A perception-transmission cooperative optimization method based on task value joint estimation

Abstract

The application discloses a kind of based on task value joint estimation's perception-transmission coordination optimization method, it is related to industrial monitoring field.The application constructs unified task value evaluation framework, the measurement information utility brought by measurement update and the transmission lag cost caused by transmission time delay are included in the same index system, by quantifying the time-effect contribution of sensor measurement data in transmission process with time evolution attenuation, establish the task value evaluation model for remote state estimation.Based on the task value index constructed by the application, further joint optimization sensor scheduling and wireless network resource allocation, under the constraint of limited communication resources, the transmission of high task value data is prioritized to be guaranteed, the collaborative design of perception decision and transmission decision is realized, so as to effectively reduce long-term state estimation error, improve the overall operation performance and resource utilization efficiency of industrial wireless monitoring system.

Description

Technical Field

[0001] This invention relates to the field of industrial monitoring, and in particular to a sensing-transmission collaborative optimization method based on joint estimation of task value. Background Technology

[0002] With the rapid development of smart manufacturing and Industrial Internet of Things (IIoT) technologies, industrial monitoring systems are placing higher demands on the real-time performance, accuracy, and resource utilization efficiency of remote state estimation. In typical industrial joint monitoring scenarios, multiple distributed sensors collaboratively observe the same dynamic process and upload measurement data to a remote estimator via a wireless network to achieve system state awareness. However, industrial wireless networks are generally limited by factors such as spectrum resources, transmission power, and bandwidth capacity, making it difficult to support simultaneous data transmission from all sensors within the same sampling period. Therefore, it is necessary to design a reasonable sensor scheduling mechanism to select some key sensors for data upload under limited communication resources to ensure the performance of remote state estimation.

[0003] The performance of remote state estimation depends not only on the quality of sensor measurement information but also significantly on the wireless transmission process. On one hand, different sensor observations contribute differently to system state estimation; identifying and prioritizing the transmission of high-value measurement information is crucial for improving estimation performance. On the other hand, resource contention in wireless networks leads to increased transmission latency, preventing measurement data from reaching the remote estimator in a timely manner, thus prolonging the prediction process and exacerbating transmission lag costs. Therefore, how to comprehensively consider the utility of measurement information and transmission lag costs under limited communication resources, and achieve collaborative optimization between the sensing and transmission layers, has become a critical issue that urgently needs to be addressed in the field of industrial remote state estimation.

[0004] Existing technologies in sensor scheduling research have explored aspects such as information value, information freshness, and information gain. However, they still exhibit a relatively disconnect between "information value modeling" and "transmission timeliness modeling." On the one hand, most methods focus on characterizing the contribution of measurement information to estimation performance, without incorporating actual transmission delay under wireless network conditions into the error propagation analysis of remote state estimation. On the other hand, existing schemes that introduce information age or communication delay mostly remain at the level of information freshness description or constraint handling, and have not yet formed a unified evaluation mechanism that can simultaneously reflect the utility of measurement information and the cost of transmission lag. Furthermore, sensor scheduling and wireless network resource allocation are usually designed independently, lacking a collaborative optimization framework across the sensing and transmission layers. Therefore, it is urgent to propose a new evaluation index that can uniformly measure the utility of measurement information and the cost of transmission lag, and based on this, to achieve joint optimization of sensor scheduling and wireless network resource allocation, thereby improving the performance of remote state estimation in resource-constrained networks.

[0005] Therefore, those skilled in the art are dedicated to developing a sensing-transmission collaborative optimization method based on joint estimation of task value. This invention proposes a sensing-transmission collaborative optimization method based on joint estimation of value-of-task (VoT). This method constructs a unified task value evaluation framework, incorporating the utility of measurement information brought about by measurement updates and the transmission lag cost caused by transmission delay into the same index system. By quantifying the comprehensive contribution of sensor measurement data to the remote state estimation task, a task value evaluation model for remote state estimation is established. Based on the constructed task value index, this invention further jointly optimizes sensor scheduling and wireless network resource allocation, prioritizing the transmission of high task value data under the constraint of limited communication resources. This achieves collaborative design of sensing and transmission decisions, thereby effectively reducing long-term state estimation errors and improving the overall operational performance and resource utilization efficiency of industrial wireless monitoring systems. Summary of the Invention

[0006] In view of the above-mentioned deficiencies of the prior art, the technical problem to be solved by the present invention is to realize the joint optimization of sensor scheduling and wireless network resource allocation, and improve the remote state estimation performance in resource-constrained networks.

