An integrated underwater information network comprehensive performance evaluation and prediction method

Abstract

The application discloses a kind of underwater information network comprehensive performance evaluation and forecast method of integrated communication, navigation and sensing, belongs to underwater network and information system technical field.The application is directed to the problem that node, channel, task in integrated network of communication, navigation and sensing are highly coupled and difficult to be quantitatively evaluated, constructs four-layer progressive mathematical model system of single node capability characterization layer, multi-node coupling characterization layer, single task performance characterization layer and comprehensive performance characterization layer.Extract the key coupling coefficient of network effect, communication conflict probability, positioning error statistical parameter and sensing coverage overlap coefficient through Monte Carlo simulation, realize the quantitative separation of network effect, then calculate network communication throughput, navigation positioning error and sensing coverage range, and fuse the normalized single task performance into network comprehensive performance index through task weight.The application also provides network comprehensive performance standardization forecast process, and the relative deviation between forecast comprehensive performance and measured value is 10.8% after sea trial, which can provide decision basis for design optimization and task planning of integrated underwater information network.

Description

Technical Field

[0001] This invention relates to the fields of underwater information systems, network communication and signal processing technology, and specifically to a method for quantitatively evaluating the comprehensive performance of an integrated underwater information network that combines communication, navigation and sensing. It is mainly a method for evaluating and predicting the comprehensive performance of an integrated underwater information network that combines communication, navigation and sensing. Background Technology

[0002] With the increasing demand for integrated underwater information support in fields such as marine observation, resource development, and marine security, integrated underwater information networks combining communication, navigation, and sensing have emerged. These networks deeply integrate underwater communication, collaborative navigation, and distributed sensing functions into a single infrastructure, aiming to improve resource utilization efficiency and mission coordination capabilities.

[0003] However, scientifically evaluating the overall performance of such integrated networks faces significant challenges. Network performance is the result of highly nonlinear coupling between node characteristics, time-varying underwater acoustic channel characteristics, multi-task requirements, and multi-user access characteristics. Traditional evaluation methods typically analyze single functions in isolation (e.g., evaluating only communication capacity or only positioning accuracy) or simply linearly superimpose the performance of individual nodes, failing to accurately quantify the gains (e.g., enhanced coverage, improved accuracy) or performance losses (e.g., multiple access interference, increased collisions) brought about by network collaboration. This lack of evaluation methods leads to a lack of theoretical guidance for network design, difficulty in optimizing resource allocation, and unpredictable task performance, severely restricting the engineering application and development of integrated underwater information networks.

[0004] Therefore, there is an urgent need for a method that can systematically decouple the aforementioned complex coupling relationships and quantitatively evaluate and accurately predict the overall performance of a network performing multiple tasks. Based on this, this application is made. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a method for comprehensive performance evaluation and prediction of integrated underwater information networks that combines communication, navigation, and sensing. This method aims to decouple and quantify the multiple coupling effects of nodes, channels, tasks, and network structure layer by layer by constructing a systematic and computable "network equation" model, thereby achieving accurate and forward-looking evaluation of the overall network performance and providing a reliable tool for network design, resource optimization, and task planning.

[0006] The objective of this invention is achieved through the following technical solution: A method for comprehensive performance evaluation and forecasting of an integrated underwater information network combining communication, guidance, and sensing, specifically comprising the following steps:

[0007] S1. Construct a four-layer progressive network performance evaluation model, which includes, in sequence: a single-node capability representation layer, a multi-node coupling representation layer, a single-task performance representation layer, and a comprehensive performance representation layer;

[0008] S2. Based on preset network parameters, environmental parameters, and task parameters, the coupling coefficients in the multi-node coupling representation layer are obtained through Monte Carlo simulation. The coupling coefficients include the collision probability for evaluating network communication performance, the positioning error statistical parameters for evaluating network navigation performance, and the coverage overlap coefficient for evaluating network perception performance.

