Smart home temperature control system opacity strengthening method based on label time Petri network

Abstract

The invention belongs to the technical field of information security, and relates to a smart home temperature control system opacity strengthening method based on a tag time Petri network. Establishing a label time Petri network model of the intelligent household temperature control system; constructing a partial correction state class diagram based on time sequence observation of which the current state is not transparent, forming a path constraint model consistent with observation logic, and identifying a path set inconsistent with observation time; screening a path which meets the opaqueness of the current state relative to the secret identifier set in the path set, and constructing a linear programming model to solve a minimum time sequence disturbance function value; and updating the original time sequence observation by using the perturbation function, and generating a new observation sequence meeting the opaqueness of the current state. According to the method, an attacker is effectively prevented from deducing the real state of the system through a time sequence, meanwhile, exhaustive search on a full-state space is avoided, the calculation complexity is reduced, and the applicability and expandability of the method in a complex discrete event system are improved.

Description

Technical Field

[0001] This invention belongs to the field of information security technology, specifically relating to a method for enhancing the opacity of a smart home temperature control system based on a tag-time Petri net. Background Technology

[0002] With the widespread adoption of IoT technology, smart home systems have become deeply integrated into daily life. These systems achieve adaptive environmental control through the collaborative work of sensors and actuators; for example, a temperature control system can automatically adjust its operation based on user habits, real-time environmental data, and remote commands. However, while providing convenience, this system also brings significant privacy and security risks. The system's internal state (such as whether the user has left home) can be inferred by external attackers through observable event sequences (such as sensor trigger records), leading to the leakage of sensitive information such as user habits and home status.

[0003] To address such information flow security issues, the field of discrete event systems (DES) has proposed opacity as a formal property of confidentiality. Its core principle is to ensure that the secrets within the system (such as a specific set of system states) remain ambiguous to external intruders who can only observe a subset of events—that is, for any truly secret trajectory, there exists a non-secret trajectory with an identical external observation sequence, making it impossible for the intruder to confirm whether the secret has occurred. Current state opacity, as an important type, requires that intruders cannot determine whether the system is currently in a secret state based on historical observations. Existing research has established relatively complete theoretical frameworks for verifying and enforcing current state opacity on models such as finite automata and general Petri nets, and has strengthened the confidentiality of non-opaque systems by implementing behavioral modifications through supervised control or event editing (such as insertion, deletion, and replacement).

[0004] However, existing technologies have significant shortcomings when directly applying the above theories to real-time physical information fusion systems such as smart homes. First, the behavior of smart home systems has inherent time constraints (such as actuator response delays and the timing requirements for command activation), while traditional DES models typically ignore the time dimension. Attackers can easily eliminate ambiguity by combining the precise time of event occurrence, thereby inferring the secret state with higher confidence. This makes the opacity verification based on non-time models potentially ineffective in practice. Second, existing opacity enhancement mechanisms (such as event editing) are mostly designed at the abstract logical level, failing to fully consider the physical constraints of smart home systems. For example, directly inserting or deleting temperature control events may disrupt system stability; while simply delaying event observation may violate the performance requirements of real-time monitoring and response. Therefore, there is an urgent need for a new method that can systematically confuse attackers through refined means (such as controllable perturbation of event timestamps) while adhering to the inherent time and logical constraints of the system, thereby achieving the goal of opacity enhancement.

[0005] Existing opacity enhancement methods primarily focus on perturbing the logical behavior of the system, such as inserting, deleting, or replacing observable events, or altering the order of events, to confuse the inference abilities of external observers. However, these methods typically introduce significant interference with the system's original behavior, easily affecting the system's functional correctness, real-time performance, and stability, and are particularly unsuitable for time-sensitive temperature control systems with strict physical constraints. Furthermore, some methods require enumeration analysis of the system's complete state space, resulting in high computational complexity and making it difficult to support online or near real-time applications. Summary of the Invention

[0006] This invention provides a technical solution that enhances system opacity by finely controlling timing observations without compromising system logic and physical feasibility, thereby effectively protecting user privacy while ensuring system functionality and engineering feasibility.

[0007] To address the aforementioned technical problems, this invention proposes a method for enhancing the opacity of a smart home temperature control system based on tag-time Petri nets. The specific technical solution includes the following steps:

[0008] S1: Establish a tagged time Petri net model for the smart home temperature control system, and formalize the sensing, decision-making and execution behaviors in the system as transitions and locations with logical labels and time constraints, thereby simultaneously characterizing the causal relationship and time evolution characteristics of the system.