[0007] To achieve the above objectives, this invention provides a sensing-transmission collaborative optimization method based on joint estimation of task value, including a sensor scheduling evaluation mechanism based on task value. The evaluation mechanism is used to uniformly quantify the utility of measurement information brought about by measurement updates and the transmission lag cost caused by transmission delay, so as to characterize the combined impact of the two on the performance of remote state estimation.

[0008] Furthermore, the evaluation mechanism is based on the evolution process of remote state estimation error, and decomposes the role of the sensor scheduling scheme into two parts: measurement information utility and transmission lag cost. For any candidate sensor set, the measurement matrix and measurement noise determine the update effect on the estimation error, which is used to characterize the measurement information utility. The transmission delay generated under the current network resource conditions determines the time for the estimator to make prior estimations based on system state transitions before the data arrives, which is used to characterize the transmission lag cost.

[0009] Furthermore, the task value index is defined as the error propagation matrix. The reciprocal of the spectral radius ,in , Represents the state transition matrix. Indicates transmission delay The transmission lag cost caused by system state transitions; Represents the identity matrix. Indicates Kalman gain, Represents the measurement matrix. This indicates the utility of measurement information resulting from measurement updates.

[0010] Furthermore, a perception-transmission collaborative design framework is constructed based on task value indicators, which incorporates sensor scheduling and network resource allocation into a unified optimization process. When evaluating the candidate sensor set, the utility of measurement information and the cost of transmission lag are considered simultaneously, and the selected sensors, the number of subcarriers, and the transmission power are jointly determined.

[0011] Furthermore, for any candidate sensor set, the measurement matrix and measurement noise determine the set's contribution to the remote estimation update; meanwhile, the subcarrier and transmit power configuration of the set under given communication constraints determines the achievable transmission delay.

[0012] Furthermore, for each candidate sensor set, the collaborative design framework first calculates the minimum transmission delay under communication constraints, then combines the measurement update effect of the set to calculate the task value, and compares different candidate solutions based on the task value.

[0013] Furthermore, a hierarchical task value-driven solution algorithm is included, which decomposes the perception-transmission joint optimization problem into two nested sub-problems: outer sensor selection and inner network resource optimization. The outer layer is responsible for constructing a set of candidate sensors, while the inner layer is responsible for determining feasible network resource configurations given the set of candidate sensors. This transforms the original combined optimization problem involving both sensor scheduling and network resource allocation into a hierarchical optimization structure that can be executed online.

[0014] Furthermore, the hierarchical task value-driven solution algorithm includes the following steps: Step 1: Initialize the selected sensor set It is an empty set; Step 2: From the set of sensors that are not currently selected Select sensors in sequence Construct a candidate sensor set .in, Represents the entire set of sensors; Step 3: For the candidate sensor set By invoking network resource optimization methods, the minimum achievable transmission latency is obtained. ; Step 4: Based on the candidate sensor set Measurement information utility and minimum achievable transmission delay Calculate the value of the task ; Step 5: After evaluating all candidate sensor sets, select the candidate sensor set with the highest task value. As a new set of selected sensors ; Step 6: Determine the selected sensor set Has the number of sensors reached the maximum number of selectable sensors? If the target is not reached, return to step 2; if the target is reached, output the final sensor scheduling result. And the corresponding network resource allocation scheme.