[0009] S3. Based on the coupling coefficient, the networked communication throughput, networked navigation and positioning error, and networked sensing coverage are calculated respectively through the single-task performance characterization layer;

[0010] S4. Normalize the performance indicators of each individual task obtained in step S3, and calculate the network's overall performance index according to the preset task weight coefficients and the overall performance evaluation equation of the comprehensive performance characterization layer. .

[0011] Furthermore, in step S1:

[0012] The single-node capability representation layer is constructed based on sonar equations and is used to quantify the basic capabilities of a single node to perform communication, navigation or perception tasks. These basic capabilities are related to the environment and the hardware capabilities of the single node.

[0013] The multi-node coupling characterization layer quantifies the networking effect generated after nodes are interconnected into a network through the coupling coefficient. The networking effect includes the gain and / or performance loss brought about by multi-node network collaboration.

[0014] The single-task performance characterization layer modifies the basic capabilities based on the coupling coefficient to obtain networked single-task performance indicators.

[0015] The comprehensive performance characterization layer is evaluated through the comprehensive performance evaluation equation. The normalized performance metrics of each individual task are then fused together, among which... These are the weighting coefficients for communication, navigation, and perception tasks, respectively. This is the normalization function.

[0016] Furthermore, the networked communication throughput ,in For the first The theoretical throughput of a communication link under collision-free conditions The probability of collisions used to evaluate network communication performance;

[0017] The networked navigation and positioning error Used to evaluate network navigation performance, the mean value of positioning error statistical parameters is selected according to the evaluation requirements. or standard deviation ;

[0018] The networked sensing coverage area ,in This refers to the sensing coverage area of ​​a single node. This is the coverage overlap coefficient used to evaluate network-aware performance.

[0019] Furthermore, the step S2, which involves obtaining the coupling coefficient through Monte Carlo simulation, includes: simulating the network's operation in a dynamic environment and statistically analyzing the proportion of conflicting data packets to obtain the conflict probability. Statistical analysis is performed on the positioning solution results from multiple simulations to obtain the positioning error statistical parameters. , Randomly generate target routes, perform multiple simulations to obtain the network's joint sensing coverage area, and compare it with the sum of the independent sensing coverage areas of all nodes to obtain the coverage overlap coefficient. .

[0020] Furthermore, the normalization process described in step S4 employs the Min-Max normalization method:

[0021] Networked communication throughput score ,and ,in This represents the minimum tolerable value for network throughput. This represents the optimal value for network throughput.

[0022] Networked navigation accuracy score ,and ,in This represents the maximum tolerance value for navigation errors. This represents the optimal value for navigation error.

[0023] Networked sensing coverage score ,and ,in This represents the minimum coverage area for network perception. This is the sum of the independent coverage areas of each node in the network;

[0024] Normalized comprehensive efficiency index .

[0025] Furthermore, the network parameters mentioned in step S2 include node characteristics, network topology, and network protocol parameters; the environmental parameters include sound velocity profile, environmental noise, and seabed sediment parameters; and the task parameters include communication error rate threshold, navigation accuracy requirements, and perception detection probability threshold.

[0026] Furthermore, in the multi-node coupled representation layer, the communication collision probability The proportion of data packets with time slot or carrier collisions is obtained through Monte Carlo simulation; the positioning error statistics are obtained through multiple simulations and statistical analyses of the long baseline / short baseline underwater acoustic positioning results; the coverage overlap coefficient is... The ratio of the target route to the sum of the coverage areas of the network's joint detection and the independent coverage areas of each node is obtained by randomly generating the target route and calculating the ratio of the coverage areas of the network's joint detection to the sum of the coverage areas of each node.

[0027] Furthermore, this invention predicts networked communication throughput, networked navigation and positioning error, networked sensing coverage, and overall performance index under different network configurations or environmental conditions by changing the input network parameters, environmental parameters, and task parameters.

[0028] Furthermore, the predicted value of the comprehensive performance index is compared with the actual sea trial measurement value to verify the accuracy of the evaluation model; the comparison uses a relative deviation index, and when the relative deviation meets the preset threshold for engineering applications, the evaluation model is confirmed to be effective.

