T i and T i ′, to determine whether the verifier’s data block is partially complete. In these verification processes, T i As a data block set d 1 ,d 2 ,d 3 ,d 4 The vector commitment, T i It is the proof of these four data blocks.

[0095] In contrast, AccEDI leverages RSA accumulators, a more concise and efficient approach. RSA accumulators compress the mapped data items into a set of multiplications, producing a fixed-size value, significantly simplifying the verification process. Boneh et al. showed that RSA accumulator proofs always remain 3000 bits in size. In addition, for sets of more than 4000 elements, RSA-based proofs outperform Merkle proofs, making them particularly suitable for MEC environments with multiple edge servers (ESs). It is worth noting that regardless of the number of data blocks or ESs, the accumulator size and proof size always remain the same, significantly reducing storage overhead. During EDI verification, two ESs compare and verify their shared set of data items, such as Figure 2 In this process, A i As the data item set d 1 ,d 2 ,d 3 ,d 4 The vector accumulator commitment of provides an efficient aggregation proof for these terms.

[0096] 2. AccEDI Framework

[0097] The present invention details the two phases of AccEDI. The framework of the present invention implements a scheme with efficient batch verification and adaptive data selection, maintaining high accuracy and low communication and storage costs.

[0098] 1. Adaptive selection stage

[0099] This paper introduces the AccEDI protocol for reaching integrity consensus. Assuming an ESs 0 Want to investigate its copy D, the integrity investigation process includes four steps, such as Figure 3 shown.

[0100] (1) Step 1: Invite

[0101] -1- System Settings

[0102] System settings are not in Figure 3 In this step, AccEDI sets the basic parameters, including time thresholds and universal hash functions. AccEDI assigns two time thresholds to each ES: d sets the start time of the EDI investigation, which can be predefined by the application provider or adjusted by the ES according to its computing resources; Δ is the maximum time each ES waits for a message in each investigation step. In addition, AccEDI must select a base g of the accumulator for all ES in the system, which is usually set to 2 in actual implementation. In addition, AccEDI selects a hash function H with a prime domain for all ES in the system. prime (). This function generates a data block digest using algorithms such as MD5, SHA-1, and SHA-256, and then maps it to the corresponding prime number. For detailed algorithms, please refer to Boneh's paper.

[0103] -2-Initiate an invitation

[0104] Initiators 0 Broadcast survey invitation inv in the system in the form of in Yes 0 The public key is used to indicate the sender of the message, R.id is the ID of the data copy to be investigated, D.id indicates the ID of the data item in the data copy to be investigated, S D All members of the designated integrity consensus committee. The research of this invention draws on the successful experience of distributed consensus protocols such as CooperEDI and EdgeWatch, which assume that each ES retains a list of IDs of all ESs storing data D. This approach trades storage for reduced communication overhead, allowing potential verifiers to be quickly identified. The present invention adopts a similar assumption.

[0105] (2) Step 2: Answer After receiving the survey invitation inv, each ESs v ∈S D Check two conditions: ●s 0 YesS D The legal data owner in ●s v Owns the data D or a subset thereof.

[0106] If both conditions are met, s v Become a validator. Next, v To s 0 Send ans. It is a tuple in Yes v The public key of s 0 and v A sequence of data items shared between s 0 The D.id is calculated, including the initiator s v Verify specific items selected for adaptive selection.

[0107] 1. Batch verification stage

[0108] (1) Step 3: Request

[0109] Received from s v After the EDI invitation is answered by ans, the initiator s 0 Select participants to form the integrity consensus committee and send req to all validators as described in Algorithm 1

[0110] First, 0 Filter irrelevant ans messages (lines 1-2). Then, s 0 Initialize the RSA accumulator A with the base g j (Line 3). Next, s 0 Use the verifier's answer ans.D bar Adaptively identify shared data, represented as sharedData j [] and use map to sharedData j The prime number p of [x] x Construct accumulator A with formula [add] j (Lines 5-8). Finally, s 0 To s v Send contains tuple A sampling request, where A j It is for v Verification proof (line 9). It is worth noting that s0 Through dynamic operation A j Implementing integrity proof A j Reuse between different datasets.

