Single hydrophone range-depth positioning method based on multi-ofdm symbol joint processing

Abstract

The application relates to a single hydrophone distance-depth positioning method based on multi-OFDM symbol joint processing, which is suitable for two-dimensional positioning of an underwater sound source. The method comprises the following steps: a plurality of OFDM symbols containing pilots are emitted by a sound source; a single hydrophone receives signals and constructs a multi-measurement vector (MMV) model based on the pilots; a maximum likelihood method is used to estimate a multi-path propagation time delay; a cost function is constructed in combination with a geometric model to inversely calculate the horizontal distance and depth of the sound source. The method does not require a multi-element structure, has the advantages of high positioning accuracy, strong robustness, simple deployment and the like, and is particularly suitable for resource-limited platforms such as UUVs, buoys and distributed underwater nodes. Pool experiments and various applications prove that the application scheme has superior spatial resolution and background noise suppression performance in an underwater sound multi-path environment, and is suitable for various underwater passive positioning scenes.

Description

Technical Field

[0001] This invention relates to the field of underwater acoustic positioning technology, and in particular to a single hydrophone distance-depth positioning method based on orthogonal frequency division multiplexing (OFDM) signal processing. Background Technology

[0002] In typical application scenarios such as marine surveying, underwater monitoring, and unmanned underwater vehicle (UUV) navigation, acoustic source localization technology serves as a core support for underwater perception and autonomous control, possessing broad engineering significance and research value. Currently, most mainstream underwater passive positioning systems rely on array structures composed of multiple hydrophones, estimating the target's position in two-dimensional or three-dimensional space by calculating the angle of arrival (DOA) or time difference of arrival (TDOA) between multiple channels. However, such systems have high requirements in terms of physical deployment space, time synchronization accuracy between nodes, and platform size, making them difficult to adapt to resource-constrained platforms such as buoys and small autonomous underwater vehicles (AUVs), thus limiting their practicality and scalability.

[0003] In architectures with only a single hydrophone, traditional passive positioning methods typically rely on analyzing characteristics such as energy changes, envelope shape, or spectral delay of the received acoustic signal to estimate the propagation delay of the target signal and thus indirectly infer its spatial location in order to reduce hardware complexity. However, in typical underwater environments, due to strong multipath reflection effects, background noise interference, and channel time-varying characteristics, these methods often face problems such as unstable delay estimation and decreased accuracy. When the signal-to-noise ratio is ≤10dB, the positioning error of traditional methods is >3m, which cannot meet the precision operation requirements such as UUV docking (requiring <0.5m).

[0004] Especially in complex boundary conditions such as shallow seas and pools, underwater acoustic channels commonly exhibit multiple propagation paths, including direct paths, surface reflection paths, and bottom reflection paths. Since the time delay differences between these paths are typically small and their frequency responses are highly coupled, effectively separating path features from signals with limited bandwidth becomes a challenge. Although multipath information contains rich clues about target location, existing methods struggle to effectively extract and utilize multipath propagation features under insufficient resolution conditions.

[0005] Existing research attempts to achieve sound source localization by constructing time-frequency graph features of signals, simulating sound field matching, or using database-based localization cost functions. However, these methods often rely on precise environmental modeling parameters, such as prior knowledge of sound velocity profiles and boundary shapes, resulting in poor system adaptability and versatility. Other deep learning-based path recognition methods, while showing certain advantages in simulation environments, heavily depend on a large number of training samples, and their generalization ability in real-world deployment environments is limited, making widespread application difficult.

[0006] Therefore, without relying on array structure or global prior environmental information, how to fully extract the multipath propagation characteristics in complex underwater acoustic channels based on a single hydrophone architecture and establish a robust and scalable sound source localization method has become one of the key issues that current underwater passive localization technology urgently needs to overcome. Summary of the Invention

[0007] To overcome the problems of low positioning accuracy, poor path resolution, and strong dependence on environmental modeling in existing single-hydrophone passive sound source localization systems in complex multipath underwater acoustic channels, this invention proposes a single-hydrophone range-depth localization method based on joint processing of multiple OFDM symbols. By introducing frequency domain pilots, joint path estimation, and cost function inversion mechanisms, the algorithm effectively improves its positioning accuracy and robustness under resource-constrained conditions. The technical solution of this invention will be described in detail below with reference to specific implementation methods.