[0029] Compared with the prior art, the present invention has the following beneficial technical effects:

[0030] (i) It is the first to propose a “four-layer network equation” model architecture, which realizes effective decoupling and comprehensive evaluation of the integrated performance of communication, guidance and sensing.

[0031] This invention addresses the fundamental differences in physical mechanisms and performance metrics among communication, navigation, and sensing tasks. Employing a systems engineering approach of "element decoupling, task decoupling, functional decoupling, and structural decoupling," it constructs a four-layer progressive mathematical model system, progressing from single-node capabilities to multi-node coupling, then to single-task performance, and finally to overall performance. A key innovation of this system lies in introducing "coupling coefficients" (communication conflict probability, positioning error statistical parameters, and sensing overlap coefficients) into the multi-node coupling representation layer. This allows the complex network effects resulting from interconnected nodes—including gains from collaboration (such as enhanced multi-station joint coverage and improved multi-primary unit joint positioning accuracy) and performance losses from competition (such as multiple access conflicts and increased mutual interference)—to be separated from overall performance and quantified as explicit parameters. This makes the coupling relationships, previously implicit within the system and difficult to directly observe and calculate, transparent, measurable, and computable. By employing a two-step strategy of first simulating and extracting coupling coefficients, and then substituting them into formulas to calculate single-task performance and overall effectiveness, this invention achieves a systematic decoupling analysis and fusion evaluation of the integrated performance of communication, conduction, and sensing, transforming it from "coupling chaos" to "structural clarity." This solves the problem that traditional methods cannot uniformly quantify the performance of the three tasks under the same framework due to their high coupling.

[0032] (ii) A standardized and programmable evaluation process has been established, which has the ability to make forward-looking predictions.

[0033] This invention transforms performance evaluation into three repeatable and programmable standardized steps: parameterized scenario construction, Monte Carlo simulation extraction of coupling coefficients, and analytical calculation of network equations. The core advantage of this process lies in the fact that once the mapping relationship between network parameters, environmental parameters, task parameters, and coupling coefficients is established through simulation, the comprehensive performance under different network designs (number of nodes, topology, protocol parameters), deployment schemes, or environmental conditions can be rapidly predicted by changing input parameters, without the need for actual network deployment or sea trials. Compared to existing post-analysis-based evaluation methods, this invention provides pre-emptive forecasting capabilities, offering quantitative decision-making basis for scheme selection in the network optimization design phase, parameter configuration in the dynamic resource scheduling phase, and performance simulation and deduction before task execution. This fills the technological gap in the field of integrated underwater information networks, which lacks forward-looking performance evaluation tools.

[0034] (iii) Through the normalization and weighted fusion mechanism, the unified quantification and flexible allocation of multi-task efficiency have been achieved.

[0035] Communication throughput (unit: bit / s), navigation and positioning error (unit: m), and sensing coverage (unit: km²) have different dimensions and physical meanings, and directly adding them together lacks physical meaning. This invention employs a Min-Max standardization method to map these three heterogeneous indicators to a unified scoring interval, and then performs a weighted sum using preset task weight coefficients to form a dimensionless comprehensive performance index. Since the weight coefficients can be flexibly adjusted according to the priority of communication, navigation, and sensing tasks in different application scenarios (e.g., increasing the sensing weight in collaborative detection scenarios and increasing the communication weight in data feedback scenarios), this evaluation framework can adapt to various task profiles and has good versatility and flexibility. The truncation process during normalization further ensures the robustness of the evaluation results, preventing extreme outliers from unduly affecting the comprehensive score. Actual sea trials have verified that the relative deviation between the predicted and measured comprehensive performance index values ​​obtained using this method is only 10.8%, meeting the prediction accuracy requirements for engineering applications. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of the four-layer model architecture of the "network equation" of this invention;

[0037] Figure 2 This is a schematic diagram of the process for extracting coupling coefficients based on Monte Carlo simulation in this invention. Detailed Implementation

[0038] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the present invention.

