User selection method and device considering privacy protection in hybrid fog computing

Abstract

The application is suitable for the computer technology field, and provides a user selection method and device considering privacy protection in hybrid crowd sensing, including: obtaining reference utility data, the reference utility data including fuzzy utility corresponding to users participating in historical sensing tasks, the fuzzy utility being utility after privacy processing based on a privacy budget of the user; adjusting a current allocation weight of each user according to the reference utility data to obtain an adjusted allocation weight, the allocation weight reflecting the ability of the user to execute a sensing task; determining a target user from each user based on the adjusted allocation weight; and allocating a current sensing task to the target user to enable the target user to execute the current sensing task. The application can effectively guarantee user privacy security.

Description

Technical Field

[0001] This application belongs to the field of computer technology, and in particular relates to a user selection method and apparatus that takes privacy protection into account in hybrid crowd sensing. Background Technology

[0002] Crowdsourcing sensing refers to a new data acquisition model that combines crowdsourcing concepts with the sensing capabilities of mobile devices. It utilizes users' mobile devices (such as smartphones) to form an interactive, participatory sensing network, distributing sensing tasks to users within the network to achieve efficient data collection at low cost. Hybrid crowdsourcing sensing, in particular, refers to a data acquisition model where users can complete sensing tasks using multiple methods, such as combined vehicles and drones. Because it integrates multiple methods for completing sensing tasks, it is more aligned with real-world applications and has significant potential to leverage the power of the collective. Therefore, it can significantly improve the efficiency and quality of data collection and has promising application prospects.

[0003] Because users' real-time location and other private data may be exposed to the crowdsourcing platform when they participate in perception tasks, privacy protection for user utility is usually carried out through methods such as differential privacy. However, since users can complete perception tasks through multiple perception methods in hybrid crowdsourcing perception, their private data is exposed at different stages or in different scenarios. Therefore, current privacy protection methods are difficult to meet the diverse privacy protection needs in hybrid crowdsourcing perception, resulting in poor protection of user privacy. Summary of the Invention

[0004] This application provides a user selection method and apparatus that considers privacy protection in hybrid crowd sensing, which can effectively protect user privacy and security.

[0005] In a first aspect, embodiments of this application provide a user selection method that considers privacy protection in a hybrid crowd sensing system, including:

[0006] Obtain reference utility data, which includes fuzzy utility corresponding to users participating in historical perception tasks. The fuzzy utility is the utility after privacy processing based on the user's privacy budget.

[0007] The current allocation weights of each user are adjusted based on the reference utility data to obtain the adjusted allocation weights, which reflect the user's ability to perform perceptual tasks.

[0008] Target users are determined from among the users based on the adjusted allocation weights;

[0009] Assign a current perception task to the target user so that the target user can perform the current perception task.

[0010] In one possible implementation of the first aspect, determining the target user from among the users based on the adjusted allocation weights includes:

[0011] For each user, the selection probability corresponding to the user is calculated based on the adjusted allocation weights, the total weights, and the set privacy parameters. The privacy parameters are used to limit the upper limit of performance loss caused by privacy protection, and the total weights are determined based on the adjusted allocation weights corresponding to each user.

[0012] In one possible implementation of the first aspect, calculating the selection probability for each user based on the adjusted allocation weights, the sum of weights, and the set privacy parameters includes:

[0013] For each user, the selection probability corresponding to the user is calculated based on the ratio of the adjusted allocated weight to the sum of the weights, the privacy parameter, and the target number, where the target number is the number of target users to be determined.

[0014] In one possible implementation of the first aspect, adjusting the current allocation weights of each user based on the reference utility data to obtain adjusted allocation weights includes:

[0015] If the user's current assigned weight is less than or equal to a set weight threshold, the set first adjustment algorithm is used to adjust the current assigned weight to obtain the adjusted assigned weight.

[0016] If the user's current assigned weight is greater than the weight threshold, the current assigned weight is adjusted using a set second adjustment algorithm to obtain the adjusted assigned weight, wherein the first adjustment algorithm and the second adjustment algorithm are different.

[0017] In one possible implementation of the first aspect, before adjusting the current allocation weights of each user based on the reference utility data to obtain the adjusted allocation weights, the method further includes:

[0018] A first sum is determined based on the number of first allocation weights and the weight threshold, wherein the first allocation weights include the current allocation weights that are greater than the weight threshold;

[0019] A second sum is determined based on a second allocation weight, wherein the second allocation weight includes the current allocation weight which is less than or equal to the weight threshold.

[0020] The weight threshold is determined based on the first sum, the second sum, the privacy parameter, the target number, and the total number of users.

[0021] In one possible implementation of the first aspect, after assigning the current perceptual task to the target user, the method further includes:

[0022] If the target user is detected to have completed the current perception task, the fuzzy utility corresponding to the target user is obtained;

[0023] For each target user, the current allocation weight of the target user is updated according to the fuzzy utility and privacy parameters to obtain the updated allocation weight.

[0024] In one possible implementation of the first aspect, obtaining the reference utility data includes:

[0025] The fuzzy utility sent by the terminal device corresponding to the user is received. The fuzzy utility is obtained by the terminal device interfering with the user's utility by using any value in the target Laplace distribution. The target Laplace distribution is determined based on the user's privacy budget and the set maximum utility difference.