[0008] In one embodiment of the present invention, a single hydrophone range-depth localization method based on multi-OFDM symbol joint processing is provided, which includes the following steps without relying on array structure and global time synchronization:

[0009] S10, the underwater target sound source continuously transmits multiple orthogonal frequency division multiplexing (OFDM) symbols containing pilot subcarriers, each symbol embedding pilot information of known structure in the frequency domain;

[0010] S20, the single hydrophone at the receiving end receives the multiple OFDM signals, extracts the frequency domain pilot subcarrier data from each symbol, and forms multiple pilot response vectors;

[0011] S30, a multi-measurement vector (MMV) model is constructed based on multiple pilot response vectors. By establishing a linear mapping relationship between the received vector matrix and the frequency domain response matrix of the candidate path, the maximum likelihood estimation is performed using the minimum residual criterion to extract the multi-path propagation delay vector.

[0012] S40, combining the geometric propagation model between the sound source and the receiver, constructs a path propagation cost function, matches the estimated multipath propagation delay with the theoretical propagation delay, and obtains the horizontal distance-depth estimate of the sound source by minimizing this cost function.

[0013] Furthermore, the pilot subcarriers are uniformly distributed in the frequency domain within each OFDM symbol, possessing frequency response distinguishability, which enhances the ability to distinguish multipath signals. The pilot subcarriers account for 5% to 20% of the total number of subcarriers.

[0014] Preferably, the channel response matrix in the MMV model is composed of frequency domain response templates under multipath propagation delay, with each column corresponding to a path propagation feature, and the total number of paths not exceeding 10.

[0015] Alternatively, the maximum likelihood estimation can be performed by constructing an orthogonal projection matrix and using the minimum residual norm or minimum trace criterion to search for the optimal path combination in the propagation delay parameter space to obtain the optimal propagation delay vector.

[0016] Furthermore, the path propagation cost function is used to describe the error of multipath delay corresponding to the hypothetical and real sound source locations. Ideally, the target location is the coordinate point that minimizes the function.

[0017] Preferably, the propagation model includes a direct path, a surface reflection path, and a bottom reflection path, and the propagation delay is obtained by dividing the propagation distance of each path by the speed of sound in the water.

[0018] Alternatively, the cost function can be minimized using a grid search or local gradient descent algorithm, and path weights can be introduced to improve estimation accuracy.

[0019] Furthermore, the multipath propagation delay feature extraction and two-dimensional sound source localization are two decoupled processing stages, which are executed independently based on frequency domain modeling and cost function optimization, respectively.

[0020] In one embodiment of the present invention, an underwater passive positioning system is provided, comprising:

[0021] The receiving module is used to receive multiple OFDM signals containing pilot signals and extract frequency domain pilot data;

[0022] The modeling module is used to construct a multi-measurement vector model and generate a frequency domain receiver matrix and a response template matrix.

[0023] The estimation module is used to estimate the path propagation delay based on the likelihood maximization criterion.

[0024] The localization module is used to construct a cost function based on the propagation geometry model and output the estimated position of the target sound source in two-dimensional space.

[0025] Alternatively, a single-node underwater acoustic passive positioning device deployed on a small platform is provided, configured to perform any of the methods described above, and output the positioning coordinates of the target sound source in the horizontal distance and depth directions.

[0026] Furthermore, the present invention also provides applications of the above-mentioned technical solutions in unmanned underwater vehicle navigation, buoy underwater acoustic monitoring, or earthquake sound source localization.

[0027] Based on the above technical solutions, this invention proposes a single hydrophone range-depth localization method based on joint processing of multiple OFDM symbols. Addressing the problems of low localization accuracy and weak anti-interference capability under single-channel reception conditions in complex underwater multipath propagation environments, this invention constructs an integrated technical framework that combines communication pilot design, frequency domain response modeling, and path parameter estimation. By jointly processing pilot information from multiple OFDM symbols during a relatively stable acoustic channel period, the robustness and resolution of multipath propagation delay estimation are significantly improved, laying a precise and reliable data foundation for two-dimensional spatial localization calculations.

[0028] Compared to traditional single-symbol processing methods, the multi-symbol redundant information constructed in this invention makes full use of the multi-symbol redundancy information, which can effectively suppress the interference of background noise on the path estimation results. It can still maintain high positioning accuracy in scenarios with small multipath intervals or uneven path power distribution, thereby enhancing the adaptability and stability of the system in complex marine environments such as strong noise, hydrodynamic disturbances and rapid channel changes.

[0029] Furthermore, this invention introduces a maximum likelihood estimation (MLE) mechanism in the path parameter estimation stage. By constructing a residual projection matrix and minimizing the time delay domain error norm, it achieves joint optimization and identification of multi-path propagation characteristics, overcoming the limitation of traditional energy peak methods that are highly dependent on path separation, and effectively improving the discriminability of short-range paths. Subsequently, by combining the underwater acoustic propagation geometry model and sound speed constraints, a spatial mapping relationship between path delay and target position is established. By constructing and optimizing the path propagation cost function, accurate joint localization of the target sound source in the distance and depth directions is achieved.

