Privacy protection-based trustworthy data aggregation method and system for crowd-sensing network

Abstract

The application relates to the technical field of data processing, and discloses a group intelligence perception network trusted data aggregation method and system based on privacy protection, which comprises the following steps: a key center distributes a cryptographic key to all participants, assigns tasks to matched workers in a manner without leaking the type of tasks that the workers can perform, and completes privacy protection task distribution; the selected workers submit encrypted perception data and related metadata to a blockchain for privacy protection data submission; a true value discovery algorithm is cooperatively executed in a privacy protection manner, the credit of the workers is iteratively updated, and more accurate data true values are calculated under privacy protection based on the updated credit, so that credit updating and privacy protection true value discovery are realized; malicious and low-credit workers are excluded in the true value discovery process; and a reward is calculated according to the contribution and credit of each worker, and is safely distributed to the corresponding worker through the blockchain, so that the reward calculation is completed.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and system for trusted data aggregation in a privacy-preserving crowd sensing network. Background Technology

[0002] With the widespread adoption of the Internet of Things (IoT) and mobile devices, crowdsourcing sensing, as a large-scale data collection method, has been widely applied in various fields such as environmental monitoring, intelligent transportation, and urban computing. This model relies on a large number of participants collecting data using their personal devices and then using a platform to aggregate and analyze the data to obtain the "true value" of the sensing task. However, existing crowdsourcing sensing still faces three core problems. First, there is the issue of data reliability. Participants responsible for data collection may submit low-quality or even forged data due to insufficient ability or malicious motives, seriously affecting the accuracy of the aggregation results. While traditional truth-finding algorithms can avoid low-quality data to some extent by iteratively calculating data weights, they are not robust to malicious data such as collaborative cheating. Second, there is the issue of privacy leakage. Sensing data often contains sensitive information about participants (such as location, daily routines, and behavioral habits). If this sensitive information is directly uploaded to a centralized server, there is a risk of misuse by the platform or attackers. Although some works have used differential privacy or simple encryption to intentionally protect privacy, the former introduces irreversible noise that affects the utility of the data, while the latter cannot support direct computation on encrypted data, affecting the efficiency of data use. Finally, there is the issue of process transparency and fairness. As a centralized management platform, there is a possibility of unilateral data tampering within it, and the weight and reward distribution results can also be modified manually, resulting in a lack of transparency in the entire process and unreliable output results, which seriously dampens the enthusiasm of participants.

[0003] Blockchain technology, as a novel data processing technology, offers a new approach to solving trust issues due to its decentralized and tamper-proof characteristics. Simultaneously, dynamic encryption technology allows for specific operations directly on ciphertext, enabling data computation with privacy protection. However, current research largely involves simply combining these two technologies without deep integration. For example, some studies only store encrypted data in the blockchain, but the crucial data aggregation and reputation update logic is still executed by a centralized server within the blockchain, failing to completely solve the single-point trust problem. Therefore, existing crowdsourced data aggregation methods suffer from low data reliability and poor privacy. Summary of the Invention

[0004] This invention provides a trusted data aggregation method and system for crowd sensing networks based on privacy protection, in order to solve the problems of low data reliability and poor privacy in existing crowd sensing data aggregation methods.

[0005] To achieve the above objectives, the present invention employs the following technical solution:

[0006] In a first aspect, the present invention provides a method for trusted data aggregation in a privacy-preserving crowd sensing network, comprising: S1. The key center distributes cryptographic keys to all participants and assigns tasks to matching workers in a way that does not reveal the types of tasks that workers can perform, thus completing the privacy-preserving task allocation. S2. The selected worker submits encrypted perception data and related metadata to the blockchain for privacy-preserving data submission. S3. Collaborately execute the truth discovery algorithm in a privacy-preserving manner, iteratively update the worker's reputation, and calculate a more accurate data truth value based on the updated reputation under privacy protection, performing reputation update and privacy-preserving truth discovery. During the truth discovery process, malicious and low-reputation workers are excluded. S4. The data requester calculates the reward based on each worker's contribution and reputation, and distributes it securely to the corresponding worker through the blockchain, thus completing the reward calculation.