[0015] Furthermore, step 3 includes the following steps: Step 3.1: For the candidate sensor set Acquire sensor information and network resource parameters, and set a lower bound for transmission delay search. Search upper bound and setting precision ; Step 3.2: Take the median value of the current search interval. As a candidate transmission delay, according to Calculate the candidate sensor set The minimum transmission rate required by each sensor is determined, and the transmit power of each sensor is set to the maximum value. Step 3.3, in the candidate transmission delay Next, a resource feasibility check is performed to search for the minimum number of subcarriers required for each sensor to meet the signal-to-noise ratio constraints and minimum transmission rate requirements. If all sensors have corresponding And the total number of subcarriers required does not exceed the total number of available subcarriers. Then determine Feasible; otherwise, determine Not feasible; Step 3.4: Update the binary search interval based on the feasibility assessment results; if If feasible, then Update to the upper bound of the search and record the current resource allocation results; if If it is not feasible, then Updated to the lower bound of the search; Step 3.5: Determine if the search interval meets the termination condition; if Greater than the set precision If so, return to step 3.2; if Not greater than the set precision If the search stops, the candidate sensor set is obtained. The corresponding minimum achievable transmission delay And network resource allocation scheme.

[0016] Furthermore, step 4 includes the following steps: Step 4.1: Construct a candidate sensor set The corresponding error propagation matrix ,in Indicates transmission delay The transmission lag cost caused by system state transitions; Represents the set of candidate sensors The corresponding Kalman gain, Represents the set of candidate sensors The corresponding measurement matrix, Indicates the selection of candidate sensor set The utility of measurement information resulting from measurement updates; Step 4.2: Calculate the error propagation matrix spectral radius ; Step 4.3: Calculate the sensor set Task value .

[0017] In resource-constrained, sensor-node-dense industrial joint monitoring scenarios, existing sensor scheduling methods for remote state estimation struggle to coordinate the trade-off between measurement information utility and transmission lag costs. Especially under conditions of limited network resources, selecting sensors solely based on measurement information utility may result in measurement data being unusable for timely state updates due to significant transmission delays; conversely, selecting sensors solely based on transmission delay may lead to insufficient observational information. Existing methods often separate sensor scheduling from network resource allocation, making it difficult to uniformly characterize the coupled impact of these two factors on remote estimation performance, thus affecting the accuracy and real-time performance of the estimation. The essence of these problems lies in the lack of a unified evaluation mechanism that can simultaneously characterize both "measurement information utility" and "transmission lag costs." To address these issues, this invention proposes a sensor scheduling evaluation mechanism based on Value-of-Task (VoT). This mechanism quantifies the measurement information utility from measurement updates and the transmission lag costs caused by transmission delays in a unified manner, characterizing their combined impact on remote state estimation performance, thereby achieving a comprehensive evaluation of sensor scheduling schemes under conditions of limited network resources. The evaluation mechanism of this invention is based on the evolution of remote state estimation error, decomposing the role of the sensor scheduling scheme into two parts: measurement information utility and transmission lag cost. Specifically, for any candidate sensor set, the measurement matrix and measurement noise determine its effect on updating the estimation error, which characterizes the measurement information utility; the transmission delay generated under the current network resource conditions determines the time it takes for the estimator to make prior estimates based on system state transitions before the data arrives, which characterizes the transmission lag cost. Based on this, this invention constructs a task value index, defined as the error propagation matrix. The reciprocal of the spectral radius ,in , Represents the state transition matrix. Indicates transmission delay The transmission lag cost caused by system state transitions; Represents the identity matrix. Indicates Kalman gain, Represents the measurement matrix. This represents the utility of measurement information brought about by measurement updates. A higher task value indicates a better overall effect of the scheduling scheme in terms of both measurement information utility and transmission lag cost, thus achieving a unified evaluation of different sensor scheduling schemes. The task value evaluation mechanism constructed in this invention can simultaneously reflect the impact of measurement information utility and transmission lag cost on remote state estimation performance. This mechanism avoids scheduling bias caused by selecting solely based on measurement information utility or solely based on transmission lag cost, making the sensor scheduling results more consistent with the estimation task requirements. In resource-constrained industrial wireless monitoring scenarios, this invention can prioritize sensor combinations that combine high measurement information utility with low transmission lag cost, thereby reducing the impact of transmission lag cost on estimation error propagation, improving the convergence speed and stability of remote state estimation, significantly improving the estimation accuracy and real-time performance under long-term system operation, and enhancing the overall performance of the industrial joint monitoring system.

