CN-121665226-B - Ocean location privacy protection task unloading method based on improved Laplace mechanism

Abstract

The invention discloses a marine location privacy protection task unloading method based on an improved Laplace mechanism. Firstly, a system model considering the serial threat of a multi-buoy server is constructed, a feasible angle interval is determined by combining buoy coverage and transmitting power constraint, and a clipping Laplacian mechanism is used in the constraint area to generate a user pseudo-position with geographic indistinguishability. And secondly, modeling the weighted sum minimization of the unloading energy consumption and the time delay as a discrete optimization problem under the virtual channel condition corresponding to the pseudo position, and realizing task unloading decision by using a gray wolf optimization algorithm. Finally, the signal intensity received by the buoy server is kept consistent with the pseudo position of the user through a marine received signal intensity simulation method based on power control. The invention can effectively reduce the position privacy leakage risk in the task unloading process under the multi-side channel attack scene, and realize safe and reliable ocean task unloading.

Inventors

• LI WENFENG
• LIU SHUAI
• ZHOU YUBIN
• ZHAO KANGLIAN

Assignees

• 南京大学

Dates

Publication Date

20260508

Application Date

20260204

Claims (4)

1. 1. The ocean location privacy protection task unloading method based on the improved Laplace mechanism is characterized by comprising the following steps of: s1, constructing a system model considering the serial threat of a multi-buoy server; S2, determining a feasible angle interval by combining buoy coverage and transmitting power constraint, and generating a user pseudo-position with geographic indistinguishability by using a clipping Laplacian mechanism, wherein the method comprises the following steps of: s2-1, determining a feasible angle interval by combining buoy coverage and transmitting power constraint: In the real position of the ship user Is the center of a circle and the disturbance radius is Generating a disturbance zone, wherein the pseudo positions generated in the disturbance zone need to satisfy the following conditions: 1) Unloading feasibility, namely, the unloading feasibility is that the unloading feasibility is located in the coverage area of at least one buoy server, namely, the unloading feasibility meets the buoy coverage constraint; 2) The received signal strength can imitate that the power area which is required to be far away from the buoy server is required to meet the transmitting power constraint; the angle interval where the pseudo position generated in the disturbance area does not meet the unloading feasibility is recorded as Pseudo position setting and buoy server Is the nearest allowable distance of (2) To take the following steps Is the center of a circle and the radius Making a circle, called a buoy server The angle interval where the disturbance area intersects the power area is recorded as The pseudo-position generated in the section does not satisfy the impersonation of the received signal strength; For each buoy server Can all obtain a group of In order to ensure that the unloading feasibility and the received signal strength impersonation can be simultaneously met under any condition, the unavailable angles are summarized as follows: , Wherein, the , Removing unusable angles from the disturbance area to obtain a feasible angle interval ; S2-2, generating a user pseudo position with geographic indistinguishability by using a clipping Laplace mechanism: To the real position of the ship user As origin of polar coordinate system, pseudo position is generated Expressed as polar coordinates The value range is , Wherein , Is a ray and Included angle of horizontal positive axis, and And (3) with Obeying independent uniform distribution, then generating pseudo-position by clipping Laplacian mechanism The joint probability density of (2) is: Wherein, the Representing privacy budgets with geographic indistinguishability, For the length of the angle interval, the constant is normalized The pseudo position is generated by adopting a process of sampling angles and resampling radii, and the method comprises the following specific steps: 1) Angle sampling: At the position of Is uniformly distributed in the interval Internally randomly generating an angle Wherein Is shown in the interval Obeying uniform distribution; 2) Radius sampling for The edge probability density is: , the cumulative distribution function is: , Wherein, the Is the current integral variable, let According to the inversion sampling principle, obtain The method comprises the following steps: , Wherein the method comprises the steps of The-1 branch of the product log function, and the final pseudo position is written as follows in a right angle system: ; S3, modeling the weighted sum minimization of unloading energy consumption and time delay as discrete optimization problem under the virtual channel condition corresponding to the pseudo position, and realizing task unloading decision by using a gray wolf optimization algorithm; and S4, enabling the strength of the signal received by the buoy server to be consistent with the pseudo position of the user through a marine received signal strength simulation method based on power control.