[0007] Optionally, in step S1, the privacy protection task allocation specifically includes the following steps: S11. Based on the maximum number of task requests m from the data requester, the key center generates a core encryption matrix, a re-encryption matrix, and a hash function for task privacy mapping, and calculates and generates keys for different purposes to be distributed to the data requester, workers, and auxiliary servers. S12. The worker uses their key to map their set of executable tasks to task numbers through the hash function, and after encrypting the cyclic matrix based on each number, uploads the encryption result to the blockchain. S13. The data requester uses its key to map and construct a polynomial for its n actual tasks and (mn) virtual tasks to be published through the hash function, generates the corresponding cyclic matrix, performs pre-encryption, and uploads the encryption result to the blockchain. S14. The auxiliary server obtains the encryption matrix uploaded by workers and data requesters from the blockchain and performs re-encryption conversion using its own key; S15. In encrypted state, the auxiliary server determines whether the task set of the data requester is a subset of the task set that a certain worker can execute by calculating the matrix trace, thereby achieving privacy-preserving task matching, and sending the information of the successfully matched worker to the data requester. S16. The data requester filters the successfully matched workers based on their reputation, forming a final set of workers to perform the task, and uploads this set to the blockchain.

[0008] Optionally, in step S2, the submission of privacy-protected data specifically includes the following steps: S21. After the worker confirms that he has been selected from the blockchain, he generates perception data and corresponding random perturbations for each task that needs to be reported. S22. The worker constructs a circular matrix from the perturbated data, the original data, and the perturbated data, and encrypts them using their keys to generate multiple encryption matrices. S23. The worker uploads the encryption matrix, the original sensing data value, and the sum of squared disturbances to the blockchain; S24. The data requester generates an initial predicted value for the true value of the task and the corresponding random perturbation, constructs a cyclic matrix and encrypts it, and sends it to the auxiliary server along with the estimated weighted maximum value and the truth perturbation factor. S25. The auxiliary server generates an over-incrementing sequence and constructs a cyclic matrix encryption. At the same time, it re-encrypts the encryption matrix sent by the data requester and sends all the processed matrices back to the data requester. S26. The data requester pulls the data uploaded by the workers from the blockchain and re-encrypts the matrix received from the auxiliary server locally. S27. The data requester calculates the weight of each worker based on the re-encrypted matrix, using a formula that includes matrix trace operations, while protecting privacy.

[0009] Optionally, in step S3, the reputation update and privacy-preserving truth discovery specifically includes the following steps: S31. Based on the worker weights calculated in Phase 2, the data requester updates the reputation of other workers using the average weight of workers falling into the set of trustworthy workers as a benchmark; the update formula satisfies the following relationship: ; In the formula, This represents the worker's creditworthiness in round t+1. This indicates the worker's creditworthiness in round t. This represents the maximum permissible difference between the weights calculated for workers in the previous stage and the average weights of reliable workers. Indicates the positive learning rate. Indicates the reverse learning rate. Indicates sensitivity; S32. Based on the updated reputation score, add workers with reputation scores above the threshold α to the trusted set and workers with reputation scores below the threshold β to the malicious set, and readjust the weights of each worker in truth discovery accordingly. S33. The data requester uses the truth perturbation factor h, the super-incrementing sequence ciphertext matrix, and the encrypted matrix of worker data to calculate the trace of the weighted sum, and with the assistance of the auxiliary server, iteratively calculates a new round of task truth estimation in a privacy-preserving manner. S34. Repeat the iterative process until the change in the true value estimation for two consecutive rounds is less than the preset threshold, and obtain the final true value of the data.

[0010] Optionally, in step S4, the reward calculation specifically includes: The data requester allocates rewards from the reward pool according to each worker's weight in the final iteration, proportionally to their total weight. The reward calculation formula satisfies the following relationship: ; In the formula, Indicates the final iteration weight. This represents the sum of the weights of all participating workers. This represents the total reward amount corresponding to the total reward pool. This represents the total reward pool.