[0009] S2: Based on the logical label sequence in the current opaque time series observation, calculate the partially corrected state class diagram of the system, and construct the constraint problem of the path set consistent with the time series observation logic, and then solve the path set that is inconsistent with the current time series observation in the time dimension.

[0010] S3: Search the set of paths in the set of inconsistent time paths where the current state of the secret identifier set is not transparent, and construct a linear programming problem under the constraints of maintaining the consistency of system logic and time feasibility, and solve for the corresponding time-series perturbation function value;

[0011] S4: Update the original time series observations using the time series perturbation function value to generate a new time series observation sequence that satisfies the current state opacity, so that external observers cannot determine whether the system is in a secret state based on the observation data.

[0012] Compared with the prior art, the present invention has at least the following beneficial effects:

[0013] (1) A more refined and covert enhancement of opacity is achieved. By introducing the dimension of time to perturb the system observation, this invention avoids direct modification of the logical events themselves, thereby effectively obfuscating the external inference process while maintaining the surface consistency of system behavior, and improving the covertness and security of privacy protection.

[0014] (2) The physical feasibility and functional integrity of the system are guaranteed. This invention only adjusts the observation timestamp, without interfering with the actual execution process of the system, and does not affect the original timing and logical relationship of sensor sampling, control calculation and execution actions, so as not to damage the core functional indicators such as system stability, real-time performance and temperature control comfort.

[0015] (3) It has high computational efficiency and engineering feasibility. By constructing a partially modified state class diagram to avoid full state space enumeration, and using linear programming to solve key parameters, the overall computational complexity is controllable, providing a feasible technical foundation for online or near real-time deployment on home gateways or edge servers.

[0016] (4) Enhanced security with greater precision and concealment: By introducing time as a key dimension, this invention can protect the privacy of logs or communication traffic generated by smart home systems that contain precise timestamps. The resulting temporal perturbation is a non-intrusive security enhancement method. Compared with traditional methods that directly insert, delete, or replace logical events, it deviates less from the original behavior of the system and is less likely to be detected by attackers, thus maintaining higher concealment while enhancing confidentiality.

[0017] (5) Ensuring the physical feasibility and functional integrity of the system: The core of the enhancement mechanism proposed in this invention is to adjust the observed timestamps, rather than interfering with the actual operating logic of the system or delaying the execution of real events. This means that all sensor sampling, controller calculation, and actuator actions of the temperature control system occur in the original designed time and logical order, and the system's stability, real-time performance, and comfort control objectives are not affected in any way. This method achieves a good balance between privacy protection and system function assurance, and has high practicality.

[0018] (6) Achieves efficient computation and feasible online application prospects: The method avoids full state space enumeration by constructing a partially modified state class diagram and uses linear programming, a mature and efficient optimization tool, to solve for key parameters (time-series perturbation function values). This makes the entire analysis and computation process have feasible time complexity, providing a technical foundation for achieving online, near real-time privacy protection on home gateways or edge servers with certain computing capabilities in the future, and enabling dynamic response to the continuously generated system observation data streams. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating an embodiment of the method for enhancing the opacity of a smart home temperature control system based on a tag-time Petri net, as described in this invention.

[0020] Figure 2 This is a schematic diagram of an intelligent home temperature control system, representing an embodiment of the method for enhancing the opacity of an intelligent home temperature control system based on a tag-time Petri net, according to the present invention.

[0021] Figure 3 This is a schematic diagram of a tag-time Petri net model, representing an embodiment of the tag-time Petri net-based method for enhancing the opacity of a smart home temperature control system according to the present invention.

[0022] Figure 4 This is a schematic diagram of the MSCG portion of the tag-time Petri net in an embodiment of the method for enhancing the opacity of a smart home temperature control system based on tag-time Petri nets according to the present invention. Detailed Implementation

[0023] To make the objectives and advantages of this invention clearer, the invention will be specifically described below with reference to embodiments. It should be understood that the following text is merely used to describe one or more specific embodiments of the invention and does not strictly limit the scope of protection specifically claimed by the invention.