[0026] The reference utility data is determined based on the received fuzzy utilities.

[0027] Secondly, embodiments of this application provide a user selection device that considers privacy protection in a hybrid crowd sensing system, comprising:

[0028] The utility data acquisition module is used to acquire reference utility data, which includes fuzzy utility corresponding to users participating in the historical perception task. The fuzzy utility is the utility after privacy processing based on the user's privacy budget.

[0029] The allocation weight adjustment module is used to adjust the current allocation weight of each user according to the reference utility data to obtain the adjusted allocation weight, which reflects the user's ability to perform perception tasks.

[0030] The target user determination module is used to determine the target user from among the users based on the adjusted allocation weights;

[0031] The allocation module is used to allocate a current perception task to the target user so that the target user can perform the current perception task.

[0032] Thirdly, embodiments of this application provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the user selection method considering privacy protection in the hybrid crowd sensing described in any of the first aspects above.

[0033] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the user selection method considering privacy protection in the hybrid crowd sensing described in any of the first aspects.

[0034] Fifthly, embodiments of this application provide a computer program product that, when run on an electronic device, causes the electronic device to execute the user selection method for considering privacy protection in the hybrid crowd sensing described in any of the first aspects above.

[0035] The beneficial effects of the embodiments in this application compared with the prior art are:

[0036] In this embodiment, the obtained reference utility data includes the fuzzy utility corresponding to users who participated in historical perception tasks. Since the fuzzy utility corresponding to a user is obtained by privacy processing based on the user's privacy budget—that is, privacy processing based on the user's privacy protection needs—the utility exists in a fuzzy form. Therefore, when obtaining the reference utility data and adjusting the current allocation weights of each user based on the reference utility data, the privacy of users who participated in historical perception tasks is not excessively exposed, effectively protecting user privacy and security. Simultaneously, adjusting the current allocation weights of each user based on the reference utility data allows for a more accurate assessment of each user's current ability to perform the perception task, combined with the performance of users who have participated in the perception task. This results in more accurate adjusted allocation weights, enabling the rapid identification of suitable users to perform the current perception task based on the adjusted allocation weights, which is beneficial for improving the completion quality of the perception task. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0038] Figure 1 This is a flowchart illustrating a user selection method that considers privacy protection in a hybrid crowd sensing system, according to an embodiment of this application.

[0039] Figure 2 This is a schematic diagram of the user selection device that considers privacy protection in the hybrid crowd sensing provided in this application embodiment;

[0040] Figure 3 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0041] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0042] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0043] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0044] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0045] References to "one embodiment" or "some embodiments" in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized.

[0046] Example 1:

[0047] Figure 1 A flowchart illustrating a user selection method considering privacy protection in a hybrid crowd sensing system provided in this application is shown below:

[0048] S101, Obtain reference utility data, which includes fuzzy utility corresponding to users participating in the historical perception task. The fuzzy utility is the utility after privacy processing based on the user's privacy budget.

[0049] The aforementioned sensing tasks refer to data collection tasks based on systems or platforms such as crowd sensing, such as environmental data collection tasks like meteorological or noise data collection.

[0050] The aforementioned utility can reflect information such as the benefits or contributions of users in performing perception tasks. For example, utility can be determined based on data such as the accuracy, speed, breadth, and depth of data in which users complete perception tasks, reflecting the quality of task completion.

[0051] The aforementioned privacy budget can reflect the degree of protection of users' privacy data, or the degree of acceptance by users of the collection, use, and disclosure of their privacy data.

[0052] Privacy processing is used to protect users' privacy data, including but not limited to processing such as interference and encryption.

[0053] It should be understood that historical perception tasks typically refer to completed perception tasks. When obtaining reference utility data, one can obtain the fuzzy utility corresponding to the users who participated in the top N (N greater than 0) historical perception tasks. N can be set or input by the user, or it can be calculated by deep learning models such as large models or other intelligent algorithms.

[0054] It should be noted that the user selection method for considering privacy protection in the hybrid crowd sensing provided in this application embodiment can be applied to hybrid crowd sensing scenarios (i.e., crowd sensing scenarios with multiple sensing methods) or crowd sensing scenarios with a single sensing method.

[0055] It should be understood that when a user has participated in multiple historical perception tasks, the reference utility data obtained may include the fuzzy utility corresponding to each historical perception task in which the user participated.

[0056] In this embodiment, since the fuzzy utility corresponding to the user is obtained by privacy processing of the utility based on the user's privacy budget, and the user's privacy budget can reflect the user's privacy data protection needs, that is, the fuzzy utility is data that is privacy protected based on the user's privacy data protection needs. Therefore, obtaining these fuzzy utilities as reference utility data will not expose the user's privacy too much and can better protect the user's privacy security.

[0057] S102, adjust the current allocation weights of each user based on the above reference utility data to obtain the adjusted allocation weights, which reflect the user's ability to perform perception tasks.

[0058] The aforementioned allocation weights are used to reflect a user's ability to perform a perception task. It should be understood that a user's allocation weight can be determined or adjusted based on data such as the user's relative importance in the perception task allocation or the user's reputation. In some embodiments, the allocation weights can be used to determine the probability of selecting a user to perform a perception task.