[0011] In a second aspect, embodiments of this application provide a trusted data aggregation system for a crowd-aware network for privacy protection, used to implement the method as described in any of the first aspects, the system comprising: The Data Requester module is used to publish task requirements, initially screen workers, collaboratively conduct privacy-preserving truth discovery and reputation updates, and finally calculate and distribute rewards. The worker client module is used for registration, receiving tasks, collecting sensing data, encrypting and protecting the privacy of task capabilities and sensing data, and uploading them to the blockchain. The blockchain network module, as a decentralized ledger, is used to store encrypted task information, sensing data, intermediate parameters, and transaction records, ensuring the immutability and transparency of the process. The auxiliary server module is configured as a semi-trusted third party, holding a specific key, to perform task matching in encrypted state and assist in completing data aggregation and truth calculation under privacy protection, without knowing the original task information or perceived data of any participant. The key center module is used to generate and securely distribute a complete set of cryptographic keys to all the aforementioned participants during the system initialization phase.

[0012] Beneficial effects: This invention provides a trusted data aggregation method for crowdsourced sensing networks for privacy protection. Through a dynamic reputation mechanism, the system can automatically identify and reduce the weight of low-quality or malicious data, significantly improving the accuracy of the final truth value estimation and maintaining strong robustness even when the proportion of malicious participants is high. Homomorphic encryption technology ensures that the entire process from data submission to aggregation calculation is conducted in encrypted form, preventing the data requester and the aggregation server from accessing the original sensing data, fundamentally preventing participant privacy leaks. Blockchain storage of all encrypted data, intermediate parameters, and computation logs ensures the immutability and traceability of the entire aggregation process, establishing a decentralized trust foundation. Reputation is directly linked to rewards, incentivizing participants to provide high-quality data. The reward distribution process based on blockchain smart contracts is automated and transparent, avoiding potential fraudulent activities on centralized platforms. Attached Figure Description

[0013] Figure 1 This is a flowchart of a preferred embodiment of the trusted data aggregation method for swarm intelligence sensing networks for privacy protection according to the present invention; Figure 2 The false ratio R in a preferred embodiment of the present invention F The following is a schematic diagram of the RMSE corresponding to the method of the present invention; Figure 3 This diagram illustrates the privacy leakage costs and proxy costs for different numbers of proxies, as provided in a preferred embodiment of the present invention. Detailed Implementation

[0014] The technical solution of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0015] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, the terms "an" or "a" and similar terms do not indicate a quantity limitation, but rather indicate the presence of at least one. The terms "connected" or "linked" and similar terms are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up," "down," "left," "right," etc., are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship also changes accordingly.

[0016] Please see Figure 1 This application provides a privacy-preserving method for trusted data aggregation in a crowd-sensing network, comprising: S1. The key center distributes cryptographic keys to all participants and assigns tasks to matching workers in a way that does not reveal the types of tasks that workers can perform, thus completing the privacy-preserving task allocation. S2. The selected worker submits encrypted perception data and related metadata to the blockchain for privacy-preserving data submission. S3. Collaborately execute the truth discovery algorithm in a privacy-preserving manner, iteratively update the worker's reputation, and calculate a more accurate data truth value based on the updated reputation under privacy protection, performing reputation update and privacy-preserving truth discovery. During the truth discovery process, malicious and low-reputation workers are excluded. S4. The data requester calculates the reward based on each worker's contribution and reputation, and distributes it securely to the corresponding worker through the blockchain, thus completing the reward calculation.

[0017] Optionally, in step S1, the privacy protection task allocation specifically includes the following steps: S11. Based on the maximum number of task requests m from the data requester, the key center generates a core encryption matrix, a re-encryption matrix, and a hash function for task privacy mapping, and calculates and generates keys for different purposes to be distributed to the data requester, workers, and auxiliary servers. S12. The worker uses their key to map their set of executable tasks to task numbers through the hash function, and after encrypting the cyclic matrix based on each number, uploads the encryption result to the blockchain. S13. The data requester uses its key to map and construct a polynomial for its n actual tasks and (mn) virtual tasks to be published through the hash function, generates the corresponding cyclic matrix, performs pre-encryption, and uploads the encryption result to the blockchain. S14. The auxiliary server obtains the encryption matrix uploaded by workers and data requesters from the blockchain and performs re-encryption conversion using its own key; S15. In encrypted state, the auxiliary server determines whether the task set of the data requester is a subset of the task set that a certain worker can execute by calculating the matrix trace, thereby achieving privacy-preserving task matching, and sending the information of the successfully matched worker to the data requester. S16. The data requester filters the successfully matched workers based on their reputation, forming a final set of workers to perform the task, and uploads this set to the blockchain.