[0018] In industrial wireless monitoring systems, sensor scheduling and network resource allocation are typically modeled and optimized separately. The sensor selection side focuses primarily on measurement updates, while the network resource allocation side focuses primarily on transmission latency, lacking effective coordination between the two. However, the selected sensor set affects the utility of measurement information, and the network resource allocation scheme for that set affects the transmission lag cost of that set's measurement information, indicating a significant coupling relationship. Existing methods often ignore this coupling, leading to system-level performance deviations. This invention further constructs a sensing-transmission collaborative design framework based on task value indicators, incorporating sensor scheduling and network resource allocation into a unified optimization process. When evaluating candidate sensor sets, this framework considers both measurement information utility and transmission lag cost, jointly determining the selected sensors, the number of subcarriers, and the transmit power, thereby achieving coordinated optimization of measurement updates on the sensing side and transmission latency on the transmission side. By introducing task value as a unified scheduling criterion, a joint decision-making mechanism across the sensing and transmission layers is achieved. This invention's collaborative design framework integrates sensor scheduling and network resource allocation into the same process. For any candidate sensor set, its measurement matrix and measurement noise determine the measurement information utility of the set for remote estimation updates; simultaneously, the subcarrier and transmit power configuration of the set under given communication constraints determines its achievable transmission delay. For each candidate sensor set, the system first calculates the minimum transmission delay under communication constraints, then combines the set's measurement information utility to calculate the task value, and compares different candidate schemes based on the task value. Through this mechanism, sensor scheduling is simultaneously affected by both measurement information utility and transmission lag cost, and network resource allocation schemes also directly participate in the task value evaluation of candidate sets, thereby achieving collaborative optimization of sensor scheduling and network resource allocation. This invention, by constructing a sensing-transmission collaborative design framework, can simultaneously consider measurement information utility and transmission lag cost during the evaluation of candidate sensor sets, enabling sensor scheduling and network resource allocation to simultaneously contribute to remote state estimation performance optimization. This framework avoids suboptimal scheduling results caused by selecting sensors solely based on measurement information utility or allocating resources solely based on transmission lag cost, ensuring that the selected sensor set satisfies communication reliability and resource constraints while also considering measurement information utility and transmission lag cost. In resource-constrained industrial wireless joint monitoring scenarios, this invention can reduce the propagation of estimation errors caused by latency and improve the accuracy of long-term remote state estimation.

[0019] When simultaneously considering sensor scheduling, subcarrier allocation, transmit power limitations, communication reliability constraints, and system observability constraints, the sensing-transmission joint optimization problem exhibits significant combinatorial characteristics and coupling constraints. Directly enumerating all sensor subsets and their resource allocation schemes results in high computational complexity, making it unsuitable for online scheduling requirements under dynamic sampling periods. Therefore, a low-complexity, highly executable solution method is urgently needed to improve the real-time scheduling capability of industrial wireless monitoring systems. This invention proposes a hierarchical task-value-driven solution algorithm, decomposing the sensing-transmission joint optimization problem into two nested subproblems: an outer-layer sensor selection problem and an inner-layer network resource optimization problem. The outer layer is responsible for constructing candidate sensor sets, while the inner layer is responsible for determining feasible network resource configurations given the candidate sets. This transforms the original combinatorial optimization problem involving both sensor scheduling and network resource allocation into a hierarchical optimization structure that can be executed online. By proposing a hierarchical task-value-driven solution algorithm, this invention can transform the complex sensing-transmission joint optimization problem into a hierarchical solution process combining outer-layer sensor selection and inner-layer network resource optimization, avoiding exhaustive searches of all sensor subsets, subcarrier allocation, and transmit power allocation schemes. This algorithm, while ensuring system observability, communication reliability, and resource constraints, can quickly screen candidate sensor sets with high task value and their corresponding resource allocation schemes. Therefore, this invention reduces computational complexity under dynamic sampling periods, improves online scheduling efficiency in resource-constrained industrial wireless monitoring systems, and enhances long-term remote state estimation performance with lower solution overhead.