[0059] It should be noted that, in the embodiments of this application, the current allocation weight of all users can be adjusted according to the reference utility data, the current allocation weight of each user to be adjusted can be adjusted according to the reference utility data, and the current allocation weight of each user to be adjusted and historical users (the historical users refer to users who participated in the historical perception task) can also be adjusted according to the reference utility data.

[0060] The users to be adjusted can be determined from users other than those who participated in historical perception tasks (let's call them historical users) (let's call them unknown users). For example, unknown users whose similarity to historical users (such as the similarity of user profiles) is greater than or equal to a similarity threshold (such as 0.7) can be considered as users to be adjusted. Adjusting the weighting of unknown users who have not participated in historical perception tasks based on objective reference utility data fully taps into the potential of unknown users to perform perception tasks. This approach effectively explores unknown users and utilizes historical users with known utility, thus improving the accuracy of perception task allocation and the quality of task completion.

[0061] Since reference utility data can reflect information such as the benefits of users who have completed historical perception tasks or their contributions to those tasks, and this data is objective and quantifiable, it can provide a reliable basis for assessing the potential of users who have participated in and those who have not. Therefore, reference utility data can be used as the basis for adjusting the allocation weights. Adjusting the current allocation weights for each user based on the reference utility data can improve the accuracy of the adjusted allocation weights.

[0062] Optionally, when adjusting the current allocation weights of each user based on reference utility data, common characteristics of each historical user can be analyzed. Then, a user score can be calculated based on the similarity between the user's characteristics (such as owned devices and activity levels) and the common characteristics. The user's current allocation weights can then be adjusted based on the user score and the reference utility data. Alternatively, the estimated utility of each user performing the current perception task can be predicted based on the reference utility data, and the current allocation weights of each user can be adjusted based on the estimated utility. The specific implementation method for adjusting the current allocation weights of users based on reference utility data can be set according to the actual application scenario, and this application embodiment does not impose specific limitations on this.

[0063] Optionally, initial allocation weights for users can be preset. These initial allocation weights can be set or input by the platform's management user, or calculated using deep learning models or other intelligent algorithms based on reference information such as user profiles. After obtaining the current perception task to be assigned, it can be determined whether reference utility data exists. If reference utility data exists, the current allocation weights for each user can be adjusted based on the reference utility data. If no reference utility data exists, the target user can be determined from among the users based on the current allocation weights. Alternatively, the current allocation weights for each user can be adjusted based on user-related reference data such as the user's reputation, resulting in an adjusted weight.

[0064] In this embodiment, since the reference utility data is the fuzzy utility corresponding to the historical users who participated in the historical perception task, it can quantify information such as the benefits or contributions of the historical users in performing the historical perception task. Therefore, adjusting the current allocation weight of each user according to the reference utility data can better realize the exploration of unknown users and the utilization of historical users with known utility, effectively explore the potential of each user in performing the current perception task, improve the accuracy of the adjusted allocation weight, and thus help improve the accuracy of the current perception task allocation.

[0065] S103, Based on the adjusted allocation weights mentioned above, determine the target user from among the users.

[0066] It should be understood that when determining the target user from among the users based on the adjusted allocation weight, the target user can be a user whose adjusted allocation weight is greater than or equal to the allocation weight threshold, or the k users whose adjusted allocation weight is sorted in descending order (k is greater than 1) can be the target user, or k users can be randomly selected from the users whose adjusted allocation weight is greater than or equal to the allocation weight threshold. The specific method for determining the target user can be set according to the actual application scenario, and no specific restrictions are made here.

[0067] S104, Assign the current perception task to the target user so that the target user can perform the current perception task.

[0068] Optionally, the current sensing task can be assigned to the target user by sending the current sensing task to the terminal device corresponding to the target user or by other means.

[0069] It should be understood that once a target user is assigned a current sensing task, they can complete that task in any way, such as by walking, driving, or using a drone.

[0070] In this embodiment, the obtained reference utility data includes the fuzzy utility corresponding to users who participated in historical perception tasks. Since the fuzzy utility corresponding to a user is obtained by processing its utility based on the user's privacy budget—that is, processing the utility based on the user's privacy protection needs—the utility exists in a fuzzy form. Therefore, when obtaining the reference utility data and adjusting the current allocation weights of each user based on the reference utility data, the privacy of users who participated in historical perception tasks is not excessively exposed, effectively protecting user privacy and security. Simultaneously, adjusting the current allocation weights of each user based on objective and quantifiable reference utility data allows for a more accurate assessment of each user's ability to perform the current perception task, combined with the performance of users who have participated in the perception task. This results in more accurate adjusted allocation weights, enabling the rapid and accurate identification of suitable users to perform the current perception task based on the adjusted allocation weights, which is beneficial for improving the completion quality of the perception task.

[0071] In some embodiments, S101 includes:

[0072] The aforementioned fuzzy utility is received from the terminal device corresponding to the user. The fuzzy utility is obtained by the terminal device interfering with the user's utility using any value in the target Laplace distribution. The target Laplace distribution is determined based on the user's privacy budget and the set maximum utility difference.

[0073] The aforementioned reference utility data is determined based on the received fuzzy utilities.