[0030] Furthermore, the method of the present invention has good versatility and scalability. It can be seamlessly integrated into the existing OFDM communication framework as an auxiliary positioning function module, or deployed as an independent passive positioning subsystem in various application scenarios such as underwater acoustic monitoring, moving target tracking, and unmanned underwater vehicle navigation. It is also compatible with resource-constrained platforms such as buoys, underwater unmanned platforms (UUVs), and distributed sensor nodes.

[0031] In summary, this invention provides a novel, practical, and engineering-feasible technical approach in the field of underwater passive sound source localization technology, with the following significant advantages:

[0032] High precision: Multi-symbol joint processing improves frequency domain resolution and path delay estimation accuracy;

[0033] High robustness: It maintains stable performance even under conditions of strong noise, dense multipath propagation, and dynamic channel changes;

[0034] Low cost: The single hydrophone architecture eliminates the need for array deployment and synchronization calibration, simplifying system implementation;

[0035] Highly adaptable: Suitable for a variety of small, low-power platforms, meeting the deployment needs of complex environments;

[0036] System integration: It has the potential to integrate communication and positioning, thereby improving the system's functional integration and task efficiency.

[0037] The invention effectively fills the gap in existing technology for achieving two-dimensional positioning under the conditions of "single hydrophone + multipath channel + no prior modeling", and provides key technical support and innovative solutions for the next generation of intelligent underwater acoustic sensing platforms. Attached Figure Description

[0038] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the embodiments will be further described below with reference to the accompanying drawings. Obviously, the described drawings are only for illustrating the present invention and are not intended to limit the scope of protection of the present invention.

[0039] Figure 1 This is a flowchart of the single hydrophone distance-depth localization method based on multi-OFDM symbol joint processing according to the present invention.

[0040] Figure 2 This is a schematic diagram of a typical three-path underwater acoustic propagation structure involved in the present invention.

[0041] Figure 3 This is a comparison chart of the two-dimensional positioning performance of the method of the present invention in a water tank experimental environment, showing the error comparison data between the traditional method and the method of the present invention in the horizontal distance and depth directions. Detailed Implementation

[0042] To better understand the technical solution of the present invention, its specific embodiments will be described in detail below with reference to the accompanying drawings. It should be understood that the described embodiments are only used to illustrate the technical principles and method flow of the present invention, and are not intended to limit the present invention. Without departing from the core concept of the present invention, those skilled in the art can make appropriate changes and substitutions to the structure, steps, or parameters, all of which should fall within the protection scope of the present invention.

[0043] The single hydrophone distance-depth localization method based on multi-OFDM symbol joint processing described in this invention will be further explained below with reference to the accompanying drawings.

[0044] This implementation method is applicable to underwater applications such as... Figure 2 The integrated communication and positioning scenario shown is particularly designed for high-precision positioning in complex underwater acoustic environments. Specific applicable scenarios include, but are not limited to, underwater robot navigation, submarine autonomous positioning, and underwater sensor network node positioning.

[0045] Based on the existing underwater acoustic channel model, the following assumptions are further clarified:

[0046] The observation area meets the horizontal stratification condition, that is, both the bottom and the surface of the water remain flat;

[0047] The sound velocity in the water is uniformly distributed and constant within the observation area, thus avoiding the impact of sound velocity variations on positioning accuracy.

[0048] Ignoring the sound wave scattering effect caused by underwater obstacles, the sound wave propagation path follows the law of ray propagation.

[0049] The model parameters are defined as follows:

[0050] The water depth is H;

[0051] The height of the transmitter above the seabed is Z. t ;

[0052] The height of the receiver above the bottom of the water is Z. r ;

[0053] The horizontal distance between the transmitting and receiving ends is R.

[0054] To address the demands for precise positioning and stable communication in complex underwater acoustic environments within the aforementioned application scenarios, a systematic solution that balances multipath propagation characteristics and channel time-varying features needs to be developed. For example... Figure 1 As shown, this embodiment, based on core technologies such as multipath channel modeling, OFDM signal design, and frequency domain processing, designs a highly robust method integrating communication and positioning functions. The key technical solutions of this embodiment will be described in detail below.