[0024] This invention provides a method for enhancing the opacity of smart home temperature control systems based on tagged temporal Petri nets, aiming to solve the problem that existing technologies cannot accurately analyze and enhance the confidentiality of physical information fusion systems with strict time constraints. The method first uses tagged temporal Petri nets to formally model the smart home temperature control system, accurately characterizing the system state, events (such as sensor triggering and command execution), and their time constraints. The core of this method lies in that, for a time-series observation (containing the logical sequence of events and their timestamps) verified as having a transparent current state (i.e., potentially revealing a secret state), this invention does not simply modify or disable events, but rather obfuscates attackers by calculating a time-series perturbation. Specifically, the method first obtains a set of system paths consistent with the logic of the time-series observation, and filters out paths inconsistent with the original observation time by solving a constraint problem. Then, it further searches within this set for paths that satisfy the current state opacity requirement and constructs a linear programming problem to find the optimal time-series perturbation function value. This function value indicates how to make the minimum and safest adjustment to the event timestamps in the original observation. Finally, this time-series perturbation is applied to update the observation, generating a new time-series observation. For an external attacker, this new observation corresponds to a time pattern that is consistent with the operating trajectory of at least one non-secret system. This makes it impossible for the attacker to be certain whether the system is currently in a secret state. Thus, without changing the actual physical behavior and logical instructions of the system, the opacity of the system's current state is effectively enhanced.

[0025] like Figure 1 As shown, a method for enhancing the opacity of a smart home temperature control system based on tag-time Petri nets includes the following steps:

[0026] S1: Establish a tag-time Petri net model for a smart home temperature control system;

[0027] S2: Based on the partially modified state class diagram of the logical label sequence calculation system in the current opaque time series observation, construct the constraint problem of the path set consistent with the time series observation logic, and solve the path set inconsistent with the time of the current opaque time series observation.

[0028] S3: Search for a set of paths whose current state is not transparent to the secret identifier set within the set of paths whose current state is not transparent to the path observation time; construct a linear programming problem to solve for the time perturbation function value.

[0029] S4: Update the time series observations using the time series perturbation function value to obtain new time series observations that satisfy the current state opacity.

[0030] This invention uses tagged temporal Petri nets to formally model the operational behavior of smart home temperature control systems. By perturbing and controlling the observable temporal sequence of the system, it protects the secret state of the system, thereby enhancing the system's opacity and privacy security. Its basic working principle is as follows.

[0031] First, in step S1, a corresponding labeled time Petri net model is constructed based on the temporal and logical relationships between various functional modules (such as temperature acquisition, control decision and execution units) in the smart home temperature control system. By introducing logical labels and time constraints into the transitions, the model can simultaneously characterize the causal relationships and temporal evolution characteristics of the system behavior, thus providing a formal basis for subsequent temporal analysis and opacity control.

[0032] Secondly, in step S2, for the time-series observations generated by the system in its current operating state, the logical label sequence is extracted, and a partially corrected state class diagram of the system is calculated accordingly. This state class diagram depicts the set of states that the system may be in under given observation conditions. Furthermore, by constructing a constraint problem of the path set consistent with the logic of the time-series observations, and solving for the path set that is inconsistent with the current time-series observations in the time dimension, distinguishable paths that may potentially reveal secret states are identified, providing candidate adjustment objects for enhancing opacity.

[0033] Then, in step S3, paths involving the secret identifier set are further filtered from the aforementioned set of inconsistent time paths. A linear programming problem is constructed, constrained by maintaining system logical consistency and time feasibility, to solve for the required temporal perturbation function. This temporal perturbation function describes the magnitude and manner of minimizing the adjustment of the observation time without changing the system's logical behavior, thereby achieving the concealment of the secret state.

[0034] Finally, in step S4, the original time-series observations are updated based on the solved time-series perturbation function values ​​to generate a new time-series observation sequence. This prevents external observers from determining whether the system is in a secret state solely based on the observation sequence, thereby enhancing the opacity of the current state. Through the above process, this invention effectively improves the privacy protection and security performance of the smart home temperature control system at the time-series level without affecting the system's normal control functions.

[0035] In S1, the label-time Petri net consists of triples. It means that, among them Representative containing collections and Change Set Petri nets; It is a change Assign two nonnegative rational numbers and Time function, and express The static emission interval, i.e., the time range in which the event occurs; It is for each change Specify a tag Set of symbols (corresponding to observable transitions) or empty strings Label functions (corresponding to unobservable transitions).