[0074] To fully protect user privacy and security, in this embodiment, the user's corresponding terminal device performs privacy processing on the user's utility based on the user's privacy budget, resulting in fuzzy utility. When obtaining reference utility data, the user's corresponding terminal device can be notified to report the fuzzy utility through methods such as sending a notification, so that the required reference utility data can be determined based on the fuzzy utility sent by each user's corresponding terminal device.

[0075] Optionally, the maximum utility difference can be set or input by the user, or it can be calculated by a deep learning model or other intelligent algorithms. Alternatively, it can be calculated based on the utility of all users participating in one or more historical perception tasks, reflecting the maximum difference between the utilities of these users.

[0076] For example, the maximum utility difference could be the maximum difference calculated based on the utility of all users who participated in the previous historical perception task.

[0077] To reduce the impact of privacy protection on data availability, when performing privacy processing on the user's utility, the terminal device can first randomly determine an interference value (also known as a noise value) based on the target Laplace distribution, and use this noise value to interfere with the user's utility to obtain the noisy utility, i.e., fuzzy utility.

[0078] Since the target Laplace distribution is determined based on the user's privacy budget and the maximum utility difference set, using any value in the target Laplace distribution as noise to interfere with the user's corresponding utility can effectively control the degree of noise's impact on utility. That is, while performing privacy processing according to the user's utility protection needs to achieve personalized privacy protection, the impact of privacy protection on utility availability can be controlled, ensuring the accuracy of subsequent weight adjustments based on fuzzy utility.

[0079] In some embodiments, the target Laplace distribution can be a Laplace distribution with a mean (also known as the location parameter) of 0 and a scale (also known as the scale parameter) of the target ratio, which can be the ratio of the maximum utility difference to the user's privacy budget.

[0080] In some embodiments, fuzzy utility can be expressed in the following form:

[0081] π p (u i ,∈ i )=u i +φ i ,φ i ~Lap(0, Δ u / ∈ i )

[0082] Where, π p (u i ,∈ i ) represents the fuzzy utility corresponding to user i, u i Represents the utility corresponding to user i, ∈ i φ represents the privacy budget for user i. i Let Lap(0, Δ) represent the interference value corresponding to user i. u / ∈ i ) represents the Laplace distribution of the target, φ i ~Lap(0, Δ u / ∈ i ) represents the interference value φ i It is a value determined based on the target Laplace distribution, Δ u / ∈ i Δ represents the scale of the target Laplace distribution. u This represents the difference in maximum utility.

[0083] In this embodiment, the user's utility is processed for privacy based on a privacy budget through the user's corresponding terminal device. The utility obtained by the platform is a fuzzy utility that exists in an obfuscated form, so that the user's utility will not be directly exposed to devices other than the corresponding terminal device, thus ensuring the security of user privacy.

[0084] In some embodiments, step S102 includes:

[0085] If the user's current assigned weight is less than or equal to the set weight threshold, the current assigned weight is adjusted according to the aforementioned reference utility data and the set first adjustment algorithm to obtain the adjusted assigned weight.

[0086] If the user's current assigned weight is greater than the weight threshold, the current assigned weight is adjusted according to the reference utility data and the set second adjustment algorithm to obtain the adjusted assigned weight. The first adjustment algorithm and the second adjustment algorithm are different.

[0087] It should be understood that the weight threshold can be set or input by the user, or it can be calculated by deep learning models such as large models or other intelligent algorithms.

[0088] To control the weights assigned to each user and reduce situations where some users' weights are too high or too low, when adjusting the current weights assigned to each user based on reference utility data, different adjustment algorithms can be used to adjust the weights based on the user's current weight relative to a weight threshold (e.g., 0.6). This results in adjusted weights, allowing for reasonable control of the adjusted weights through appropriate algorithms. This helps avoid situations where some users' weights are extremely high or low, thereby improving the fairness and accuracy of the perception task allocation.

[0089] For example, the first adjustment algorithm can be used to make a small adjustment to the current allocation weight (e.g., the adjustment is less than or equal to the first ratio, such as 10%), and the second adjustment algorithm can be used to make a large adjustment to the current allocation weight (e.g., the adjustment is less than or equal to the second ratio, such as 30%, where the second ratio is greater than the first ratio).

[0090] Optionally, the first adjustment algorithm can be used to adjust the user's current assigned weight, such as increasing, decreasing, or keeping it unchanged, based on the weight difference between the weight threshold and the current assigned weight. For example, if the weight difference is greater than the difference threshold (e.g., 0.2), the first adjustment algorithm can be used to increase the user's current assigned weight, that is, to appropriately increase the user's current assigned weight based on the reference utility data and the first adjustment algorithm; if the weight difference is less than or equal to the difference threshold, the first adjustment algorithm can be used to decrease the user's assigned weight, that is, to appropriately decrease the user's current assigned weight based on the reference utility data and the first adjustment algorithm, or, the user's assigned weight can be kept unchanged.

[0091] Optionally, the second adjustment algorithm can be used to reduce the user's current assigned weight. When the user's current assigned weight is greater than a weight threshold, i.e., when it is determined that the user's current assigned weight is too large, the user's current assigned weight is adjusted in a decreasing direction according to the reference utility data and the second adjustment algorithm, so that the adjusted assigned weight is less than or equal to the weight threshold, thereby reducing the imbalance and unfairness of the perceived task allocation.