[0111] (2) Step 4: Vote

[0112] Received from the initiator 0 After the req, the verifier s v Perform verification, vote on each data item involved, and broadcast its own voting information, see Algorithm 2 for details.

[0113] First, v Filter its data items according to D.bar in the Answer step to determine sharedData j [] (line 1). Then, similar to the Request step, s v Use map to sharedData j The prime number p of [x] x 'Build accumulator A j '(lines 2-5). Next, s v Proof of its calculation A′ j With the proof received j .Aj is compared to determine voteResult[] (lines 6-9). If the verification result is true, it means that s 0 and v All data items between s v All data items will be voted (lines 6-7). If there is no match, it means s 0 and v There are differences in the data items, and the different data items need to be further identified. By using Algorithm 3 to identify the specific differences, s v Voting is performed on data items other than the identified items (lines 8-9). Finally, s v Create and broadcast a voting message (lines 10-11), tuple where voteResult[] is s v Voting information.

[0114] When vWhen verification fails and differences are found in data items, it becomes critical to identify the specific different items. For details on the relevant algorithm, see Algorithm 3. v Use from s 0 Proof A and the prime number set pSet[] calculated by itself are used for positioning calculation. As described in Lemma 2 and Formula 5, by verifying the accumulator result temp after the deletion operation, s can be determined. 0 Prove that A contains v The hash value of the data item. According to Lemma 2, if X is divisible by n, then accValue is confirmed to contain PrimeSet[x] and temp should be an integer (Lines 2-4). Conversely, if X is not divisible by n, accValue lacks PrimeSet[x], indicating that the two ESs have a difference in the data item at position x, and temp will not be an integer (Lines 5-6). It should be noted that since the modulo N environment ensures that all calculation results are still integers, there is no need to apply the modulo operation in this deletion process.

[0115] In all s v After voting, the integrity consensus committee reaches a consensus based on the collected voting data. x If the number of affirmative votes obtained exceeds half of the number of ESs that cache it, it is considered intact; otherwise, it is considered damaged. i After collecting all votes, s i From the accumulated vot.D bar Calculate n x , and then count the total number of votes totalVote for the data items involved x If totalVote x >n x / 2, then confirm d x Complete. The correctness of AccEDI consensus is rigorously proved in Theorem 2 in Section 5. In addition, according to different data items d x Using different total number of ES n x It is also a reflection of AccEDI’s adaptability.

[0116] 3. Performance Analysis

[0117] The performance of AccEDI is theoretically analyzed from three aspects: correctness, efficiency and security.

[0118] 1. Correctness

[0119] The correctness of AccEDI depends on two key aspects: EDI verification of corrupted data copies and location of corrupted data items.

[0120] Lemma 1. If the attacker cannot construct the accumulator, element x, and valid membership proof w x , then the accumulator is safe.

[0121] Lemma 1 is from Boneh et al. and is the basis for Theorem 1. The proof of Lemma 1 can be found in Definition 6 of the paper Accumulator Security (Non-repudiation).

[0122] Theorem 1. Intersection datasets cached in ES If the hash function H prime () is collision resistant, the corrupted data item will fail the AccEDI validation process.

[0123] Lemma 1 and Theorem 1 ensure that if the accumulator, its algorithm, and the hash function are correct and collision-resistant, then the RSA accumulator can be used correctly to solve the EDI authentication problem. This guarantees the correctness of the verification of corrupted copies of the data.

[0124] Lemma 2. Let g, n, and X be integers, and g and n are different prime numbers. Then, if n does not divide X, then Must be a non-integer.

[0125] Lemma 2 ensures that AccEDI can correctly and precisely locate the damaged data items when it detects corruption.

[0126] Theorem 2. For a single data item d i If you have d i More than half of the ESs confirmed their version of d i If the version of the initiator matches the version of the initiator, the initiator's data item d i is complete. On the contrary, if you have d i More than half of the ESs confirmed their version of d i If the version does not match the initiator's version, the initiator's data item d i Damaged.

[0127] Theorem 2 shows that AccEDI can correctly determine the integrity of the reviewed dataset.

