A multi-layer double-ring coupling-based array distributed synchronization system and method

By using a multi-layered dual-ring coupled array distributed synchronization system, the problems of timing synchronization, topology structure and signal interference separation of sensor arrays are solved. It achieves global self-synchronization and interference hierarchical localization, reduces costs and improves robustness, and is suitable for unified standardization of sensor arrays in multiple scenarios.

CN123074487APending Publication Date: 2026-11-10周伟华
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611259863.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-19
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing distributed sensor arrays have many problems in terms of timing synchronization, topology, signal aliasing, heterogeneous architecture, and panoramic multi-view acquisition, resulting in poor robustness, high cost, limited scalability, difficulty in separating signal interference and feature sharing, and inability to achieve large-area and irregular curved surface detection.

Method used

A distributed synchronization system based on multi-layer dual-ring coupling is adopted. Through the ring topology structure of the bottom cell, intermediate cell and convergent cell, global self-synchronization is achieved, eliminating external clock dependence. A multi-layer coupling network is formed by self-synchronization module, interference injection unit and comparison judgment unit to realize signal self-frequency stabilization, self-phase locking and interference hierarchical localization.

Benefits of technology

It achieves crystal-free self-synchronization across the entire domain, reducing hardware power consumption and cost, improving synchronization stability, supporting continuous operation in industry 24/7, providing fast response and high reliability, and is suitable for unified standardization of sensor arrays in multiple scenarios.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123074487A_ABST
    Figure CN123074487A_ABST
Patent Text Reader

Abstract

This invention provides a multi-layer dual-ring coupling array-based distributed synchronization system and method, relating to the field of sensor data synchronization. It includes multiple bottom-level cells, intermediate cells, and convergence cells. Multiple bottom-level cells are constructed according to a ring topology to form at least one bottom-level coupled sensing layer, achieving single-layer global self-synchronization. At least one intermediate-dimensional classification layer is constructed using multiple intermediate cells. At least one one-to-one corresponding top-level feature convergence layer is constructed, which outputs a globally fused feature signal. This system can be expanded planarly or vertically. It achieves global self-locking without an external crystal oscillator, microsecond-level hardware response, strong anti-interference capabilities, and is suitable for scenarios such as panoramic vision, industrial monitoring, bionic robots, and passive IoT.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of sensor data synchronization, specifically to an array-based distributed synchronization system and method based on multi-layer dual-ring coupling. Background Technology

[0002] Existing distributed sensor arrays, image sensor arrays, and distributed synchronous acquisition systems generally suffer from five major pain points, which are also the underlying problems that have long been difficult to eradicate in the fields of live streaming imaging, industrial sensing, and robot vision: 1. Timing synchronization relies on external references, resulting in poor robustness and high cost. Traditional arrays rely entirely on external crystal oscillators, central timing chips, and global clock trees for node synchronization, requiring additional synchronization traces, calibration circuits, and compensation algorithms. In scenarios with strong electromagnetic interference, weak networks, and large-scale array deployments, clock skew, phase lockout, frame tearing, and data packet loss are prone to occur. At the same time, external clocks and supporting circuits increase hardware size, power consumption, and material costs, making adaptation to mobile and embedded devices difficult.

[0003] 2. Weak topological correlation limits large-scale array expansion. Mainstream arrays use linear topology, star topology, and standard rectangular pixel arrays. Array nodes only achieve data interconnection and lack hardware-level coupling synchronization links. Array nodes, single-layer, and global arrays cannot achieve multi-level linkage synchronization. As the array size increases, the temporal dispersion continues to rise, making it difficult to support large-area, irregular curved surface, and ring-shaped distributed collaborative detection.

[0004] 3. Signal aliasing and lack of dimensional isolation make interference tracing difficult. Traditional sensor arrays transmit multiple signals in parallel without physical layering or dimensional isolation design, allowing local interference and signal distortion to spread throughout the entire array. They rely on backend MCUs, DSPs, and AI algorithms to complete signal separation, noise reduction, feature extraction, and interference localization, resulting in high computational overhead and high latency, making it impossible to achieve rapid hardware-level fault tracing.