[0092] In some embodiments, the second adjustment algorithm can be represented as follows:

[0093]

[0094] Among them, w i This represents the current assigned weight for user i. This represents the adjusted allocation weight for user i. This represents the adjusted allocation weight for user i whose current allocation weight is greater than the weight threshold. In other words, the first adjustment algorithm is used to adjust the current allocation weight that is greater than the weight threshold to the weight threshold.

[0095] In this embodiment, a weight threshold is set. When adjusting the user's current assigned weight based on reference utility data, different adjustments are made based on the size of the user's current assigned weight relative to the weight threshold. This can tap the potential of each user to perform the current perception task, improve the accuracy of the adjusted assigned weight, and effectively control the balance of the adjusted assigned weight, avoiding extreme values ​​such as maximum or minimum values, thereby improving the fairness and accuracy of perception task allocation.

[0096] In some embodiments, prior to step S102 described above, the method further includes:

[0097] The first sum is determined based on the number of first allocation weights and the aforementioned weight threshold, wherein the first allocation weights include the current allocation weights that are greater than the aforementioned weight threshold.

[0098] The second sum is determined based on the second allocation weight, wherein the second allocation weight includes the current allocation weight which is less than or equal to the weight threshold.

[0099] The aforementioned weight thresholds are determined based on the first sum, the second sum, the privacy parameters, the target number, and the total number of users.

[0100] The aforementioned privacy parameters are used to limit the upper limit of performance loss caused by privacy protection. Optionally, these privacy parameters can be set or entered by platform administrators, or calculated by deep learning models or other intelligent algorithms based on each user's privacy budget.

[0101] To control the balance of user allocation weights while controlling the range of performance losses (such as computational efficiency or data availability) caused by privacy protection, a weight threshold for allocation weights can be calculated by combining privacy parameters, the current allocation weights of each user, the number of target users to be determined, and the total number of users.

[0102] It should be noted that since the weight threshold is currently an unknown value, the determined first sum and second sum are also unknown values. The first sum and second sum can be represented by the weight threshold. Then, by combining the privacy parameters, the target number, and the total number of users, the specific value of the weight threshold can be calculated according to the set calculation rules.

[0103] In some embodiments, the first sum can be expressed in the following form:

[0104]

[0105] Among them, w sum1 Let w represent the first sum, λ represent the weight threshold, and w represent the weight threshold. i This represents the current assigned weight for user i. This represents the sum of current assigned weights that are greater than the weight threshold, where m represents the number of current assigned weights (i.e., the first assigned weight) that are greater than the weight threshold.

[0106] The second sum can be expressed in the following form:

[0107]

[0108] Among them, w sum2 This represents the second sum, and ∑ represents the cumulative sum operation. This represents the sum of the currently assigned weights that are less than or equal to the weight.

[0109] In some embodiments, the defined calculation rules can be expressed in the following form:

[0110]

[0111] Where γ represents the privacy parameter, k represents the target number, N represents the total number of users, and w sum1 w represents the first sum. sum2 This indicates the second sum.

[0112] Corresponding to the above calculation rules, the weight threshold can be expressed in the following form:

[0113]

[0114]

[0115] It should be noted that if no weight threshold satisfying the above expression exists, the current allocation weights of all users can be directly adjusted based on the reference utility data and the first adjustment algorithm to obtain the adjusted allocation weights. Alternatively, if a weight threshold satisfying the preset conditions exists, the current allocation weights of users can be adjusted based on the reference utility data; if no weight threshold satisfying the above preset conditions exists, the current allocation weights of users do not need to be adjusted. The calculation rule set is the calculation rule expressed in the above expression, that is, the weight threshold is less than or equal to the value calculated based on the first sum, the second sum, the privacy parameter, and the target quantity.

[0116] In some embodiments, step S103 includes:

[0117] For each user, the selection probability corresponding to the user is calculated based on the adjusted allocation weights, the total weights, and the set privacy parameters. The privacy parameters are used to limit the upper limit of performance loss caused by privacy protection, and the total weights are determined based on the adjusted allocation weights for each user.

[0118] The target users are determined from among the users based on the selection probabilities described above.

[0119] Since noise in utility may affect its availability, thus having an unpredictable impact on processing such as updating the allocation weights, and these impacts will continue to accumulate in multiple rounds of perception tasks, affecting the accuracy of subsequent target user selection, when determining the target user from each user based on the user's adjusted allocation weights, the adjusted allocation weights of the user can be quantified based on privacy parameters to obtain the user's corresponding selection probability, and then the target user can be determined from each user based on the selection probability.

[0120] In some embodiments, the above privacy parameters can be expressed in the following form:

[0121]

[0122] Where γ represents the privacy parameter, Δ u Indicates the maximum utility difference. This represents the overall privacy requirements, ln represents the logarithm (base e), and T represents the task allocation round corresponding to the current sensing task. It should be understood that one sensing task corresponds to one task allocation round; that is, one round typically allocates one sensing task.

[0123]

[0124] Where, ∈ i This represents user i's privacy budget.