[0055] Step 1: Multipath Channel Modeling

[0056] In real-world scenarios, not only direct sound paths, surface reflection paths, and bottom reflection paths are considered, but also possible multiple reflection paths, such as combined bottom-surface reflection and surface-bottom reflection paths, are further considered to construct a more complete and diverse multipath propagation model.

[0057] Typical path distance is represented as:

[0058] Direct sound path:

[0059]

[0060] Water surface reflection path:

[0061]

[0062] underwater reflection path:

[0063]

[0064] The corresponding transmission delay is:

[0065]

[0066] The explicit definition of the theoretical time delay vector τ:

[0067] Define the set of multipath propagation paths as P = {1, 2, ..., N}, where N is the total number of paths considered (including direct paths, surface reflections, underwater reflections, and their combinations). For each path i ∈ P, let its propagation distance be d. i The corresponding propagation delay is:

[0068]

[0069] Where, d i Let be the propagation distance along the i-th path (in meters), and c be the speed of sound in water, assumed to be constant within the observation region. The overall time delay vector is then constructed as follows:

[0070] τ=[τ1,τ2,…,τ N ] T

[0071] Each element is arranged in path number order, with a dimension of N×1, and is used to describe the propagation delay characteristics of all observed paths. This vector serves as the basic input parameter for maximum likelihood estimation and subsequent localization optimization calculations.

[0072] The path numbers 1-N correspond to direct, surface reflection, bottom reflection and their combination paths, respectively. The numbering order can be set according to the propagation distance or the number of reflections, and can also be adaptively adjusted in actual deployment to enhance the versatility of modeling.

[0073] Step 2: OFDM Signal Design

[0074] To improve positioning accuracy and communication reliability, this implementation increases the pilot density in OFDM symbols and employs differentially coded pilot sequence design to enhance signal anti-interference performance and improve the accuracy of time delay estimation.

[0075] The l-th OFDM symbol is represented as:

[0076]

[0077] in:

[0078] The pilot symbol for the l-th symbol and the k-th subcarrier;

[0079] K is the total number of subcarriers.

[0080] Step 3: Frequency Domain Reception Model

[0081] This embodiment performs frequency domain filtering on the received signal to reduce the impact of noise interference, and uses a frequency domain adaptive estimation method to improve the estimation accuracy of the path complex fading coefficient.

[0082] The frequency domain vector of the received l-th symbol is represented as:

[0083] y (l) =A(τ)·s (l) +n (l) ,l=1,2,...,L

[0084] in:

[0085] A(τ)∈C K×P The channel matrix consists of P paths, and its columns are the frequency domain responses corresponding to different time delays.

[0086] s (l) ∈C P×1 Let be the complex fading coefficients of each path in the l-th symbol;

[0087] n (l) ∈C K×1 This is an additive noise vector.

[0088] Based on the above model, and further based on the channel stationarity assumption, let L receive vectors be used. They share the same channel response matrix A(τ), which reflects the frequency domain response characteristics of each path, providing a consistent modeling basis for multi-symbol collaborative processing. Considering the randomness of the underwater acoustic channel, the path gain coefficient... The channels are independent and identically distributed, satisfying a zero-mean complex Gaussian random process, meaning the channel gain satisfies the Rayleigh fading model.

[0089] The channel response matrix A(τ) can be expressed in the form of:

[0090]

[0091] Where {f1,f2,...,f K Let f be the frequency corresponding to the K pilot subcarriers. k Let τ be the frequency corresponding to the k-th pilot subcarrier. p Let τ be the propagation delay of the p-th candidate path. Let τ be a column vector of length P, representing the set of propagation delays for all candidate paths. p Let p be the p-th component.

[0092] Step 4: Construction of Multi-Measure Vector Model (MMV)

[0093] This implementation proposes a multi-symbol cooperative processing method, which combines received vectors from different times to form a unified measurement matrix, and further improves the multipath delay estimation accuracy by utilizing matrix sparse reconstruction technology.

[0094] By concatenating L received vectors, a total measurement matrix can be constructed:

[0095] Y=A(τ)S+N

[0096] Y = [y (1) ,...,y (L) ]∈C K×L S = [s (1) ,...,s (L) ]∈C P×L ;N∈C K×L This is the noise matrix.

[0097] Step 5: Delay Estimation and Location

[0098] Maximum likelihood time delay estimation:

[0099] Iterative optimization algorithms, such as gradient descent or Newton's method, can be used to improve the computational efficiency and accuracy of maximum likelihood estimation and to estimate the time delay vector.

[0100]

[0101] in, To estimate the time delay vector, Tr[.] denotes the trace operation of the matrix, i.e., the sum of all diagonal elements. This indicates that the matrix is ​​taking its pseudo-inverse.