[0036] In S1, the identifier of the tag-time Petri net is represented as follows: ,in Represents the places of Petri nets The token number in M; the set of transitions enabling M is represented as A time transition sequence is represented as Its logical transition sequence is as follows ;from Launch a time-varying sequence Generate reachable identifier The process is denoted as .

[0037] In S2, the time-series observations of the labeled time Petri net system are represented as follows: Its meaning is at a certain moment Observed time-stamped sequences ; The logical label sequence is .

[0038] In S2, the modified state class graph (MSCG) of the label-time Petri net system is a directed graph whose nodes are state classes. ,in For reachability indicators, Edges connected to the state class are marked as " ",in Indicates from The target class to the edge The permitted launch interval.

[0039] In S2, in a portion of the MSCG, and The relevant path is defined as follows:

[0040] (1)

[0041] in, and .

[0042] In S2, a Petri net system with given labels and time-series observations that are opaque to the current state are used. The relevant part of MSCG is related to The logically consistent set of paths is:

[0043] (2)

[0044] in This is the current identifier for a portion of MSCG.

[0045] In S2, the Petri net system with given labels and the time-series observations with opaque current states are... in and Then with A time-consistent path set is defined as:

[0046] (3)

[0047] in yes The logical transition sequence; and The set of paths with inconsistent times is represented as ;and Time-consistent identifier set is represented as .

[0048] In S2, a Petri net system with given labels and time series observations are used. and a set of secret identifiers If with Time-consistent identifier sets satisfy Then time series observation It's about a set of secret symbols. The current status is opaque.

[0049] In S3, given the labeled time Petri net system and the time-series observations where the current state is opaque. Then the path set whose current state is opaque regarding the secret identifier set is represented as:

[0050] (4)

[0051] In S3, given the labeled time Petri net system and the time-series observations where the current state is opaque. ,in and Timing perturbation function It is a time series observation and Tags in Allocate observation time and The function of timing perturbation. The solution algorithm includes the following steps:

[0052] Input: Tag-Time Petri Net System Time series observations with opaque current state and with The relevant part is MSCG;

[0053] S31: Calculation and The relevant part is MSCG; ;

[0054] S32: Order For all of the conditions in part of the MSCG path A set;

[0055] S33: For all paths , construct constraint (5); if constraint (5) cannot be satisfied, then ;

[0056] S34: In Find the set of paths whose current state is opaque regarding the secret identifier set. ;

[0057] S35: For all paths Construct LPP(6) computation ,in , This allows us to obtain each path. Timing perturbation cost ,in and ;

[0058] S36: For all paths ,calculate Finally, the time-series perturbation function value is obtained. );

[0059] Output: New time series observations .

[0060] In step S33, the satisfiability of the time constraint (5) related to the transition is determined to obtain the result. A set of paths with inconsistent time. That is, if constraint (5) has no feasible solution, then the path... and The timing is inconsistent.

[0061] (5)

[0062] Finally, due to Time-inconsistent path sets It can be concluded that... Inconsistent set of identifiers .

[0063] In S35, each path The time-series perturbation cost is expressed as ,in and To find the path The minimum order perturbation cost is calculated by constructing the following LPP first. :

[0064] (6)

[0065] Then, the path can be derived. Timing perturbation cost .

[0066] In S36, for all paths Find the minimum perturbation cost, and find the cost that makes the cost Minimum observation time and In this way, the time-series perturbation function value is obtained. ).

[0067] In step S4, the timing observation is updated using the timing perturbation function value. In order to obtain a time-series observation that satisfies the opacity of the current state. ,in and .

[0068] This invention provides a method for enhancing the opacity of smart home temperature control systems based on tag-time Petri nets. This method offers privacy protection for smart home temperature control systems and improves system security. The code implementation of the timing perturbation function algorithm can be generated using Python and run in the PyCharm interpreted environment, or it can be implemented using other software tools such as Matlab; no limitation is made here. Furthermore, the timing perturbation function algorithm used in this invention is not limited to the embodiments provided in this method and can also be applied to other embodiments.