[0018] Optionally, in step S2, the submission of privacy-protected data specifically includes the following steps: S21. After the worker confirms that he has been selected from the blockchain, he generates perception data and corresponding random perturbations for each task that needs to be reported. S22. The worker constructs a circular matrix from the perturbated data, the original data, and the perturbated data, and encrypts them using their keys to generate multiple encryption matrices. S23. The worker uploads the encryption matrix, the original sensing data value, and the sum of squared disturbances to the blockchain; S24. The data requester generates an initial predicted value for the true value of the task and the corresponding random perturbation, constructs a cyclic matrix and encrypts it, and sends it to the auxiliary server along with the estimated weighted maximum value and the truth perturbation factor. S25. The auxiliary server generates an over-incrementing sequence and constructs a cyclic matrix encryption. At the same time, it re-encrypts the encryption matrix sent by the data requester and sends all the processed matrices back to the data requester. S26. The data requester pulls the data uploaded by the workers from the blockchain and re-encrypts the matrix received from the auxiliary server locally. S27. The data requester calculates the weight of each worker based on the re-encrypted matrix, using a formula that includes matrix trace operations, while protecting privacy.

[0019] Optionally, in step S3, the reputation update and privacy-preserving truth discovery specifically includes the following steps: S31. Based on the worker weights calculated in Phase 2, the data requester updates the reputation of other workers using the average weight of workers falling into the set of trustworthy workers as a benchmark; the update formula satisfies the following relationship: ; In the formula, This represents the worker's creditworthiness in round t+1. This indicates the worker's creditworthiness in round t. This represents the maximum permissible difference between the weights calculated for workers in the previous stage and the average weights of reliable workers. Indicates the positive learning rate. Indicates the reverse learning rate. Indicates sensitivity; S32. Based on the updated reputation score, add workers with reputation scores above the threshold α to the trusted set and workers with reputation scores below the threshold β to the malicious set, and readjust the weights of each worker in truth discovery accordingly. S33. The data requester uses the truth perturbation factor h, the super-incrementing sequence ciphertext matrix, and the encrypted matrix of worker data to calculate the trace of the weighted sum, and with the assistance of the auxiliary server, iteratively calculates a new round of task truth estimation in a privacy-preserving manner. S34. Repeat the iterative process until the change in the true value estimation for two consecutive rounds is less than the preset threshold, and obtain the final true value of the data.

[0020] Optionally, in step S4, the reward calculation specifically includes: The data requester allocates rewards from the reward pool according to each worker's weight in the final iteration, proportionally to their total weight. The reward calculation formula satisfies the following relationship: ; In the formula, Indicates the final iteration weight. This represents the sum of the weights of all participating workers. This represents the total reward amount corresponding to the total reward pool. This represents the total reward pool.

[0021] In this embodiment, three representative static data fusion algorithms were selected as baselines. These methods integrate multi-source data without dynamic weight adjustments: CRH: A classic iterative fusion algorithm that updates weights based on the deviation between each DP's report and the current true value estimate. Here, we use the version without confidence adjustment, relying solely on error statistics.

[0022] AVE: A simple averaging method that assigns equal weights to all DPs, assuming uniform reliability. It provides a basic benchmark but is highly sensitive to noise.

[0023] MID: A median-based method that mitigates the impact of outliers and extreme values. While robust to noise, it ignores the quality of dynamic programming (DP) and remains vulnerable to large-scale tampering.

[0024] To evaluate the effectiveness of TDRPP in multi-source data fusion tasks within the MCS system, we compared the root mean square error (RMSE) metric with three baseline methods: the traditional CRH algorithm, the simple average method (AVE), and the median method (MID). Four spurious ratios (RF) were set up for the experiments (20%, 35%, 50%, 65%), generating 12 experimental scenarios. For each scenario, six different K values ​​(10, 13, 16, 19, 22, 25) were tested to assess performance trends. The RMSE metric measures the deviation between the final result and the true value; a lower value indicates better performance. Experimental results are as follows: Figure 2 As shown, the method of this invention exhibits superior performance across all configurations. Its RMSE is significantly lower than the other three algorithms. This demonstrates that the method of this invention possesses strong noise resistance and can accurately identify and fuse information sources in noisy environments.