[0039] like Figure 1 and Figure 2 As shown, this invention provides a method for comprehensive performance evaluation and prediction of integrated underwater information networks, whose core lies in constructing and applying a four-layer progressive "network equation" model. This model adopts the systems engineering approach of "element decoupling, hierarchical representation, and step-by-step fusion," transforming the complex problem of comprehensive network performance evaluation into a clearly structured and computable problem. The four-layer model includes:

[0040] 1. Single-node capability representation layer: Based on the sonar equations, establish the relationship between the environment, target and node parameters, and quantify the basic physical capabilities of a single node to perform communication, navigation or perception tasks under ideal conditions.

[0041] 2. Multi-Node Coupling Representation Layer: As a key bridge connecting single-point capabilities and network capabilities, this layer quantifies the networking effect generated after nodes are networked by introducing a "coupling coefficient." The coupling coefficient includes at least: the probability of communication collisions representing multi-user interference. Navigation and positioning error statistical parameters (such as mean) characterizing network geometric layout and error propagation and standard deviation ), and the sensing overlap coefficient representing coverage redundancy and blind spots. These coefficients were obtained through Monte Carlo simulation combined with specific network application scenarios.

[0042] 3. Single-task performance characterization layer: Under the correction of the coupling coefficient, a single-task performance equation for networked scenarios is established. Specifically, this includes networked communication throughput. Networked navigation and positioning errors The coverage area of ​​the networked sensing system was obtained through statistical analysis of the positioning results from multiple simulations. .

[0043] 4. Comprehensive Performance Characterization Layer: Weighting coefficients determined based on task priority or decision-making requirements. The normalized performance indicators of each individual task are then combined to form the final comprehensive performance evaluation equation: in, The value directly reflects the overall performance level of the network in performing multiple tasks under a given configuration.

[0044] Based on the aforementioned "network equation", the performance evaluation and prediction process of this invention includes:

[0045] Step 1: Parametric Scenario Construction. Input three types of parameters: network parameters (node ​​characteristics, topology, network protocols, etc.), environmental parameters (sound velocity profile, noise, seabed sediment, etc.), and task parameters (business requirements, performance thresholds, etc.).

[0046] Step 2: Coupling Coefficient Simulation Extraction. Based on the parameters from Step 1, run a Monte Carlo simulation to dynamically simulate network operation, and statistically calculate the probability of communication collisions. Positioning error statistical characteristics, sensing overlap coefficient .

[0047] Step 3: Network Equation Calculation and Prediction. Substitute the coupling coefficients obtained in Step 2 into the "network equation," and calculate it sequentially through each representation layer. The final output includes predicted values ​​for networked communication throughput, navigation and positioning error, sensing coverage, and a comprehensive performance index. .

[0048] Example:

[0049] 1. Parametric Scene Construction

[0050] According to step one of the present invention, the input parameters required for simulation evaluation are set:

[0051] 1.1 Network Parameters: The network is configured with 4 nodes (3 user nodes and 1 central node), using the TDMA protocol for multiple access, with a time slot length of 2 minutes, and a specific modulation and coding scheme at the physical layer. Network topology, transmitter level, receiver sensitivity, and other parameters are set according to the actual experimental implementation.

[0052] 1.2 Environmental Parameters: The propagation loss was calculated using the measured sound velocity profile and seabed parameters from the test sea area, and a sound field model (such as Bellhop). The simulated background noise level was set to the measured background noise level from the test sea area.

[0053] 1.3 Task Parameters: The communication task requires the bit error rate to not exceed a certain threshold (e.g., ...). The perception task targets a sound source at a set level and requires the detection probability to meet a certain threshold.

[0054] 2. Monte Carlo simulation and coupling coefficient extraction

[0055] Based on the above parameters, construct a high-fidelity network simulation environment and execute step two.

[0056] For communication tasks, the simulation models the transmission of data packets within TDMA time slots. Due to the use of a collision-free TDMA protocol, the theoretical collision probability is... The simulation also considered packet loss caused by physical layer bit errors, and statistically obtained the equivalent effective data reception probability.