[0128] 2. Efficiency

[0129] The performance of the proposed scheme in terms of computation, communication and storage costs is analyzed from the perspective of ES. Theoretical analysis shows that AccEDI has significant advantages in terms of communication and storage overhead, and after aggregation optimization, the computational overhead is still within an acceptable range.

[0130] The results are shown in Table 4, where m is the number of all ESs storing data replicas D, k is the number of data items in D, and h is The number of data items in |S D |=m,|D|=k and In order to evaluate the efficiency of the proposed scheme, a prototype is introduced. Table 5 lists the symbols used in this section to represent the operation costs.

[0131] Table 4: Efficiency analysis

[0132] Table 5: Operating cost symbols symbol Operating costs <![CDATA[T hash ]]> Hash function calculation <![CDATA[T mul ]]> Modular multiplication <![CDATA[T exp ]]> Modular exponentiation

[0133] Finally, an extensive evaluation of AccEDI is performed to compare its performance with two established MHT-based decentralized EDI schemes and an accumulator-based centralized EDI scheme in terms of accuracy, computation time, communication overhead, and storage cost. A prototype system is developed and simulated under various environmental conditions and operating parameters.

[0134] The results consistently show that AccEDI achieves high accuracy comparable to existing methods. Moreover, it exhibits relatively low time overhead and reduced communication and storage costs, confirming its superior efficiency and accuracy.

[0135] Experimental setup

[0136] Environment Setup

[0137] All experiments were conducted on an Alibaba Cloud server equipped with an Intel Xeon (Ice Lake) processor (2.7 GHz / 3.3 GHz), 64 GB memory, and Ubuntu Server 22.04 LTS. ES was simulated as different processes, and the network latency between servers fluctuated between 60 ms and 120 ms.

[0138] Benchmark Methodology

[0139] In order to compare with AccEDI, the present invention uses three methods for experiments: CAPDP (centralized, based on RSA accumulator), CooperEDI (decentralized, based on MHT) and EdgeWatch (decentralized, based on MHT and blockchain).

[0140] CAPDP: A distributed data integrity assurance scheme for MEC environments. It uses a distributed consensus mechanism to create a self-managed edge cache system, where ESs collaborate to ensure and repair data integrity.

[0141] CooperEDI: A distributed data integrity assurance solution for MEC environments. It uses a distributed consensus mechanism to create a self-managed edge cache system where ESs collaborate to ensure and repair data integrity.

[0142] EdgeWatch: A blockchain-based collaborative edge data integrity investigation scheme that adopts a novel integrity consensus collaboration with incentive mechanisms, reputation systems, and leader randomization techniques.

[0143] Performance Indicators

[0144] Precision: The ratio of detected corrupted data copies to the total number of corrupted data copies. Higher precision indicates better performance.

[0145] Time consumption: The total computation and communication time required by all ESs to complete the data integrity investigation process. The lower the value, the better.

[0146] Communication overhead: The total communication overhead of all ESs during the data integrity investigation process.

[0147] Storage overhead: The average storage overhead of each ES during the data integrity investigation.

[0148] Parameter settings

[0149] In order to comprehensively evaluate the performance of AccEDI, the present invention sets the parameters shown in Table 6. In each experiment, only one parameter is modified and the other parameters remain at the default settings. Each experiment is repeated 100 times and the average results are recorded. The last row of the Byzantine scale includes three default values ​​to dynamically maintain one-third of the edge scale.

[0150] Edge size (m), the total number of ESs.

[0151] Data scale (k), the total number of data items in AppVend. Each ES stores a portion of the data, and each ES caches more than half of the data items to better reflect the real scenario.

[0152] Data block size (bz), the size of each data block (in KB).

[0153] Sampling size (ss), in AccEDI, refers to the number of primes sampled for the second time in each data copy. In EdgeWatch, it refers to the number of data blocks sampled each time.

[0154] Corrupted block ratio (cr), the ratio of the number of corrupted blocks in the corrupted data replica to the total number of blocks.