[0020] Compared with the prior art, the present invention has the following obvious substantive features and significant advantages: This invention addresses resource-constrained industrial wireless joint monitoring scenarios by proposing a task-value-based sensing-transmission collaborative optimization method. By jointly optimizing sensor scheduling and network resource allocation, it improves the accuracy and timeliness of remote state estimation and has promising engineering application prospects.

[0021] 1. Technical advantages of this invention

[0022] High network resource utilization efficiency: This invention incorporates sensor scheduling and network resource allocation into a unified optimization process, while considering the utility of measurement information and the cost of transmission lag. It prioritizes uploading data from sensors that are more valuable to the current estimation task, thereby improving the efficiency of network resource utilization.

[0023] 2. Performance indicators of the present invention

[0024] Performance Guarantee: This invention comprehensively considers the utility of measurement information and the cost of transmission lag, and combines system observability, communication reliability and resource constraints for collaborative optimization, so that the remote estimator can obtain effective measurement data in a timely manner, and ensures the performance of remote state estimation under the condition of limited communication resources.

[0025] 3. Production Implementation of the Invention

[0026] Industry Applicability: This invention is applicable to multi-sensor joint monitoring scenarios such as aerospace manufacturing, equipment inspection, industrial process monitoring, and intelligent manufacturing, and is especially suitable for processes with high requirements for real-time performance and accuracy.

[0027] In summary, this invention can improve the network resource utilization efficiency and remote state estimation performance of resource-constrained industrial wireless monitoring systems, and has good engineering implementation value and potential for widespread application.

[0028] The following will further explain the concept, specific structure, and technical effects of the present invention in conjunction with the accompanying drawings, so as to fully understand the purpose, features, and effects of the present invention. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of an industrial joint monitoring system architecture according to a preferred embodiment of the present invention; Figure 2 This is a schematic diagram of a preferred embodiment of the perception-transmission collaborative design framework based on task value of the present invention; Figure 3 This is a flowchart of a preferred embodiment of the hierarchical task value-driven sensor scheduling and resource allocation algorithm of the present invention; Figure 4 This is a flowchart of network resource optimization under a fixed set of candidate sensors according to a preferred embodiment of the present invention. Detailed Implementation

[0030] The following description, with reference to the accompanying drawings, illustrates several preferred embodiments of the present invention to make its technical content clearer and easier to understand. The present invention can be embodied in many different forms, and the scope of protection of the present invention is not limited to the embodiments mentioned herein.

[0031] In the accompanying drawings, components with the same structure are indicated by the same numerical designation, and components with similar structures or functions are indicated by similar numerical designations. The dimensions and thicknesses of each component shown in the drawings are arbitrary, and the present invention does not limit the dimensions and thicknesses of each component. To make the illustrations clearer, the thickness of some components has been appropriately exaggerated in the drawings.

[0032] Figure 1This paper presents an industrial joint monitoring system architecture for remote state estimation in resource-constrained industrial wireless networks. Taking the measurement of the moving surface angle during aircraft final assembly as an example, multiple heterogeneous sensors deployed near the moving surface collect angle, displacement, and attitude-related measurement data, which are then uploaded to a remote estimation layer via a wireless transmission layer. The remote estimation layer estimates the moving surface angle state based on the received measurement data and generates subsequent sensor scheduling and resource allocation instructions. The system architecture mainly consists of an industrial field monitoring layer, a wireless transmission layer, and a remote estimation layer.

[0033] Industrial Site Monitoring Layer: This layer comprises the monitored object and multiple sensors deployed around it. Taking the measurement of the angle of an aircraft assembly moving surface as an example, the monitoring layer includes the aircraft's moving surface and multiple sets of sensors deployed near it. The moving surface includes movable components such as ailerons, elevators, rudders, flaps, and spoilers. Different sensor sets are used to monitor different moving surfaces. Each sensor set consists of multiple heterogeneous sensors, including angle sensors, acceleration sensors, vision measurement devices, and laser measurement devices, used to collect data related to the angle state of the moving surface from different measurement dimensions. This layer is primarily responsible for the perception and collection of industrial site data.