[0057] For the navigation task, the target is set to move along a typical trajectory. In each Monte Carlo simulation run, random perturbations consistent with actual conditions are injected into the positions of each navigation node (simulating ocean current drift and deployment errors), sound speed (simulating measurement errors), and signal arrival time measurements. A positioning algorithm based on the long baseline principle is then run (positioning error...). By including the ideal observation matrix and the observation matrix containing perturbations equation (Calculation) Statistical analysis was performed on a large number of positioning results to obtain the statistical distribution parameters of the positioning error. and .

[0058] For the perception task, the perception coverage of each node is first calculated based on the sonar equations and the sound field model. Then, a target route is randomly generated, and the joint detection coverage area is obtained through multiple simulations. The ratio of this coverage area to the sum of the coverage areas of a single node is the overlap coefficient. In this example, the simulation calculation yields... .

[0059] 3. Network equation calculation and performance prediction

[0060] Substitute the coupling coefficients obtained in step two into the "network equation" for calculation (step three):

[0061] Networked communication throughput forecast: .

[0062] Networked navigation and positioning error prediction: Based on the statistical distribution of positioning errors obtained from simulation, the average predicted positioning error is calculated as follows: standard deviation of error .

[0063] Networked sensing coverage forecast: .

[0064] Comprehensive performance evaluation: If the task weights are set using the analytic hierarchy process (AHP), then... After normalizing the above single-task performance values, they are substituted into the equation. This yields the comprehensive performance index. Normalization aims to make physical quantities with different dimensions dimensionless; this can be achieved using the Min-Max standardization method.

[0065] Throughput Score ,like Below It can be truncated to 0; higher than Cut off to 1.

[0066] Navigation accuracy score The smaller the error, the closer the score is to 1. When the error reaches the maximum tolerance, the score is 0.

[0067] Coverage score .

[0068] Therefore, the overall efficiency after normalization is .

[0069] 4. Sea trials and verification

[0070] To verify the accuracy of this assessment method, sea trials were conducted under the same configuration, and the measured results were compared with the aforementioned forecast results:

[0071] The measured communication throughput value is The relative deviation of the forecast is 16.13%.

[0072] The measured average value of navigation and positioning error is The relative deviation of the forecast is 6.32%.

[0073] The measured value of the sensing coverage is The relative deviation of the forecast is 14.76%.

[0074] Based on the test results, on-site experts determined that the network throughput was lower than the required level. The network communication performance score is truncated to 0, and the network throughput exceeds [a certain threshold]. The network communication performance score is truncated to 1; the average navigation error exceeds [a certain threshold]. The network navigation performance score is truncated to 0, and the error is better than 0. The network navigation performance score is truncated to 1; the network coverage area is smaller than the coverage area of ​​a single node. The network perception performance score is truncated to 0, which is greater than the sum of the individual detection coverage of 3 nodes. The network perception performance score is truncated to 1. Substituting the above parameters, the overall network performance is predicted. Actual measured network overall performance The relative deviation was 10.8%, which meets the requirements of engineering applications for forecast accuracy, proving the accuracy and practicality of the evaluation model proposed in this invention.

[0075] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for comprehensive performance evaluation and forecasting of an integrated underwater information network combining communication, guidance, and sensing, characterized in that, Includes the following steps: S1. Construct a four-layer progressive network performance evaluation model, which includes, in sequence: a single-node capability representation layer, a multi-node coupling representation layer, a single-task performance representation layer, and a comprehensive performance representation layer; S2. Based on preset network parameters, environmental parameters, and task parameters, the coupling coefficients in the multi-node coupling representation layer are obtained through Monte Carlo simulation. The coupling coefficients include the collision probability for evaluating network communication performance, the positioning error statistical parameters for evaluating network navigation performance, and the coverage overlap coefficient for evaluating network perception performance. S3. Based on the coupling coefficient, the networked communication throughput, networked navigation and positioning error, and networked sensing coverage are calculated respectively through the single-task performance characterization layer; S4. Normalize the performance indicators of each individual task obtained in step S3, and calculate the network's overall performance index according to the preset task weight coefficients and the overall performance evaluation equation of the comprehensive performance characterization layer. .