[0025] To evaluate the ability of the method of this invention to identify worker trustworthiness under different prior information in the MCS system, an initial known ratio was introduced, representing the proportion of known benign workers in the initial stage, and the false data ratio R was measured. F The recognition accuracy (IA) within the range {0.2, 0.35, 0.5, 0.65} is defined as the ratio of the number of correctly identified benign digital product (DP) nodes to the total number of benign DP nodes. Experimental results are as follows... Figure 3 As shown.

[0026] As IKR increases from 0.07 to 0.25, the method of the present invention improves in all R values. F All values ​​improved, showing a strong positive correlation. At IKR = 0.07, for R... F =0.2, 0.35, 0.5, 0.65, with values ​​of 0.93, 0.90, 0.93, and 0.91 respectively, showing some fluctuation. When IKR = 0.10, for R... F =0.35 and 0.5, stabilizing at 0.88 and 0.86, indicating robustness. When IKR ≥ 0.13, all R values... F The values ​​below all exceed 0.94, when IKR = 0.16 and R F The method achieves a reliability of 0.94 at IKR = 0.65, indicating effective trust propagation in environments where malicious data exceeds 50%. It converges to 0.95–0.98 when IKR ≥ 0.19, and stabilizes after IKR ≈ 0.2. This invention's method achieves high reliability in high-relationship environments. F The system maintains a high level of trust, leveraging its dynamic trust mechanism to amplify limited prior benign information and filter out malicious workers. This aligns with the law of large numbers, as the behavior of trusted workers tends to stabilize during iterations, thus enabling precise trust aggregation and task allocation.

[0027] This application also provides a privacy-preserving trusted data aggregation system for a swarm intelligence sensing network, which is used to implement a trusted data aggregation method for privacy protection. The system includes: The Data Requester module is used to publish task requirements, initially screen workers, collaboratively conduct privacy-preserving truth discovery and reputation updates, and finally calculate and distribute rewards. The worker client module is used for registration, receiving tasks, collecting sensing data, encrypting and protecting the privacy of task capabilities and sensing data, and uploading them to the blockchain. The blockchain network module, as a decentralized ledger, is used to store encrypted task information, sensing data, intermediate parameters, and transaction records, ensuring the immutability and transparency of the process. The auxiliary server module is configured as a semi-trusted third party, holding a specific key, to perform task matching in encrypted state and assist in completing data aggregation and truth calculation under privacy protection, without knowing the original task information or perceived data of any participant. The key center module is used to generate and securely distribute a complete set of cryptographic keys to all the aforementioned participants during the system initialization phase.

[0028] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. A method for trusted data aggregation in a privacy-preserving crowd-sensing network, characterized in that, include: S1. The key center generates cryptographic keys and distributes them to all participants. In a way that does not reveal the types of tasks that workers can perform, tasks are assigned to matching workers, thus completing the task allocation with privacy protection. S2. The selected worker submits encrypted perception data and related metadata to the blockchain for privacy-preserving data submission. S3. Collaborately execute the truth discovery algorithm in a privacy-preserving manner, iteratively update the worker's reputation, and calculate a more accurate data truth value based on the updated reputation under privacy protection, performing reputation update and privacy-preserving truth discovery. During the truth discovery process, malicious and low-reputation workers are excluded. S4. The data requester calculates the reward based on each worker's contribution and reputation, and distributes it securely to the corresponding worker through the blockchain, thus completing the reward calculation.

2. The trusted data aggregation method for a privacy-preserving crowd-sensing network according to claim 1, characterized in that, In step S1, the allocation of privacy protection tasks specifically includes the following steps: S11. Based on the maximum number of task requests m from the data requester, the key center generates a core encryption matrix, a re-encryption matrix, and a hash function for task privacy mapping, and calculates and generates keys for different purposes to be distributed to the data requester, workers, and auxiliary servers. S12. The worker uses their key to map their set of executable tasks to task numbers through the hash function, and after encrypting the cyclic matrix based on each number, uploads the encryption result to the blockchain. S13. The data requester uses its key to map and construct a polynomial for its n actual tasks and (mn) virtual tasks to be published through the hash function, generates the corresponding cyclic matrix, performs pre-encryption, and uploads the encryption result to the blockchain. S14. The auxiliary server obtains the encryption matrix uploaded by workers and data requesters from the blockchain and performs re-encryption conversion using its own key; S15. In encrypted state, the auxiliary server determines whether the task set of the data requester is a subset of the task set that a certain worker can execute by calculating the matrix trace, thereby achieving privacy-preserving task matching, and sending the information of the successfully matched worker to the data requester. S16. The data requester filters the successfully matched workers based on their reputation, forming a final set of workers to perform the task, and uploads this set to the blockchain.