[0034] Wireless Transmission Layer: This layer connects the field sensors to the remote estimation layer, responsible for transmitting the selected sensor measurement data to the remote estimator. This layer uses a shared wireless network for data transmission, dividing the wireless bandwidth into multiple subcarriers and allocating communication resources to different sensors. This layer primarily reflects the transmission process of field data from the sensor side to the remote estimation side.

[0035] Remote Estimation Layer: The remote estimation layer includes a remote estimator and a scheduling decision module. The remote estimator receives sensor measurement data from the wireless transmission layer and estimates the operating status of the monitored object based on the system model. The scheduling decision module determines the sensors that need to upload data in the next sampling period and the corresponding resource allocation scheme based on the estimation requirements and network resource status, and feeds back the scheduling instructions to the field sensors. This layer is mainly responsible for state estimation and scheduling instruction generation.

[0036] This system architecture is based on industrial field equipment condition monitoring, connecting field sensing, wireless transmission, and remote estimation into a complete monitoring closed loop. The industrial field monitoring layer provides multi-source measurement data of the monitored object; the wireless transmission layer enables the uploading of measurement data from the field end to the remote estimation end; and the remote estimation layer completes condition estimation based on the received data and feeds back subsequent data upload instructions to the field end. This architecture intuitively illustrates how sensor measurement data is transmitted from the industrial field to the remote estimator, and how the remote estimation results further influence subsequent monitoring processes, providing a clear structural foundation for describing the system's operation.

[0037] Figure 2 A task-value-based perception-transmission collaborative design framework is demonstrated. This framework is used in… Figure 1 In the industrial joint monitoring system shown, the measurement information utility and transmission lag cost of a candidate sensor set are uniformly evaluated, and the evaluation results are used as the basis for subsequent sensor scheduling. This framework performs a collaborative evaluation of the candidate sensor set, where the sensing side assesses the measurement information utility of the sensor set for remote state estimation, and the transmission side analyzes the transmission lag cost of the set's measurement information under current network resource conditions. Since both measurement information utility and transmission lag cost jointly affect the performance of remote state estimation, this invention proposes a task value index, incorporating both into the same evaluation index to characterize the comprehensive effect of the candidate sensor set on both measurement information utility and transmission lag cost.

[0038] Within this framework, the candidate sensor set Conduct collaborative evaluation. The perception side, based on the set... The measurement matrices and measurement noise of each sensor are used to construct an aggregated measurement matrix. and measurement noise matrix And calculate the corresponding Kalman gain. .in, The measurement information utility used to characterize the candidate sensor set for remote state estimation. The transmission side uses the candidate sensor set... The data packet size, wireless channel status, and available network resource configuration of each sensor are used to determine the transmission latency required for the entire dataset to complete data upload. This refers to the time required for the remote estimator to perform prior estimations before the measurement data arrives. Used to characterize the transmission lag cost of measurement information.

[0039] Based on the evaluations at the sensing and transmission sides, a candidate sensor set is further constructed. The corresponding error propagation matrix:

[0040] This error propagation matrix incorporates the utility of measurement information and the cost of transmission lag into the same expression, and is used to characterize the impact of the candidate sensor set on the evolution of remote state estimation error.

[0041] Furthermore, the error propagation matrix is ​​calculated. spectral radius And its reciprocal is defined as the candidate sensor set. Task value:

[0042] The greater the task value, the better the overall performance of the candidate sensor set in balancing the utility of measurement information and the cost of transmission lag. (The candidate sensor set...) The task value is used as the evaluation result of this set and is used for comparison of candidate sets in subsequent perception-transmission collaborative design.

[0043] Figure 3 It is layered The flowchart of the sensor scheduling and resource allocation algorithm driven by the system includes the following steps: Step 1: Obtain the sensor set The prior estimation error covariance at the current moment Maximum number of selectable sensors And network resource status, initialize the selected sensor set. It is an empty set.