[0102] Target location estimation:

[0103] Consider the target's two-dimensional position q = [R, D], where R represents the horizontal distance and D represents the depth, respectively. The propagation delay of path i is:

[0104]

[0105] Define the positioning error cost function:

[0106]

[0107] The final estimated location is obtained by minimizing this cost function, yielding the optimal location estimate:

[0108]

[0109] Based on the constructed positioning error cost function, this implementation uses Kalman filtering or particle filtering methods to dynamically estimate the target position, thereby achieving continuous tracking of the target state and optimal position estimation. In practical applications, the specific filtering algorithm can be selected based on the complexity of the target's motion state: when the target motion model is relatively linear and the system noise is approximately Gaussian, Kalman filtering is preferred to improve computational efficiency; while under nonlinear, non-Gaussian modeling conditions, particle filtering is preferred to enhance tracking robustness.

[0110] In practical application verification, such as Figure 3 As shown, the method of this invention was experimentally tested in a standardized 50m × 50m × 5m water tank environment. Compared with the traditional single OFDM symbol processing method, the positioning accuracy is significantly improved when three OFDM symbols are used for joint processing. This embodiment has been comprehensively verified in various experimental environments, including computer simulation and actual water body testing. Experimental results show:

[0111] The positioning error of a single OFDM symbol is approximately ±0.35m;

[0112] After joint positioning using multiple OFDM symbols, the error was significantly reduced to ±0.12m;

[0113] In low signal-to-noise ratio environments, the positioning error can be reduced by up to 60% through multi-symbol joint processing;

[0114] In dynamic target positioning scenario testing, the positioning and tracking error of this implementation method is reduced by about 40% compared with the traditional method; under different water depths and changes in the transmitting and receiving positions, this implementation method shows good robustness and environmental adaptability.

[0115] Compared with existing positioning methods based on single symbol or energy analysis, this invention can achieve smaller positioning errors and higher multipath resolution under the same channel conditions. In particular, it still exhibits excellent stability in low signal-to-noise ratio and strong reflection backgrounds, and has stronger environmental adaptability and engineering feasibility.

[0116] This invention's method can be flexibly integrated into various underwater platforms, such as UUV (Unmanned Underwater Vehicle) navigation systems or buoy monitoring nodes. By configuring and switching different water depth environmental parameters through software, it can quickly adapt to different underwater environments such as nearshore, deep sea, and lakes, achieving low power consumption, high versatility, and rapid deployment underwater target positioning capabilities.

[0117] To further verify the applicability of the method of the present invention in different application scenarios, four typical embodiments are listed below.

[0118] Application Example 1: Buoy Platform

[0119] In a preferred embodiment, the single hydrophone range-depth localization method based on multi-OFDM symbol joint processing described in this invention is integrated into a buoy platform to construct a passive underwater acoustic localization system in nearshore environments. This system can achieve autonomous localization of underwater targets without requiring multi-point deployment or array coordination, and is suitable for scenarios such as illegal intrusion monitoring, nearshore operation status awareness, and underwater target tracking.

[0120] The buoy platform has a compact structure, consisting of a floating body, hydrophone components, a signal processing module, and a data communication module. The hydrophone is fixed to the lower part of the buoy, pointing vertically into the underwater space, and is used to receive OFDM acoustic signals propagating in the water. The signal processing module is an integrated low-power embedded system with functions including signal acquisition, pilot extraction, frequency domain modeling, path estimation, and position calculation. Its core processing flow runs the method of this invention, which, through joint processing of multiple consecutive OFDM symbols, achieves path delay estimation and two-dimensional spatial positioning of the target sound source.

[0121] In practical deployment, the target sound source emits an OFDM signal containing a pilot signal from underwater at a fixed period. This signal propagates through the water via a direct path, a surface reflection path, and a bottom reflection path before being received by a hydrophone located at the bottom of the buoy. The receiver performs frequency domain analysis on multiple symbols, constructs a multi-measurement vector model, and jointly extracts multi-path propagation features under the channel stationary assumption. Subsequently, the propagation delay parameters for each path are obtained using the maximum likelihood estimation method. Combined with known water depth conditions and a geometric sound propagation model, the horizontal distance and depth coordinates of the target relative to the buoy are then retrieved.