[0125] In some embodiments, when determining the target user from among the users based on the selection probability, a dependency-based random selection strategy can be used to select k users as the target users. The aforementioned dependency-based random selection strategy refers to iteratively rounding the selection probability of each user, while ensuring that the rounded selection probability satisfies a predetermined marginal distribution (e.g., ...). This represents the selection probability after random rounding and a cardinality constraint. Through the above processing, by introducing a user privacy budget and a dependency-based random selection strategy, we can maximize the expected cumulative utility of the perception task while protecting user privacy, thereby improving the accuracy of perception task allocation and the quality of task completion.

[0126] Alternatively, the cardinality constraint described above can be expressed in the following form:

[0127]

[0128] That is, the sum of the random rounding probabilities of each user's selection equals the target number.

[0129] In this embodiment, when determining the target user from among all users based on the user's adjusted allocation weight, a privacy parameter restriction is introduced. The probability of a user being selected is calculated based on the adjusted allocation weight and the sum of all adjusted allocation weights. Then, the target user is determined from among all users based on the calculated selection probability. This method can accurately calculate the selection probability while limiting the performance loss caused by privacy protection, thereby improving the accuracy of subsequent target user determination.

[0130] In some embodiments, the calculation of the selection probability for each user based on the adjusted assigned weights and the sum of weights includes:

[0131] For each user, the selection probability corresponding to the user is calculated based on the ratio of the adjusted assigned weights to the sum of the weights, the privacy parameters, and the target number, where the target number is the number of target users to be determined.

[0132] In some embodiments, the probability of a user's choice can be expressed in the following form:

[0133]

[0134] Where, p i This represents the probability of choice for user i. This represents the adjusted allocation weight for user i, and k represents the target number. This represents the adjusted allocation weight for user j. This represents the sum of the adjusted weights for N users, i.e., the total weights. N is usually the number of all users.

[0135] In this embodiment, the probability of a user being selected is quantified based on the user's adjusted assigned weight relative to the total weight and the number of target users to be selected. At the same time, privacy parameters are introduced to limit the selection, which can ensure the fairness and accuracy of the selection probability while protecting user privacy.

[0136] In some embodiments, after step S104 described above, the method further includes:

[0137] If the target user is detected to have completed the current perception task, the fuzzy utility corresponding to the target user is obtained.

[0138] For each of the aforementioned target users, the current allocation weights of the target users are updated based on the aforementioned fuzzy utility and privacy parameters to obtain the updated allocation weights.

[0139] Since utility can intuitively reflect the benefits or contributions of a target user in performing the current perception task, when a target user completes the current perception task, the fuzzy utility corresponding to the target user can be obtained. Then, for each target user, the current allocation weight is updated according to its corresponding fuzzy utility and the set privacy parameters to obtain the updated allocation weight. This allows the updated allocation weight to more accurately reflect the user's ability to perform the current perception task, while also helping to incentivize users to actively complete the perception task and improve the quality of task completion.

[0140] It should be noted that in this embodiment, the current allocation weight of the target user is updated based on the fuzzy utility and privacy parameters corresponding to the target user. That is, the allocation weight of the user is updated only based on the user's own utility and privacy parameters. When adjusting the current allocation weight of the user, the current allocation weight of each user is adjusted according to the fuzzy utility of all users participating in the historical perception task.

[0141] In some embodiments, the updated allocation weights can be represented in the following form:

[0142]

[0143] Among them, w i ′ represents the updated assigned weight, w i The weight assigned to user i is the current weight (usually the adjusted weight obtained in step S102), e is the natural constant, k represents the number of targets, γ represents the privacy parameter, and u′ represents the target value. i Let p represent the fuzzy utility corresponding to target user i, N represent the total number of users, and p represent the total number of users. i This represents the selection probability corresponding to target user i.

[0144] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0145] Example 2:

[0146] Corresponding to the user selection method that considers privacy protection in the hybrid crowd sensing described in the above embodiments, Figure 2 This diagram illustrates the structure of a user selection device that considers privacy protection in a hybrid crowd sensing system provided in this application embodiment. For ease of explanation, only the parts relevant to this application embodiment are shown.

[0147] Reference Figure 2 The device includes: a utility data acquisition module 21, an allocation weight adjustment module 22, a target user determination module 23, and an allocation module 24. Among them,

[0148] The utility data acquisition module 21 is used to acquire reference utility data, which includes the fuzzy utility corresponding to the user who participated in the historical perception task. The fuzzy utility is the utility after privacy processing based on the user's privacy budget.

[0149] The weight allocation adjustment module 22 is used to adjust the current weight allocation of each user based on the above reference utility data to obtain the adjusted weight allocation, which reflects the user's ability to perform perception tasks.

[0150] The target user determination module 23 is used to determine the target user from among the users based on the adjusted weights assigned above.

[0151] The allocation module 24 is used to allocate the current perception task to the target user so that the target user can perform the current perception task.