[0044] Step 2: From the set of sensors that are not currently selected Select sensors in sequence Construct a candidate sensor set .in, Represents the entire set of sensors; Step 3: For the candidate sensor set , call Figure 4 The network resource optimization method shown is used under a fixed set of candidate sensors to determine the minimum achievable transmission delay corresponding to that set of candidate sensors. .

[0045] Step 4: Based on the candidate sensor set Based on the measurement information utility and minimum achievable transmission delay, construct the error propagation matrix corresponding to the candidate set. and task value .

[0046] Step 5, Judgment Check if all sensors in the list have completed candidate evaluation. If there are any unevaluated sensors, return to step 2.

[0047] Step 6: Compare the task value of each candidate sensor set and select the candidate sensor set with the highest task value. As a new set of selected sensors .

[0048] Step 7: Determine the selected sensor set Has the number of sensors reached the maximum number of selectable sensors? If not achieved, return to step 2.

[0049] Step 8: Output the final sensor scheduling result And the corresponding network resource allocation scheme.

[0050] Figure 4 This is a flowchart of network resource optimization under a fixed set of candidate sensors, including the following steps: Step 1: Input a fixed set of candidate sensors And the data packet size of each sensor s within the set. Channel gain Distance attenuation coefficient Subcarrier bandwidth Total number of available subcarriers Maximum transmission power Signal-to-noise ratio threshold and time slot length .

[0051] Step 2: Set the lower bound for transmission delay search and the upper bound of the search .

[0052] Step 3: Take the median value of the current search interval. As a candidate transmission delay.

[0053] Step 4: Resource feasibility check. For the candidate sensor set... Each sensor in Search for the minimum number of subcarriers that satisfy the signal-to-noise ratio constraint and the minimum transmission rate requirement. If all sensors have corresponding... And the total number of subcarriers required does not exceed Then determine the current candidate transmission delay. If feasible; otherwise, determine the current candidate transmission delay. Not feasible.

[0054] Step 5: Update the binary search interval based on the resource feasibility check results. If If feasible, then Update to the upper bound of the search and record the current resource allocation results; if If it is not feasible, then Updated to the lower bound of the search.

[0055] Step 6: Determine if the search interval meets the termination condition. If Greater than the set precision If so, return to step 3 to continue the search; if Not greater than the set precision If the search fails, proceed to step 7.

[0056] Step 7: Output a fixed set of candidate sensors The corresponding minimum achievable transmission latency and the corresponding network resource allocation scheme.

[0057] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. A sensing-transmission collaborative optimization method based on joint estimation of task value, characterized in that, This includes a sensor scheduling evaluation mechanism based on task value, which is used to uniformly quantify the utility of measurement information brought about by measurement updates and the cost of transmission lag caused by transmission delay, so as to characterize the combined impact of the two on the performance of remote state estimation.

2. The perception-transmission collaborative optimization method based on joint estimation of task value as described in claim 1, wherein the evaluation mechanism is based on the evolution process of remote state estimation error, and decomposes the role of sensor scheduling scheme into two parts: measurement information utility and transmission lag cost; for any candidate sensor set, the measurement matrix and measurement noise determine the update effect on the estimation error, which is used to characterize the measurement information utility; the transmission delay generated under the current network resource conditions determines the time for the estimator to make prior estimation based on system state transition before the data arrives, which is used to characterize the transmission lag cost.

3. In the perception-transmission collaborative optimization method based on joint estimation of task value as described in claim 2, the task value index is defined as the error propagation matrix. The reciprocal of the spectral radius ,in , Represents the state transition matrix. Indicates transmission delay The transmission lag cost caused by system state transitions; Represents the identity matrix. Indicates Kalman gain, Represents the measurement matrix. This indicates the utility of measurement information resulting from measurement updates.

4. The sensing-transmission collaborative optimization method based on joint estimation of task value as described in claim 1 constructs a sensing-transmission collaborative design framework based on task value indicators, and incorporates sensor scheduling and network resource allocation into a unified optimization process; when evaluating the candidate sensor set, it simultaneously considers the utility of measurement information and the cost of transmission lag, and jointly determines the selected sensors, the number of subcarriers and the transmission power.