2. 2. The ocean location privacy preserving task offloading method of claim 1, wherein the step S1 is specifically as follows: S1-1, laying ship users and a buoy server: Laying 1 ship user and in target sea area The station buoy server is characterized in that the real position coordinates of the ship user are as follows Buoy server Is the position coordinates of (a) The set of distances between the vessel user and each buoy server can be expressed as Wherein For Euclidean distance, all distances are assembled into a set Dividing the calculation task of the ship user into a plurality of subtasks Wherein each sub-task is independently and parallelly executed, and for any sub-task Allowing only one processing mode to be selected, i.e. locally executed or offloaded to a single buoy server Definition of For subtasks Is offloaded to If (1) For any buoy server All are true, then the subtasks are represented Local processing, noted as ; S1-2, constructing a multi-buoy server collusion threat model: The buoy server deduces the true position of the ship user according to the task unloading decision and the received signal strength side channel; task offloading decision side channel, first The number of offloading tasks carried by each buoy server in one observation window is: , aggregation Statistical correlation exists between the relative distances between the ship user and each buoy server, and an attacker constructs mapping based on an empirical model Obtaining a position information estimation based on unloading behaviors: , Wherein, the Mapping learned from an empirical model; Received signal strength side channel first The buoy server is in time slot The observed received signal strength is noted as Link gain Expressed as: , Wherein, the For the path gain to be a function of the path gain, As an index of the path loss at the sea surface, Represents the shadow fading caused by the occlusion of sea waves, The instantaneous received signal strength observed by the buoy server is as follows: , Wherein, the The power is transmitted for the user of the vessel, For the bandwidth of the system, For noise power spectral density, attacker pairs Time-averaged cancellation of small-scale fading At this point the attacker can expect the back-thrust distance based on statistics: , , when three or more untrusted buoy servers are in series, an attacker can reversely solve the position of a ship user by using trilateral positioning The task unloading decision side channel and the receiving signal intensity side channel are mutually independent, and an attacker can respectively initiate inference attack to obtain position estimation And 。
3. 3. The ocean location privacy preserving task offloading method of claim 1, wherein the step S3 is specifically as follows: Given the current pseudo-position The ship user and each buoy server Is defined as The corresponding channel gain is The virtual transmission rate at the pseudo-position is: , calculating the unloading energy consumption under the pseudo position according to the virtual transmission rate And time delay Is provided with To unload the balance coefficient of energy consumption and time delay, subtasks are carried out In time slot At the cost of task processing Its value is determined by the pseudo-position Lower offloading decision-making Determining, solving the pseudo-position by using the gray wolf optimization algorithm Optimal offloading decision under The goal is to guarantee offloading decisions and pseudo-locations Minimizing task processing cost on the premise of consistency, regarding each gray wolf as a candidate unloading scheme, and regarding the individual with the minimum fitness value in the population as Wolf, sub-optimal and third optimal individuals are respectively Wolf and Chinese wolf Wolves, the rest are The wolf is used for iteratively updating the positions of all individuals by simulating the trapping and hunting behavior of the wolf, and outputting after the iteration is finished Wolf corresponding offloading decision as pseudo-position The following optimal offloading scheme.
4. 4. The ocean location privacy preserving task offloading method of claim 1, wherein the step S4 is specifically as follows: S4-1, statistical matching of received signal strength under marine channels: when the ship user is in the real position Using camouflage power The expected actual received signal strength of the buoy server is: , while the vessel user is in a pseudo position Using transmit power The expected received signal strength for the buoy server is: , S4-2, marine received signal strength simulation based on power control: camouflage power adjusted by ship user Is required to meet the requirement of being in a pseudo position The expected received signal strength expected from the buoy server is in the true position The observed actual received signal strength is expected to be consistent, let The resulting base mimic power is: , Wherein the method comprises the steps of , When the pseudo-position is far from the buoy service, The required transmit power is reduced, and when the pseudo-position approaches the buoy server, A greater transmit power is required for disguising, and it should be noted that the disguised signal must still ensure link reachability, i.e., the transmit power cannot be below the minimum power lower bound that ensures reliable reception by the buoy server ; Introduction of sea state dependent random disturbance factor Obtaining corrected camouflage power The method comprises the following steps: , Wherein, the Representing the mean value as 1, the disturbance variance of the compensation imitation error as Taking into account the maximum transmit power constraints of the vessel users The final camouflage power adopted is as follows: , And pseudo-position Must be located in an area that satisfies the power constraint, i.e. pseudo-position and buoy server The distance between must satisfy the constraint: , Wherein, the Is the upper limit of the transmit power for the vessel user.