3. The trusted data aggregation method for a privacy-preserving crowd-sensing network according to claim 1, characterized in that, In step S2, the submission of privacy-protected data specifically includes the following steps: S21. After the worker confirms that he has been selected from the blockchain, he generates perception data and corresponding random perturbations for each task that needs to be reported. S22. The worker constructs a circular matrix from the perturbated data, the original data, and the perturbated data, and encrypts them using their keys to generate multiple encryption matrices. S23. The worker uploads the encryption matrix, the original sensing data value, and the sum of squared disturbances to the blockchain; S24. The data requester generates an initial predicted value for the true value of the task and the corresponding random perturbation, constructs a cyclic matrix and encrypts it, and sends it to the auxiliary server along with the estimated weighted maximum value and the truth perturbation factor. S25. The auxiliary server generates an over-incrementing sequence and constructs a cyclic matrix encryption. At the same time, it re-encrypts the encryption matrix sent by the data requester and sends all the processed matrices back to the data requester. S26. The data requester pulls the data uploaded by the workers from the blockchain and re-encrypts the matrix received from the auxiliary server locally. S27. The data requester calculates the weight of each worker based on the re-encrypted matrix, using a formula that includes matrix trace operations, while protecting privacy.

4. The trusted data aggregation method for a privacy-preserving crowd-sensing network according to claim 1, characterized in that, In S3, the reputation update and privacy protection truth discovery specifically includes the following steps: S31. Based on the worker weights calculated in Phase 2, the data requester updates the reputation of other workers using the average weight of workers falling into the set of trustworthy workers as a benchmark; the update formula satisfies the following relationship: ； In the formula, This represents the worker's creditworthiness in round t+1. This indicates the worker's creditworthiness in round t. This represents the maximum permissible difference between the weights calculated for workers in the previous stage and the average weights of reliable workers. Indicates the positive learning rate. Indicates the reverse learning rate. Indicates sensitivity; S32. Based on the updated reputation score, add workers with reputation scores above the threshold α to the trusted set and workers with reputation scores below the threshold β to the malicious set, and readjust the weights of each worker in truth discovery accordingly. S33. The data requester uses the truth perturbation factor h, the super-incrementing sequence ciphertext matrix, and the encrypted matrix of worker data to calculate the trace of the weighted sum, and with the assistance of the auxiliary server, iteratively calculates a new round of task truth estimation in a privacy-preserving manner. S34. Repeat the iterative process until the change in the true value estimation for two consecutive rounds is less than the preset threshold, and obtain the final true value of the data.

5. The trusted data aggregation method for a privacy-preserving crowd-sensing network according to claim 1, characterized in that, In S4, the reward calculation specifically includes: The data requester allocates rewards from the reward pool according to each worker's weight in the final iteration, proportionally to their total weight. The reward calculation formula satisfies the following relationship: ； In the formula, Indicates the final iteration weight. This represents the sum of the weights of all participating workers. This represents the total reward amount corresponding to the total reward pool. This represents the total reward pool.

6. A trusted data aggregation system for a crowd-aware sensing network for privacy protection, used to implement the method as described in any one of claims 1 to 5, characterized in that, The system includes: The Data Requester module is used to publish task requirements, initially screen workers, collaboratively conduct privacy-preserving truth discovery and reputation updates, and finally calculate and distribute rewards. The worker client module is used for registration, receiving tasks, collecting sensing data, encrypting and protecting the privacy of task capabilities and sensing data, and uploading them to the blockchain. The blockchain network module, as a decentralized ledger, is used to store encrypted task information, sensing data, intermediate parameters, and transaction records, ensuring the immutability and transparency of the process. The auxiliary server module is configured as a semi-trusted third party, holding a specific key, to perform task matching in encrypted state and assist in completing data aggregation and truth calculation under privacy protection, without knowing the original task information or perceived data of any participant. The key center module is used to generate and securely distribute a complete set of cryptographic keys to all the aforementioned participants during the system initialization phase.

Images