[0152] In this embodiment, the obtained reference utility data includes the fuzzy utility corresponding to users who participated in historical perception tasks. Since the fuzzy utility corresponding to a user is obtained by processing its utility based on the user's privacy budget—that is, processing the utility based on the user's privacy protection needs—the utility exists in a fuzzy form. Therefore, when obtaining the reference utility data and adjusting the current allocation weights of each user based on the reference utility data, the privacy of users who participated in historical perception tasks is not excessively exposed, effectively protecting user privacy and security. Simultaneously, adjusting the current allocation weights of each user based on objective and quantifiable reference utility data allows for a more accurate assessment of each user's ability to perform the current perception task, combined with the performance of users who have participated in the perception task. This results in more accurate adjusted allocation weights, enabling the rapid and accurate identification of suitable users to perform the current perception task based on the adjusted allocation weights, which is beneficial for improving the completion quality of the perception task.

[0153] In some embodiments, the weighting adjustment module 22 includes:

[0154] The probability calculation unit is used to calculate the selection probability of each user based on the adjusted allocation weights, the total weights, and the set privacy parameters. The privacy parameters are used to limit the upper limit of performance loss caused by privacy protection, and the total weights are determined based on the adjusted allocation weights for each user.

[0155] The selection unit is used to determine the target user from among the users based on the selection probability mentioned above.

[0156] In some embodiments, the weighting adjustment module 22 includes:

[0157] The probability calculation unit is used to calculate the selection probability of each user based on the ratio of the adjusted allocated weights to the sum of the weights, the privacy parameters, and the target number, where the target number is the number of target users to be determined.

[0158] In some embodiments, the weighting adjustment module 22 includes:

[0159] The first adjustment unit is used to adjust the current allocation weight by using a set first adjustment algorithm when the user's current allocation weight is less than or equal to a set weight threshold, so as to obtain the adjusted allocation weight.

[0160] The second adjustment unit is used to adjust the current allocation weight by using a set second adjustment algorithm when the user's current allocation weight is greater than the weight threshold, so as to obtain the adjusted allocation weight. The first adjustment algorithm and the second adjustment algorithm are different.

[0161] In some embodiments, the user selection device for considering privacy protection in the above-described hybrid crowd sensing further includes:

[0162] The first sum calculation module is used to determine the first sum based on the number of first allocation weights and the aforementioned weight threshold, wherein the first allocation weights include the current allocation weights that are greater than the aforementioned weight threshold.

[0163] The second sum calculation module is used to determine the second sum based on the second allocation weight, wherein the second allocation weight includes the current allocation weight that is less than or equal to the weight threshold.

[0164] The weight threshold calculation module is used to determine the weight threshold based on the first sum, the second sum, the privacy parameter, the target number, and the total number of users.

[0165] In some embodiments, the user selection device for considering privacy protection in the above-described hybrid crowd sensing further includes:

[0166] The target acquisition module is used to acquire the fuzzy utility corresponding to the target user when the target user is detected to have completed the current perception task.

[0167] The weight allocation update module is used to update the current weight allocation of each of the above target users based on the fuzzy utility and privacy parameters, so as to obtain the updated weight allocation.

[0168] In some embodiments, the utility data acquisition module 21 includes:

[0169] The receiving unit is used to receive the aforementioned fuzzy utility sent by the terminal device corresponding to the user. The fuzzy utility is obtained by the terminal device interfering with the user's utility using any value in the target Laplace distribution. The target Laplace distribution is determined based on the user's privacy budget and the set maximum utility difference.

[0170] The utility data determination unit is used to determine the aforementioned reference utility data based on the received fuzzy utilities.

[0171] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0172] Example 3:

[0173] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 3As shown, the electronic device 3 of this embodiment includes: at least one processor 30 ( Figure 3 The diagram shows only one processor, a memory 31, and a computer program 32 stored in the memory 31 and executable on the at least one processor 30, wherein the processor 30 executes the computer program 32 to implement the steps in any of the above method embodiments.

[0174] The electronic device 3 can be a desktop computer, laptop, handheld computer, or cloud server, etc. This electronic device may include, but is not limited to, a processor 30 and a memory 31. Those skilled in the art will understand that... Figure 3 This is merely an example of electronic device 3 and does not constitute a limitation on electronic device 3. It may include more or fewer components than shown in the figure, or combine certain components, or different components. For example, it may also include input / output devices, network access devices, etc.

[0175] The processor 30 may be a Central Processing Unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0176] In some embodiments, the memory 31 may be an internal storage unit of the electronic device 3, such as a hard disk or memory of the electronic device 3. In other embodiments, the memory 31 may be an external storage device of the electronic device 3, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the electronic device 3. Furthermore, the memory 31 may include both internal and external storage units of the electronic device 3. The memory 31 is used to store the operating system, applications, bootloader, data, and other programs, such as the program code of the computer program. The memory 31 can also be used to temporarily store data that has been output or will be output.

[0177] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0178] This application also provides a network device, which includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement the steps in any of the above method embodiments.

[0179] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, can implement the steps in the above-described method embodiments.

[0180] This application provides a computer program product that, when run on an electronic device, enables the electronic device to implement the steps described in the various method embodiments above.