[0122] In experiments conducted in nearshore waters of a port, the buoy was deployed at a depth of approximately 10 meters, and the target sound source moved slowly within a 30-meter radius around the buoy. The buoy platform processed three consecutive OFDM symbols in each signal reception cycle, with an average system cycle time of 4 seconds. Results showed that, without relying on any external synchronization equipment, the platform achieved an average horizontal positioning error of 0.9 meters and a depth error of 0.6 meters, with a maximum recognition radius exceeding 45 meters. Multipath resolution was improved by approximately 38% compared to traditional single-symbol processing, while maintaining good stability even in the presence of background noise.

[0123] Furthermore, the platform supports remote configuration and parameter adjustment, enabling it to switch OFDM symbol structure, pilot distribution mode, and sound velocity model parameters according to the aquatic environment, achieving rapid adaptation to different deployment areas. The platform's overall power consumption is less than 2W, allowing for continuous operation powered by solar panels, and it possesses long-term deployment and unattended operation capabilities.

[0124] Therefore, the method of the present invention, when applied to a buoy platform, can still achieve high-precision two-dimensional positioning of underwater sound sources in environments with limited space, single deployment, and isolated communication, and has significant practical value and engineering application prospects.

[0125] Application Example 2: Deep-sea UUV Platform

[0126] In another preferred embodiment, the single hydrophone range-depth localization method based on multi-OFDM symbol joint processing described in this invention is applied to a deep-sea unmanned underwater vehicle (UUV) platform to construct a navigation module with autonomous passive localization capabilities. In traditional deep-water operations, UUVs, due to space and power consumption constraints, often cannot be equipped with multi-element receiver arrays or high-precision synchronization systems, limiting their autonomous localization capabilities in complex underwater environments. The localization method proposed in this invention minimizes hardware configuration requirements, relying solely on a single hydrophone and an embedded processing module to achieve spatial localization of target sound sources in deep-water environments, providing independent and low-cost passive navigation support for the UUV platform.

[0127] In this embodiment, the UUV is equipped with an integrated acoustic sensing and positioning system, including a hydrophone installed at the front of the hull, an analog front-end circuit for receiving underwater OFDM signals, a low-power signal processing unit based on a DSP or ARM architecture, and a task management module based on a real-time operating system (RTOS). The hydrophone continuously monitors pilot-containing OFDM signals emitted by the work vessel, underwater beacon, or other autonomous sound source nodes in the environment. The signal processing module preprocesses the received multi-symbol data according to a preset frequency, extracts frequency domain pilot subcarriers to construct a multi-measurement vector model, and performs maximum likelihood estimation of the path propagation delay based on the method described in this invention.

[0128] Considering the nonlinear variation of sound speed with depth in the deep-sea environment, the navigation module introduces a ray acoustic propagation model to correct the path structure. Specifically, based on environmental parameters obtained from the depth gauge and CTD sensor onboard the UUV, the system autonomously constructs a layered sound speed profile and solves for the curved trajectory corresponding to the propagation path, replacing the assumption of straight-line propagation in shallow water. During path inversion, a position cost function is constructed based on the corrected propagation distance and estimated time delay to achieve two-dimensional positioning of the sound source relative to the UUV. In this embodiment, the UUV is equipped with a CTD sensor, which updates the sound speed profile data every 30 seconds and inputs it into the positioning algorithm. The measured sound speed at a depth of 300 meters is 1523 m / s, which reduces the positioning error by 22% compared to the default value of 1500 m / s.

[0129] In a deep-sea test range at a depth of 500 meters, the UUV conducted path tracking and sound source proximity tests within an operating radius of no more than 1 kilometer. Under conditions of joint processing using three OFDM symbols, the system could complete a positioning update every 8 seconds. Test results show that at different depth levels, the horizontal distance error to the sound source obtained by the UUV is less than 2.3 meters, and the depth estimation error is controlled within 1.0 meter. Even in the presence of sound velocity jump layers or multipath interference, the module still exhibits good stability and convergence.

[0130] This navigation module can be integrated as an independent positioning unit into multiple UUV models without relying on external base stations or underwater navigation cables. It is suitable for long-term autonomous operation scenarios in tasks such as oil and gas exploration, deep-sea resource surveys, and seabed structure inspection. The system supports rapid switching between different OFDM standards and propagation models through a parameter table configuration interface, improving the algorithm's versatility and platform adaptability.

[0131] This embodiment verifies the effectiveness and robustness of the method of the present invention in the deep-sea environment, and demonstrates its broad application potential in space-constrained, heterogeneous underwater platforms.