[0069] In actual operation, this invention, such as Figure 1 As shown, in this embodiment of the invention, the method for enhancing the opacity of a smart home temperature control system based on a tag-time Petri net includes the following steps:

[0070] S1: Establish a tag-time Petri net model for a smart home temperature control system;

[0071] S2: Based on the partially modified state class diagram of the logical label sequence calculation system in the current opaque time series observation, construct the constraint problem of the path set consistent with the time series observation logic, and solve the path set inconsistent with the time of the current opaque time series observation.

[0072] S3: Search for a set of paths whose current state is not transparent to the secret identifier set within the set of paths whose current state is not transparent to the path observation time; construct a linear programming problem to solve for the time perturbation function value.

[0073] S4: Update the time series observations using the time series perturbation function value to obtain new time series observations that satisfy the current state opacity.

[0074] When a couple leaves home for work at 9 a.m., the system will shut down within 20 to 30 seconds (change). ), and is in a closed state (warehouse) At the end of the workday, i.e., 6 PM, the smart thermostat will activate 10 to 20 seconds before running (change). Switching from the off state to the standby state (warehouse) When the system is in standby mode and detects someone, the system will (change) within 10 to 20 seconds. Switch to the open state (warehouse) Conversely, when the system is on and no one is detected, in order to save energy, the system will (change) within 10 to 20 seconds. Switch to standby mode (warehouse) Facial recognition monitors can operate within 5 to 11 seconds (variable). The system identifies a user, specifically in three scenarios: when the system identifies the husband (change). The temperature will be set to a suitable temperature for him, which will be based on the warehouse. Modeling is performed; when the system identifies the wife (transition) The temperature will be set to her preferred temperature (stock). When the system identifies someone who is neither husband nor wife, it sets the temperature to a different value (transition). Facial recognition monitors typically respond within 10 to 26 seconds (this varies). Two users were identified. If both spouses are present in the room (warehouse) at the same time... And it was identified by the system (transition) The system will set the temperature to an appropriate value (in the library) based on their preferences. If no one adjusts the selected temperature within 20 to 30 seconds (change) , , and The temperature will then be determined and saved (in the warehouse). Then it intelligently adjusts to the selected temperature. Figure 3 In the labeled time Petri net model shown, the observable transition set is Their labels are respectively , , , The remaining transitions are unobservable. The initial identifier of the system is... .

[0075] The following is combined with Figure 3 , Figure 4 Step S2 in this implementation will be explained.

[0076] Suppose an intruder is observing the activity of a smart thermostat. If the wife is alone, there is a potential risk of intrusion; the secret lies in the wife being home alone, characterized by... When the wife returned home from get off work, the intruder was positioned at the right time. Observed ,Right now Then it can be calculated with Related parts of MSCG, such as Figure 4 As shown.

[0077] Based on this time series observation, and starting from the initial state class with an observable projection, The path set associated with the logical transition sequence is

[0078] By testing all the above paths, it can be seen that... and Consistent time, that is So, with A consistent set of identifiers is .because For intruders, time series observation It's not that the current state is opaque. Therefore, an intruder can observe secret information and deduce that only the wife is home. Furthermore, due to... and Can be launched

[0079] .

[0080] The following is combined with Figure 3 , Figure 4 Step S3 in this implementation will be explained.

[0081] because None of the paths point to the marker belonging to State classes, therefore have According to the algorithm, it should be... For each path, a constraint set is constructed to solve for the corresponding time-series perturbation function value. Consider the path... Solve the following LPP:

[0082] (7)

[0083] It can be known ,in , , , .

[0084] Similarly, the set can be obtained. Solution for all paths: , , , , , ,as well as Then, find the minimum perturbation cost for all paths: , , , , , ,as well as Choose from those that satisfy path And obtain the time perturbation function value. ), ),as well as ).

[0085] The following is combined with Figure 3 , Figure 4 Step S4 in this implementation will be explained.

[0086] Based on the aforementioned time-series perturbation function values, a modified time-series observation can be obtained. ,in and .because Established, this observation is for The current state is opaque.

[0087] Therefore, the present invention provides a method for enhancing the opacity of a smart home temperature control system based on a tag-time Petri net, which solves the problems in the background technology, reduces computational complexity, and makes it more suitable for complex systems.