[0155] Each data copy consists of 1024 data blocks. The size of each data block varies according to the size of the data copy, from 16KB to 256KB, and the default size is 64KB. Table 6: Parameter settings

[0156] Experimental Results

[0157] The present invention examines the performance of AccEDI, CAPDP, CooperEDI and EdgeWatch in detail under various indicators, including accuracy, efficiency and storage utilization. The present invention first reviews the accuracy of these methods, especially their ability to detect corrupted data copies. Subsequently, the present invention conducts a comprehensive evaluation of their overall performance, including time efficiency, communication overhead and storage cost. The experimental results show that AccEDI has excellent investigation accuracy and efficiency.

[0158] Accuracy

[0159] Experimental results show that AccEDI achieves a high level of accuracy, comparable to existing EDI schemes. It is noteworthy that variables such as edge scale (m), data scale (k), and data block size (bz) do not affect the accuracy of EDI. Figure 3 The accuracy performance of the four schemes under different sampling scales (ss) and damaged block ratios (cr) is demonstrated.

[0160] Accuracy and damaged block ratio Figure 4 The relationship between the corrupted block ratio (cr) and accuracy is described. EdgeWatch uses probabilistic integrity verification, which is affected by changes in cr. At a fixed sampling scale, an increase in cr enhances the ability of probabilistic schemes such as EdgeWatch to detect data corruption, which is theoretically supported in multiple studies. In contrast, deterministic schemes such as CooperEDI, CAPDP, and AccEDI remain at 100.

[0161] Precision and sampling size Figure 4 The change in accuracy of the four schemes is shown as the sampling size (ss) increases from 32 to 512. The increase in ss means that more data blocks are sampled, leading to a more thorough investigation. Therefore, schemes like EdgeWatch are better at detecting data corruption. The results of CAPDP, CooperEDI, and AccEDI are similar to Figure 4 Similar to what is shown in , AccEDI always achieves the highest accuracy. Since the principle is the same, further details are omitted here.

[0162] Time consumption

[0163] Figure [Ex_Time] compares the time efficiency of four different schemes in various EDI scenarios. Although the modular exponentiation operation of AccEDI increases the computation time, compared with EdgeWatch, when k and m increase, the batch computing characteristics of AccEDI become more and more obvious, and its computation time is more advantageous than other schemes. The measured time includes the entire EDI process, covering the computation time and communication delay of ESs. Since the corruption rate (cr) and sampling size (ss) do not affect the time consumption, the relevant comparisons are omitted.

[0164] Time consumption and data block size Figure 5 It illustrates how the data block size (bz) affects the time consumption. When bz increases from 16KB to 256KB, the time consumption of EdgeWatch increases by 4.28%, while that of CooperEDI increases by 420.16%. In contrast, the time consumption of CAPDP and AccEDI stabilizes at approximately 463.18ms and 187.94ms, respectively, with sample standard deviations of approximately 4.32 and 1.90. Due to the slower calculation speed of the RSA accumulator, the overall calculation time of CAPDP is higher than that of CooperEDI and EdgeWatch. Thanks to AccEDI's novel aggregation algorithm, the calculation time of AccEDI is only slightly higher than that of EdgeWatch, but this is acceptable considering AccEDI's advantages in other aspects such as efficiency and accuracy.

[0165] Time consumption and data size Figure 5 The variation of time cost under different data sizes (k) is explored. As k increases from 128 to 2048, the need for more hash operations drives the increase in time consumption in most schemes. CooperEDI and EdgeWatch show significant increases of 201.08 times and 20.17 times, respectively. However, AccEDI and CAPDP use fixed-size RSA accumulators and only increase by 15.35 times and 13.58 times, respectively. While CAPDP and AccEDI are generally slower due to RSA modulus calculations, AccEDI's batch processing and adaptive data selection provide significant advantages when the data size increases.

[0166] Time consumption and edge size Figure 5 The impact of increasing edge size (m) on time consumption is evaluated. Increasing the number of ES nodes increases the time cost of integrity verification for all EDI solutions. Notably, all decentralized systems significantly reduce their time cost with changes in m compared to the centralized system CAPDP. As with the previous metrics, AccEDI initially requires more time in small-scale scenarios; however, its advantage becomes increasingly evident as the scale increases.