[0132] Application Example 3: Lake Distribution Nodes

[0133] In another preferred embodiment, the single-hydrophone distance-depth localization method based on multi-OFDM symbol joint processing described in this invention is applied to static water environments such as lakes or reservoirs to construct a distributed sound source localization system that does not require synchronization or coordination. This system consists of several independently deployed single-hydrophone monitoring nodes, each autonomously running the method of this invention to estimate the relative position of underwater sound sources without relying on array structures or external time synchronization systems. This embodiment fully demonstrates the adaptability of this invention in decentralized, low-deployment-complexity scenarios.

[0134] In practical application, the system is deployed in a large reservoir area with a water width of approximately 1 kilometer and a maximum depth of 30 meters. Three hydrophone nodes are equidistantly positioned along the reservoir bank. Each node consists of an anchoring structure, a hydrophone, a float power system, an edge processing module, and a wireless data transmission terminal. The hydrophones are located at a designated underwater depth, directionally monitoring OFDM pilot signals periodically emitted by underwater target devices or artificial sound sources within the target water area. Because the monitoring nodes are not interconnected, the system as a whole does not rely on GPS timing or wireless link synchronization. Instead, each node independently locates itself, and then the relative results are transmitted back to the central system for fusion.

[0135] At each monitoring node, the method of this invention takes three consecutively received OFDM symbols as input and extracts the multipath propagation delay through frequency domain pilot extraction, MMV model construction, and maximum likelihood estimation. Then, combining the node deployment depth and terrain / water depth data, a propagation geometry model is established, and the horizontal offset and depth coordinates of the target relative to that node are estimated. Each node uploads the positioning results and confidence index to the host computer. The central terminal performs fusion calculations based on the node's spatial location and positioning information, and outputs the target's positioning result in the global coordinate system.

[0136] Because this system does not require synchronous operation of all nodes, it offers extremely high deployment flexibility, making it particularly suitable for temporary operational scenarios with limited communication, no base station support, or requiring rapid deployment. Actual testing shows that in a typical reservoir environment, when a target moves slowly within the coverage area formed by the three monitoring nodes, the system achieves an average positioning error of 1.4 meters and a depth error of approximately 0.9 meters. Even under adverse conditions such as partial signal loss or enhanced local reflection, the positioning error remains within 2.5 meters, demonstrating excellent overall system stability and scalability.

[0137] The application of the method in this invention not only significantly reduces node costs and system deployment difficulty, but also provides a technical foundation for building a low-coordination, highly autonomous, flexible, and scalable underwater positioning network by improving the positioning capabilities of individual nodes. The system can be widely used in still water areas such as lake ecological monitoring, illegal fishing early warning, and aquatic behavior analysis.

[0138] Application Example 4: Earthquake Early Warning Node

[0139] In another preferred embodiment of the present invention, the single hydrophone distance-depth localization method based on multi-OFDM symbol joint processing is integrated into an underwater early warning node for earthquake monitoring and sound source inversion, enabling autonomous localization of earthquake sound source locations in deep-sea environments. Currently, most seabed earthquake monitoring systems rely on cable communication or underwater array structures to obtain epicenter information. However, in application scenarios lacking real-time communication capabilities, with limited deployment density, or requiring high positioning accuracy, existing methods struggle to meet the dual requirements of response speed and spatial resolution.

[0140] In this embodiment, the underwater early warning node is fixedly deployed on a seabed base and equipped with a single hydrophone, a data acquisition unit, a high-performance signal processing module, and a local storage system. The entire system possesses breakpoint retransmission and delayed reporting capabilities, supporting long-term unattended operation. When an earthquake occurs, the node receives underwater acoustic signals generated by the seismic source. These signals typically contain multiple multipath acoustic energy components in the near-field region, exhibiting a frequency domain energy distribution approximating an OFDM structure. Alternatively, an OFDM signal containing pilot frequencies can be constructed through digital preprocessing to adapt to the method of this invention.

[0141] The signal processing module performs short-time Fourier transform (STFT) processing on the received vibration signal, extracting multiple consecutive frames as "equivalent OFDM symbols" to construct a pilot distribution structure in the frequency domain. Subsequently, the multi-measurement vector modeling and maximum likelihood estimation method described in this invention is used to extract the main path propagation delay from the sound source to the node. Combined with the layered sound velocity profile and terrain structure model, the sound wave propagation path is corrected, and a time-delay-space inversion model is constructed to estimate the relative two-dimensional position of the earthquake sound source.

[0142] In a simulated environment, taking a scenario with a horizontal distance of 1.8 kilometers from the epicenter to the node and a water depth of approximately 500 meters as an example, the node can complete a full sound source inversion process within 20 seconds, with an average positioning error controlled at 3.2 meters and a depth error of less than 1.5 meters. The results demonstrate that even without satellite signals or array structure support, effective localization of sudden sound sources on the seabed can still be achieved, improving the timeliness and accuracy of the seabed monitoring system.