[0088] Example 1: System Modeling and State Class Construction Based on Tag-Time Petri Nets

[0089] This embodiment models a discrete-event system comprising a temperature control host, indoor sensor nodes, and actuators. First, the system is abstracted as a Petri net structure containing a set of places and a set of transitions. Places represent the system's operating and resource states, while transitions represent logical events that change between states. Each transition is assigned an emission time interval to describe the allowed time range for the event to occur. Logical labels are assigned to each observable transition, while empty labels are assigned to unobservable transitions. This method of constructing a labeled-temporal Petri net model allows for a unified description of the system's structure, time constraints, and observable behavior.

[0090] Based on this, a partially modified state class graph is constructed according to the initial identifier and time constraints. Each state class consists of a reachability identifier and a set of time constraints, with edges marked by transitions, corresponding logical labels, and allowed emission time intervals. This state class graph enables the enumeration of reachable paths in the system while satisfying time constraints, and provides the basic data structure for subsequent logically consistent path search and time consistency judgment.

[0091] Example 2: Logically Consistent Path Search Based on Time Series Observations

[0092] This embodiment processes a set of time-series observations generated during system operation. Each observation consists of a logical label sequence and its corresponding observation time, representing a sequence of observable events observed at a given time point. The system searches the state class graph for all paths whose logical label sequences match the observed sequence, filtering out paths with mismatched logical labels to obtain a set of candidate paths consistent with the observation logic.

[0093] This set of paths allows for further analysis of all potential labeled states the system might have been in when the observation was generated, providing a basis for subsequent determination of whether all of these states are secret states. This process is independent of specific system applications, relying only on the labeled temporal Petri net model and the observation sequence, thus possessing universality.

[0094] Example 3: Path Time Consistency Determination and Time Inconsistency Path Identification

[0095] After obtaining the set of logically consistent paths, corresponding time constraints are established for each path, including launch time interval constraints for each transition, path time accumulation constraints, and observation time matching constraints. By determining whether there are feasible solutions to these constraints, paths that cannot simultaneously satisfy all time constraints are identified as time-inconsistent paths.

[0096] This process separates logically consistent but temporally unrealizable paths from the candidate set, enabling the system to accurately distinguish which paths are temporally inconsistent with the current observations, thus providing input data for the subsequent construction of the perturbation function.

[0097] Example 4: Determining the Opacity of the Current State

[0098] Check if there exists at least one identifier in the set of reachable identifiers corresponding to the time-consistent path that does not belong to the secret identifier set. If such an identifier exists, it means that the observer cannot determine whether the system is in a secret state, thus the system satisfies the current state opacity.

[0099] If all time-consistent paths correspond to identifiers that belong to the secret identifier set, it indicates that the system has leaked secret state information under the current observation, and perturbation processing is required. This determination mechanism can be directly applied to real-time privacy protection and state hiding scenarios.

[0100] Example 5: Candidate Perturbation Path Screening

[0101] From the set of time-inconsistent paths, those paths terminating at non-secret identifiers are selected as candidate perturbation paths. This set represents the paths through which the system can reintroduce non-secret possible states with a small time perturbation, thereby restoring the system's opacity.

[0102] This screening process ensures that perturbation operations always serve the goal of privacy restoration without introducing new privacy leakage paths, thereby improving system security.

[0103] Example 6: Solving a linear programming problem with minimum perturbation cost

[0104] For each candidate perturbation path, a linear programming model is constructed, with the launch time of each transition in the path as a variable, and time interval constraints, path cumulative time constraints, and observation time constraints are applied. The objective function is the sum of the difference between the path termination time and the current observation termination time, as well as the time differences corresponding to each logical label.

[0105] By solving this linear programming model, the minimum time perturbation cost required for each path is obtained, which is used to measure the amount of modification needed to adjust the path to match new observations, thereby minimizing the perturbation.

[0106] Example 7: Optimal Perturbation Path Selection and Perturbation Function Generation

[0107] The minimum perturbation cost of all candidate perturbation paths is compared, and the path with the minimum cost is selected as the optimal perturbation path. New logically labeled observation times and new observation termination times are generated based on the optimal time variable values ​​corresponding to this path.

[0108] This results in a time-series perturbation function from the original observation to the new observation, which guarantees that the modification magnitude is minimized and satisfies the time consistency constraint and the opacity requirement.

[0109] Example 8: Time-series observation update and closed-loop reinforcement

[0110] The existing observations are updated using a timing perturbation function, replacing the observation time and termination time of the logical labels to generate new timing observations. The system then re-executes the opacity determination based on the new observations.