[0167] Communication overhead

[0168] AccEDI is able to perform batch processing and aggregate verification of different data items, and has significant advantages over other solutions in terms of communication overhead. Figure 6 The communication overhead of different schemes is compared, which is only affected by the edge scale (m), data scale (k) and sampling scale (ss).

[0169] like Figure 6 As shown, regardless of the edge scale (m), data scale (k) or sampling scale (ss), the communication cost of AccEDI is significantly lower than that of the other three solutions. This is because AccEDI batches and verifies data items more flexibly, making its communication overhead less affected by the number of data items. No matter how many data items there are, they can be aggregated into witnesses of a fixed size. Although CAPDP also uses RSA accumulators, it does not adopt aggregation operations. EdgeWatch and CooperEDI verify data items individually and require the same number of verification rounds as the number of data items. Therefore, the communication overhead of these three solutions is high, while AccEDI greatly reduces the communication burden of EDI in the MEC environment.

[0170] Storage overhead

[0171] Compared with the other three solutions, AccEDI shows excellent stability in terms of storage overhead. Figure 7 The storage requirements of the four schemes under different data block sizes (bz) and data scales (k) are evaluated, and it is shown that the average storage load of all participating ESs during the data integrity investigation is only affected by bz and k.

[0172] like Figure 7 As shown, the change in storage load due to changes in bz and k is consistent. This consistency is because the product of bz and k, as defined in Table 4, remains constant, thereby minimizing the impact on storage cost. Although AccEDI initially incurs slightly higher storage overhead, it becomes more storage efficient than other solutions as bz or k increases. This efficiency is attributed to the RSA accumulator used in AccEDI, which, unlike the scalable Merkle tree adopted by CooperEDI and EdgeWatch, maintains a consistent storage size due to its modular nature. It is worth noting that while both CAPDP and AccEDI use RSA accumulators, CAPDP's centralized approach and lack of batch processing techniques result in significantly higher storage requirements. On average, AccEDI consumes 47.22 less storage than CAPDP, EdgeWatch, and CooperEDI, respectively.

[0173] In summary, in order to address the deficiencies in the prior art, the present invention proposes AccEDI, a decentralized EDI solution based on RSA accumulators, which can achieve efficient batch verification and adaptive data selection. AccEDI utilizes the aggregation characteristics of RSA accumulators to promote batch verification of data items, and implements adaptive data selection based on data subsets cached by different ESs. The method of the present invention enables the EDI initiator to adaptively select the data items of the verifier and simultaneously complete the damage location in the batch proof in each ES, which significantly improves efficiency and greatly reduces overhead.

[0174] The main contributions of the present invention are as follows:

[0175] Efficient Batch Verification with Adaptive Validator Selection: AccEDI employs a novel cryptographic integrity verification protocol based on RSA accumulators, enabling the integrity consensus committee to efficiently reach consensus on EDI findings. By leveraging the aggregation capabilities of RSA accumulators, AccEDI can verify all data items in a single voting event, significantly reducing communication overhead and verification time. In addition, AccEDI allows the initiator to adaptively select data items based on a subset of data cached by the validator, further improving efficiency.

[0176] Accurate corruption localization: Unlike MHT-based methods, AccEDI uses RSA accumulators to accurately identify damaged data items in batch verification, achieve targeted data repair, and minimize unnecessary data retransmissions.

[0177] Theoretical Analysis and Evaluation: We provide a theoretical analysis of the correctness and efficiency of AccEDI, demonstrating its advantages over existing approaches. In addition, we conduct a comprehensive evaluation in a test system, comparing AccEDI with representative decentralized and centralized data integrity schemes.

[0178] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for determining the integrity of decentralized edge data, characterized in that: include: The edge server adaptively determines the target data item in the verifier based on the cache data status of other edge servers in its system; The edge server performs batch verification on the target data items through the verifier; and The edge server determines the integrity of the target data item according to the result of the batch verification.