[0181] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a photographic device / electronic device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0182] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0183] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0184] In the embodiments provided in this application, it should be understood that the disclosed apparatus / network devices and methods can be implemented in other ways. For example, the apparatus / network device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0185] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0186] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A user selection method considering privacy protection in hybrid crowd sensing, characterized in that, include: Obtain reference utility data, which includes fuzzy utility corresponding to users participating in historical perception tasks. The fuzzy utility is the utility after privacy processing based on the user's privacy budget. The current allocation weights of each user are adjusted based on the reference utility data to obtain the adjusted allocation weights, which reflect the user's ability to perform perceptual tasks. Target users are determined from among the users based on the adjusted allocation weights; Assign a current perception task to the target user, so that the target user can perform the current perception task; The step of adjusting the current allocation weights of each user based on the reference utility data to obtain the adjusted allocation weights includes: If the user's current assigned weight is less than or equal to the set weight threshold, the current assigned weight is adjusted according to the reference utility data and the set first adjustment algorithm to obtain the adjusted assigned weight. If the user's current allocation weight is greater than the weight threshold, the current allocation weight is adjusted according to the reference utility data and the set second adjustment algorithm to obtain the adjusted allocation weight. The adjustment range of the first adjustment algorithm on the current allocation weight is less than or equal to a first ratio, and the second adjustment algorithm is used to reduce the user's current allocation weight, and the adjustment range of the current allocation weight is less than or equal to a second ratio, which is greater than the first ratio.

2. The user selection method considering privacy protection in hybrid crowd sensing as described in claim 1, characterized in that, The process of determining the target user from among the users based on the adjusted allocation weights includes: For each user, the selection probability corresponding to the user is calculated based on the adjusted allocation weights, the total weights, and the set privacy parameters. The privacy parameters are used to limit the upper limit of performance loss caused by privacy protection, and the total weights are determined based on the adjusted allocation weights corresponding to each user. The target user is determined from among the users based on the selection probability.

3. The user selection method considering privacy protection in hybrid crowd sensing as described in claim 2, characterized in that, The step of calculating the selection probability for each user based on the adjusted allocation weights, the sum of weights, and the set privacy parameters includes: For each user, the selection probability corresponding to the user is calculated based on the ratio of the adjusted allocated weight to the sum of the weights, the privacy parameter, and the target number, where the target number is the number of target users to be determined.

4. The user selection method considering privacy protection in hybrid crowd sensing as described in claim 3, characterized in that, Before adjusting the current allocation weights of each user based on the reference utility data to obtain the adjusted allocation weights, the method further includes: A first sum is determined based on the product of the number of first allocation weights and the weight threshold, wherein the first allocation weights include the current allocation weights that are greater than the weight threshold, and the value of the first sum is equal to the sum of the individual first allocation weights. A second sum is determined based on the cumulative value of each second allocation weight, wherein the second allocation weight includes the current allocation weight that is less than or equal to the weight threshold; The weight threshold is determined based on the first sum, the second sum, the privacy parameter, the target number, and the total number of users. The weight threshold is: in, This represents the weight threshold. Indicates the first sum. Indicates the second sum. Indicates the number of weights allocated in the first assignment. This represents the current assigned weight for user i. This represents the cumulative value of each of the second allocation weights. This represents a privacy parameter, and this represents the target quantity. This indicates the total number of all users.

5. The user selection method considering privacy protection in hybrid crowd sensing as described in claim 1, characterized in that, After assigning the current perception task to the target user, the method further includes: If the target user is detected to have completed the current perception task, the fuzzy utility corresponding to the target user is obtained; For each target user, the current allocation weight of the target user is updated according to the fuzzy utility and privacy parameters to obtain the updated allocation weight.

6. The user selection method considering privacy protection in hybrid crowd sensing as described in any one of claims 1 to 5, characterized in that, The acquisition of reference utility data includes: The fuzzy utility sent by the terminal device corresponding to the user is received. The fuzzy utility is obtained by the terminal device interfering with the user's utility by using any value in the target Laplace distribution. The target Laplace distribution is determined based on the user's privacy budget and the set maximum utility difference. The reference utility data is determined based on the received fuzzy utilities.

7. A user selection device that considers privacy protection in a hybrid crowd sensing system, characterized in that, include: The utility data acquisition module is used to acquire reference utility data, which includes fuzzy utility corresponding to users participating in the historical perception task. The fuzzy utility is the utility after privacy processing based on the user's privacy budget. The allocation weight adjustment module is used to adjust the current allocation weight of each user according to the reference utility data to obtain the adjusted allocation weight, which reflects the user's ability to perform perception tasks. The target user determination module is used to determine the target user from among the users based on the adjusted allocation weights; The allocation module is used to allocate a current perception task to the target user, so that the target user can perform the current perception task; The weighting adjustment module includes: The first adjustment unit is used to adjust the current allocation weight according to the reference utility data and the set first adjustment algorithm when the user's current allocation weight is less than or equal to a set weight threshold, so as to obtain the adjusted allocation weight. The second adjustment unit is configured to adjust the current allocation weight according to the reference utility data and a set second adjustment algorithm when the user's current allocation weight is greater than the weight threshold, thereby obtaining an adjusted allocation weight. The adjustment magnitude of the current allocation weight by the first adjustment algorithm is less than or equal to a first ratio, and the second adjustment algorithm is configured to reduce the user's current allocation weight, with the adjustment magnitude being less than or equal to a second ratio, where the second ratio is greater than the first ratio.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 6.

9. A computer program product, characterized in that, When the computer program product is run on an electronic device, it causes the electronic device to perform the method as described in any one of claims 1 to 6.

Images