[0111] If the opacity requirement is still not met, the perturbation process is repeated until the opacity requirement is met, thus forming a closed-loop reinforcement mechanism to achieve automatic privacy protection adjustment.

[0112] Example 9: Application in smart home temperature control and privacy protection

[0113] In smart home temperature control systems, temperature changes, device startup and shutdown are abstracted as observable transitions, while the internal regulation logic is abstracted as unobservable transitions. Attackers can attempt to infer whether a resident is at home or engaged in a specific activity by observing the behavior of the temperature control device.

[0114] By fine-tuning the observation time using the above method, attackers cannot uniquely determine that the system is in a certain secret state, thereby protecting the privacy of residents' behavior.

[0115] Example 10: Application to other time-constrained discrete event systems

[0116] This method is also applicable to scenarios with time constraints and observable events, such as production scheduling systems, traffic control systems, and network protocol systems, to prevent observers from inferring sensitive internal states of the system through time series.

[0117] This generality ensures that the method does not depend on specific business models, but only on formal modeling and temporal perturbation mechanisms, thus possessing good promotional value and engineering applicability.

[0118] The above description is merely a preferred embodiment of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention. Structures, devices, and operating methods not specifically described or explained in this invention are implemented according to conventional methods in the art unless otherwise specified or limited.

Claims

1. A method for current state opacity path decision based on labeled time Petri nets, characterized in that, The method comprises the following steps: A label timed Petri net model is established, which comprises a place set and a transition set, and a firing time interval and a logical label are set for each transition; A partially revised state class diagram is constructed according to the time sequence observation, and the state class diagram is composed of reachable labels and corresponding time constraint sets; A path set in which a logical label sequence is consistent with the time sequence observation is searched in the state class diagram; A label set in which time is inconsistent with the time sequence observation is calculated according to the path set; Whether the label set in which time is inconsistent contains a non-secret label is judged, so as to determine the path set of the current state opacity of the secret label set.

2. The method of claim 1, wherein, The label timed Petri net is represented by a triple, which comprises a Petri net structure, a time interval function and a label function.

3. The method of claim 1, wherein, The node of the partially revised state class diagram is a state class, each state class is composed of a reachable label and a set of time constraints about enabled transitions, and the edge is marked by a transition label, a logical label and an allowed firing time interval.

4. A method for constructing a timing perturbation function based on path time inconsistency, characterized in that, The method comprises the following steps: A time inconsistent path set is constructed by identifying a path in which a time constraint cannot be satisfied in the path set consistent with the time sequence observation logic; A candidate path set is screened by filtering a path reaching a non-secret label in the time inconsistent path set; A linear programming problem is constructed for each candidate path, and a minimum time disturbance cost for making the path satisfy the time constraint is calculated; The path with the minimum disturbance cost is selected from the candidate path set, and a new observation time is determined as the output of the time sequence disturbance function according to the path.

5. The method of claim 4, wherein, The time inconsistent path is determined by the fact that the time variable on the path cannot simultaneously satisfy the transition firing time interval constraint and the observation time constraint.

6. The method of claim 4, wherein, The minimum time disturbance cost is determined by the difference between the observation termination time and the path termination time and the sum of the differences between the observation time of each logical label and the firing time of the corresponding transition.

7. A timing perturbation function based system current state opacity reinforcement method, characterized in that, The method comprises the following steps: An initial time sequence observation of a system is obtained; Whether the time sequence observation satisfies the current state opacity is judged according to the method in any one of claims 1 to 3; When the current state opacity is not satisfied, a time sequence disturbance function is calculated according to the method in any one of claims 4 to 6; The time sequence observation is updated by using the time sequence disturbance function to generate a new time sequence observation; The current state opacity is judged again based on the new time sequence observation until the current state opacity is satisfied.

8. The method of claim 7, wherein, The time sequence disturbance function is a mapping function of labels and observation times, which is used to assign a new observation time to each logical label and determine a new observation termination time.

9. The method of claim 7, wherein, The updated time sequence observation is realized by replacing the observation time and the termination time corresponding to the original logical label sequence.

10. The method of claim 7, wherein, The method is applied to the privacy protection, information hiding or state security control scene of a time constraint discrete event system.

Images