5. The sensing-transmission cooperative optimization method based on joint estimation of task value as described in claim 4, wherein for any candidate sensor set, the measurement matrix and measurement noise determine the measurement information utility of the set for remote state estimation; and the subcarrier and transmit power configuration of the set under given communication constraints determine the achievable transmission delay.

6. The sensing-transmission collaborative optimization method based on joint estimation of task value as described in claim 4, wherein the collaborative design framework, for each candidate sensor set, first calculates the minimum transmission delay under communication constraints, then calculates the task value by combining the measurement update effect of the set, and compares different candidate schemes based on the task value.

7. The perception-transmission collaborative optimization method based on joint estimation of task value as described in claim 1 includes a hierarchical task value-driven solution algorithm, which decomposes the perception-transmission joint optimization problem into two nested sub-problems: an outer sensor selection layer and an inner network resource optimization layer. The outer layer is responsible for constructing a set of candidate sensors, and the inner layer is responsible for determining feasible network resource configurations given the set of candidate sensors. This transforms the original combined optimization problem involving both sensor scheduling and network resource allocation into a hierarchical optimization structure that can be executed online.

8. The perception-transmission collaborative optimization method based on joint estimation of task value as described in claim 7, wherein the hierarchical task value-driven solution algorithm includes the following steps: Step 1: Initialize the selected sensor set It is an empty set; Step 2: From the set of sensors that are not currently selected Select sensors in sequence Construct a candidate sensor set .in, Represents the entire set of sensors; Step 3: For the candidate sensor set By invoking network resource optimization methods, the minimum achievable transmission latency is obtained. ; Step 4: Based on the candidate sensor set Measurement information utility and minimum achievable transmission delay Calculate the value of the task ; Step 5: After evaluating all candidate sensor sets, select the candidate sensor set with the highest task value. As a new set of selected sensors ; Step 6: Determine the selected sensor set Has the number of sensors reached the maximum number of selectable sensors? If the target is not reached, return to step 2; if the target is reached, output the final sensor scheduling result. And the corresponding network resource allocation scheme.

9. The sensing-transmission cooperative optimization method based on joint estimation of task value as described in claim 8, wherein step 3 includes the following steps: Step 3.1: For the candidate sensor set Acquire sensor information and network resource parameters, and set a lower bound for transmission delay search. Search upper bound and setting precision ; Step 3.2: Take the median value of the current search interval. As a candidate transmission delay, according to Calculate the candidate sensor set The minimum transmission rate required by each sensor is determined, and the transmit power of each sensor is set to the maximum value. Step 3.3, in the candidate transmission delay Next, a resource feasibility check is performed to search for the minimum number of subcarriers required for each sensor to meet the signal-to-noise ratio constraints and minimum transmission rate requirements. If all sensors have corresponding And the total number of subcarriers required does not exceed the total number of available subcarriers. Then determine Feasible; otherwise, determine Not feasible; Step 3.4: Update the binary search interval based on the feasibility assessment results; if If feasible, then Update to the upper bound of the search and record the current resource allocation results; if If it is not feasible, then Updated to the lower bound of the search; Step 3.5: Determine if the search interval meets the termination condition; if Greater than the set precision If so, return to step 3.2; if Not greater than the set precision If the search stops, the candidate sensor set is obtained. The corresponding minimum achievable transmission delay And network resource allocation scheme.

10. The sensing-transmission cooperative optimization method based on joint estimation of task value as described in claim 8, wherein step 4 includes the following steps: Step 4.1: Construct a candidate sensor set The corresponding error propagation matrix ;in Indicates transmission delay The transmission lag cost caused by system state transitions Represents the set of candidate sensors The corresponding Kalman gain, Represents the set of candidate sensors The corresponding measurement matrix, Indicates the selection of candidate sensor set The utility of measurement information resulting from measurement updates; Step 4.2: Calculate the error propagation matrix spectral radius ; Step 4.3: Calculate the sensor set Task value .

Images