[0143] By embedding the method of this invention into early warning nodes, the system can be deployed in seismically active zones, underwater fault zones, or submarine volcanic areas to form a distributed acoustic early warning network, achieving large-scale, low-latency, and low-energy-consumption epicenter sensing capabilities. It has multiple application values, including emergency response, earthquake scientific research, and submarine structure risk management.

[0144] In summary, this invention provides a single hydrophone range-depth localization method based on multi-OFDM symbol joint processing. By constructing a multi-measurement vector model and a multi-path delay joint estimation mechanism, it achieves high-precision two-dimensional sound source localization relying solely on a single node without requiring multiple array elements or synchronous coordination. This method fully exploits the underwater multipath propagation characteristics, exhibiting good robustness and adaptability in environments with strong noise, complex hydrology, and limited space, significantly reducing system hardware dependence and deployment complexity.

[0145] As can be seen from the description of several specific embodiments, the method of the present invention is applicable to a variety of typical application scenarios, such as nearshore buoy platforms, deep-sea UUV systems, lake distributed monitoring networks, and seabed seismic sound source inversion. It has significant advantages such as strong platform versatility, flexible deployment, low power consumption, and accurate positioning, and can effectively meet the comprehensive requirements of current underwater sensing systems for accuracy, stability, and engineering feasibility.

[0146] This invention not only breaks through the limitations of existing single hydrophone positioning methods in terms of accuracy and resolution, but also provides a new technological path for building high-performance, low-cost underwater passive positioning systems, and has broad prospects for engineering promotion and research expansion value.

[0147] It should be noted that although this specification has described the invention in detail with reference to the accompanying drawings and specific embodiments, the scope of protection of the invention is not limited to the above embodiments. Various equivalent modifications or substitutions made to the invention by those skilled in the art without departing from the spirit and essence of the invention should also be included within the scope of protection of the invention.

Claims

1. A single hydrophone range-depth localization method based on multi-OFDM symbol joint processing, characterized in that, Includes the following steps: S10, the underwater target sound source continuously transmits multiple orthogonal frequency division multiplexing (OFDM) symbols containing pilot subcarriers, each symbol embedding pilot information of known structure in the frequency domain; S20, the single hydrophone at the receiving end receives the multiple OFDM signals, extracts the frequency domain pilot subcarrier data from each symbol, and forms multiple pilot response vectors; S30: Based on multiple pilot response vectors, a multi-measurement vector (MMV) model is constructed. By establishing a linear mapping relationship between the receiver matrix and the frequency domain response matrix of the candidate path, maximum likelihood estimation is performed to extract the multi-path propagation delay vector. S40, combining the geometric propagation model between the sound source and the receiver, constructs a path propagation cost function, matches the estimated multipath propagation delay with the theoretical propagation delay, and obtains the horizontal distance-depth estimate of the sound source by minimizing this cost function.

2. The method according to claim 1, characterized in that, The pilot subcarriers are uniformly distributed in the frequency domain within each OFDM symbol, providing frequency response distinguishability and enhancing the ability to distinguish multipath signals.

3. The method according to claim 1, characterized in that, The channel response matrix in the MMV model is composed of frequency domain response templates under multipath propagation delay, with each column corresponding to a path propagation feature.

4. The method according to claim 1, characterized in that, The maximum likelihood estimation obtains the optimal propagation delay vector by constructing an orthogonal projection matrix and searching for the optimal path combination in the propagation delay parameter space.

5. The method according to claim 1, characterized in that, The path propagation cost function is used to describe the error of multipath delay corresponding to the hypothetical and real sound source locations. Ideally, the target location is the coordinate point that minimizes the function.

6. The method according to claim 5, characterized in that, The propagation model includes a direct path, a surface reflection path, and a bottom reflection path. The propagation delay is obtained by dividing the propagation distance of each path by the speed of sound in the water.

7. The method according to claim 5 or 6, characterized in that, The cost function can be minimized using grid search or local gradient descent algorithms, and path weights can be introduced to improve estimation accuracy.

8. The method according to claim 1, characterized in that, Multipath propagation delay feature extraction and two-dimensional sound source localization are two decoupled processing stages, which are executed independently based on frequency domain modeling and cost function optimization, respectively.

9. A single-node underwater acoustic passive positioning device deployed on a small platform, characterized in that, Configured to perform the method described in any one of claims 1 to 8, and output the positioning coordinates of the target sound source in the horizontal distance and depth directions.

Images