2. The method according to claim 1, characterized in that The edge server adaptively determines the target data items in the verifier based on the cache data status of other edge servers in the system, including: The edge server broadcasts an invitation message in the system, wherein the invitation message includes a public key of the edge server, an identifier of a target data copy, and an identifier of the target data item in the target data copy; The edge server receives a series of response messages from the other edge servers, wherein the response messages include public keys of one or more edge servers among the other edge servers, an identification of the target data copy, and a shared data item sequence, wherein the shared data item sequence is calculated by the one or more edge servers based on the identification of the target data item; and The edge server adaptively determines the verifier of the target data item according to the response message.

3. The method according to claim 2, characterized in that Before the edge server performs batch verification on the target data items through the verifier, the method further includes: The edge server initializes a first RSA accumulator using the base g; The edge server identifies first shared data using the sequence of shared data items; The edge server constructs the first RSA accumulator using a prime number mapped to the first shared data; and The edge server calculates the verification proof of the verifier using the RSA accumulator.

4. The method according to claim 3, characterized in that The edge server batch verifies the target data items through the verifier, including: The edge server sends a sampling request to the verifier, wherein the sampling request includes a public key of the edge server, an identifier of the target data copy, the sequence of shared data items, and the verification certificate; The verifier determines the second shared data according to the shared data item sequence; The verifier constructs a second RSA accumulator using a prime number mapped to the second shared data; The verifier calculates the proof using the second RSA accumulator; The verifier compares the calculated proof with the verification proof to obtain a voting result; The validator broadcasts the voting result in the system.

5. The method according to claim 4, characterized in that The verifier uses the second RSA accumulator to calculate the proof according to the following formula: Where A is the calculated proof, g is the base, H(d i ) is the target data item d i The hash value of , n is the number of target data items, and N is the RSA modulus.

6. The method according to any one of claims 1 to 5, characterized in that The edge server determines the integrity of the target data item according to the result of the batch verification, including: The edge server determines whether the number of true results among the plurality of voting results exceeds half, and if the determination result is yes, the edge server determines that the target data item is complete.

7. A device for determining the integrity of decentralized edge data, characterized in that: include: An adaptive determination module, configured to adaptively determine a target data item in the verifier according to a cache data status of other edge servers in the edge server system; A batch verification module, configured to enable the edge server to perform batch verification on the target data items through the verifier; as well as The integrity determination module is used to enable the edge server to determine the integrity of the target data item according to the result of the batch verification.

8. The device according to claim 7, characterized in that The adaptive determination module comprises: an invitation message broadcasting unit, configured to enable the edge server to broadcast an invitation message in the system, wherein the invitation message includes a public key of the edge server, an identifier of a target data copy, and an identifier of the target data item in the target data copy; a response message receiving unit, configured to enable the edge server to receive a response message from the other edge servers, wherein the response message includes a public key of one or more edge servers among the other edge servers, an identifier of the target data copy, and a shared data item sequence, wherein the shared data item sequence is calculated by the initiator and the verifier based on the identifier of the target data item; The verifier determination unit is used to enable the edge server to adaptively determine the verifier of the target data item according to the response message.

9. The device according to claim 8, characterized in that The device also includes: An RSA accumulator initialization module, configured to enable the edge server to initialize a first RSA accumulator using a base g; a shared data identification module, configured to enable the edge server to identify first shared data using the shared data item sequence; An RSA accumulator construction module, configured to enable the edge server to construct the first RSA accumulator using a prime number mapped to the first shared data; The verification proof calculation module is used to enable the edge server to calculate the verification proof of the verifier using the RSA accumulator.

10. The device according to claim 9, characterized in that The batch verification module includes: a sampling request sending unit, configured to enable the edge server to send a sampling request to the verifier, wherein the sampling request includes a public key of the edge server, an identifier of the target data copy, the shared data item sequence, and the verification certificate; a second shared data determining unit, configured to enable the verifier to determine the second shared data according to the shared data item sequence; A second RSA accumulator construction unit, configured to enable the verifier to construct a second RSA accumulator using a prime number mapped to the second shared data; a proof calculation unit, configured to enable the verifier to calculate a proof using the second RSA accumulator; a voting result determination unit, configured to enable the verifier to compare the calculated proof with the verification proof to obtain a voting result; A voting result broadcasting unit is used to enable the verifier to broadcast the voting result in the system.

Images