2. The method for comprehensive performance evaluation and forecasting of integrated underwater information networks with communication, guidance, and sensing capabilities as described in claim 1, characterized in that, In step S1: The single-node capability representation layer is constructed based on sonar equations and is used to quantify the basic capabilities of a single node to perform communication, navigation or perception tasks. These basic capabilities are related to the environment and the hardware capabilities of the single node. The multi-node coupling characterization layer quantifies the networking effect generated after nodes are interconnected into a network through the coupling coefficient. The networking effect includes the gain and / or performance loss brought about by multi-node network collaboration. The single-task performance characterization layer modifies the basic capabilities based on the coupling coefficient to obtain networked single-task performance indicators. The comprehensive performance characterization layer is evaluated through the comprehensive performance evaluation equation. The normalized performance metrics of each individual task are then fused together, among which... These are the weighting coefficients for communication, navigation, and perception tasks, respectively. This is the normalization function.

3. The method for comprehensive performance evaluation and forecasting of integrated underwater information networks with communication, guidance, and sensing capabilities as described in claim 2, is characterized in that... The networked communication throughput ,in For the first The theoretical throughput of a communication link under collision-free conditions The probability of collisions used to evaluate network communication performance; The networked navigation and positioning error Used to evaluate network navigation performance, the mean value of positioning error statistical parameters is selected according to the evaluation requirements. or standard deviation ; The networked sensing coverage area ,in This refers to the sensing coverage area of ​​a single node. This is the coverage overlap coefficient used to evaluate network-aware performance.

4. The method for comprehensive performance evaluation and forecasting of integrated underwater information networks with communication, guidance, and sensing capabilities as described in claim 3, is characterized in that... The step S2, obtaining the coupling coefficient through Monte Carlo simulation, includes: simulating network operation in a dynamic environment and statistically analyzing the proportion of conflicting data packets to obtain the conflict probability. Statistical analysis is performed on the positioning solution results from multiple simulations to obtain the positioning error statistical parameters. , Randomly generate target routes, perform multiple simulations to obtain the network's joint sensing coverage area, and compare it with the sum of the independent sensing coverage areas of all nodes to obtain the coverage overlap coefficient. .

5. The method for comprehensive performance evaluation and forecasting of integrated underwater information networks with communication, guidance, and sensing capabilities as described in claim 4, is characterized in that... The normalization process described in step S4 uses the Min-Max normalization method: Networked communication throughput score ,and ,in This represents the minimum tolerable value for network throughput. This represents the optimal value for network throughput. Networked navigation accuracy score ,and ,in This represents the maximum tolerance value for navigation errors. This represents the optimal value for navigation error. Networked sensing coverage score ,and ,in This represents the minimum coverage area for network perception. This is the sum of the independent coverage areas of each node in the network; Normalized comprehensive efficiency index .

6. The method for comprehensive performance evaluation and forecasting of integrated underwater information networks with communication, guidance, and sensing capabilities as described in claim 5, is characterized in that... The network parameters mentioned in step S2 include node characteristics, network topology, and network protocol parameters; the environmental parameters include sound velocity profile, environmental noise, and seabed sediment parameters; and the mission parameters include communication error rate threshold, navigation accuracy requirements, and perception detection probability threshold.

7. The method for comprehensive performance evaluation and forecasting of integrated underwater information networks with communication, guidance, and sensing capabilities as described in claim 6, is characterized in that... In the multi-node coupled representation layer, the communication conflict probability The proportion of data packets with time slot or carrier collisions is obtained through Monte Carlo simulation; the positioning error statistics are obtained through multiple simulations and statistical analyses of the long baseline / short baseline underwater acoustic positioning results; the coverage overlap coefficient is... The ratio of the target route to the sum of the coverage areas of the network's joint detection and the independent coverage areas of each node is obtained by randomly generating the target route and calculating the ratio of the coverage areas of the network's joint detection to the sum of the coverage areas of each node.

Images