Description

Ocean location privacy protection task unloading method based on improved Laplace mechanism Technical Field The invention belongs to the technical field of privacy protection, and particularly relates to a marine position privacy protection task unloading method based on an improved Laplace mechanism. Background With the comprehensive promotion of intelligent ocean strategy, tasks such as ocean environment monitoring, offshore emergency rescue, ocean scientific investigation and the like provide serious challenges for the real-time performance and reliability of data processing. Due to the relative scarcity of offshore communication infrastructure and the excessive transmission distance of traditional cloud computing centers, it is difficult to meet the instant processing requirements. The use of buoys deployed at the sea surface as edge servers to build maritime mobile edge computing (Mobile Edge Computing, MEC) architecture, providing near computing offload services for vessel users, has become a key solution to the above-mentioned problems. However, the openness of the maritime MEC architecture also carries a serious location privacy risk. In complex maritime application scenarios, the position of a ship user often has high sensitivity, for example, in military cruising and maritime training tasks, the ship position can be seriously threatened once exposed, real-time position leakage of rescue or law enforcement forces can be utilized by an opponent in maritime emergency rescue and law enforcement patrol processes, and key ship track exposure can also cause commercial benefit damage in the fields of marine resource exploration and commercial shipping. Therefore, under the maritime MEC architecture, how to effectively protect the location privacy of the ship user during the task offloading process is a problem that needs to be solved. In the architecture, on one hand, a buoy server is usually operated by a third party, when a ship user unloads tasks to each buoy, the task unloading amount of the buoy server has statistical correlation with the physical distance from each buoy, an untrusted buoy can infer the position of the user through analyzing task distribution to form a task unloading decision side channel, on the other hand, the received signal strength (RECEIVED SIGNAL STRENGTH Indication, RSSI) is monotonically attenuated along with the propagation distance, and a plurality of untrusted buoys can accurately lock the position of the ship user by trilateral positioning based on long-term RSSI observation to form the RSSI side channel. Aiming at the problem of position privacy protection, the prior art generally adopts a differential privacy or simple power control method. For example, patent literature "a location privacy protection method based on differential privacy" (application date: 7/24/2020, application number: 202010143726.6, application publication number: CN111447181 a, the content of which can be cited) implements location data blurring by superimposing laplace noise on a user location and using differential privacy theory, but mainly aims at general location service scenarios, task offloading energy consumption and time delay constraint in edge computation are not considered, RSSI side channels are not protected, and communication coverage constraint is not combined, so that unreachable pseudo locations falling into a communication blind area are easily generated, resulting in offloading failure. The applicant's Guilin university introduces power control in its application patent literature ' a task offloading and transmission power allocation optimizing method for location privacy protection ' (application day: 2024, 3, 6, application number: 202410256537.8, application publication number: CN 1183383580A, the content of which can still be cited) to resist RSSI side channels, but it is designed based on a deterministic path loss model of land communication, does not consider random shadow fading characteristics specific to marine environment, and the generated RSSI signal has obvious deviation from real sea conditions in statistical characteristics and is easily identified by an attacker through statistical analysis, resulting in defending failure. In summary, in the prior art, when facing the high-sensitivity position privacy requirement in a complex maritime environment, there are generally problems that the defending dimension is single, the buoy coverage and the power constraint are not considered in the pseudo position generation, the pseudo is easy to be recognized and the like, and the problem that the pseudo is difficult to be recognized due to channel model mismatch is difficult to effectively guarantee the position privacy of a ship user in a multi-side channel attack scene, so that the real safe and reliable ocean task unloading cannot be realized. Disclosure of Invention The invention aims to solve the problems that the existing position privacy protection method has single d