[0005] 4. Severe architectural heterogeneity and lack of a universal, shared feature base. Existing imaging and sensing devices have independent hardware architectures, with image acquisition, industrial sensing, and depth perception using different circuit systems and feature formats, making feature sharing impossible. Device iteration and multi-scenario adaptation require redesigning the entire hardware, resulting in long R&D cycles, low reusability, and difficulty in forming a unified technical standard.

[0006] 5. Panoramic multi-view acquisition and large-area visual arrays have problems such as difficulty in multi-channel temporal alignment, stitching gaps, uneven brightness, and dynamic misalignment; traditional solutions rely on back-end algorithm compensation, which further aggravates the consumption of terminal and cloud computing power and bandwidth resources.

[0007] Therefore, in order to solve the above-mentioned technical problems and to facilitate the implementation of the cyclic phase transition topology system in the applicant's application 2026107263443 "A Phase Encoding Method and System Based on Cyclic Phase Transition Topology", there is an urgent need for an array distributed synchronization system and method based on multi-layer double-ring coupling. Summary of the Invention

[0008] To overcome the existing technical problems, the present invention provides an array distributed synchronization system and method based on multi-layer dual-ring coupling.

[0009] The present invention adopts the following technical solution.

[0010] A distributed synchronization system based on multi-layer dual-ring coupling includes multiple bottom cells and intermediate cells capable of receiving signals, outputting signals, and self-synchronizing, and also includes at least one convergent cell capable of receiving signals, outputting signals, and self-synchronizing. The bottom cells, intermediate cells, and convergent cells all have the same ring topology. Based on the ring topology, multiple bottom cells are constructed to form at least one bottom coupled sensing layer. The bottom cells within the same bottom coupled sensing layer have a weak coupling relationship to form a single-layer global self-synchronization. At least one intermediate dimension classification layer is formed by constructing multiple intermediate unit cells. The intermediate dimension classification layer is used to receive the output signal of the corresponding lower coupling sensing layer. The intermediate unit cells in the same intermediate dimension classification layer are independent of each other. At least one top-level feature convergence layer is formed by constructing at least one convergence cell. The top-level feature convergence layer is used to receive the output signal of the corresponding intermediate dimension classification layer and output the global fusion feature signal.

[0011] As a further improvement of the present invention, the bottom cell, the intermediate cell and the convergent cell all include a timing reference unit, an interference injection unit and a comparison and judgment unit; The timing reference unit includes a self-synchronization module, which can output a self-stabilized frequency and self-locked phase reference voltage. The self-synchronization module has a preset sensitive node. The input of the interference injection unit is used to receive external interference signals so that the bottom cell and intermediate cell can receive signals. The output of the interference injection unit is used to couple the external interference signals to the sensitive node to change the voltage of the sensitive node and form the disturbed voltage. The comparison and judgment unit includes a voltage comparator and a voltage divider potentiometer. The input terminal of the voltage divider potentiometer is electrically connected to the self-synchronization module, and the output terminal of the voltage divider potentiometer is electrically connected to the inverting terminal of the voltage comparator to provide an adjustable reference voltage to the voltage comparator. The non-inverting terminal of the voltage comparator is electrically connected to the sensitive node to compare the disturbed voltage of the sensitive node with the adjustable reference voltage to form a node interference indication signal.

[0012] As a further improvement of the present invention, the self-synchronization module includes a first inner loop oscillation component, a first outer loop oscillation component, a node coupling link, and an output interface to form a ring topology. Both the first inner ring oscillation component and the first outer ring oscillation component are formed by cascading at least two oscillation nodes; The node coupling link is located between the first inner loop oscillation component and the first outer loop oscillation component. It can inject the output signal of the first inner loop oscillation component with a frequency of the first initial oscillation frequency into the first outer loop oscillation component with a preset coupling coefficient, and inject the output signal of the first outer loop oscillation component with a frequency of the second initial oscillation frequency into the first inner loop oscillation component with a preset coupling coefficient, so that the oscillation frequencies of the first inner loop oscillation component and the first outer loop oscillation component tend to be consistent and the phase difference is locked. The oscillation node is equipped with a frequency configuration network, which is used to determine the first initial oscillation frequency and the second initial oscillation frequency. The output interface is used to output a reference voltage.

[0013] As a further improvement of the present invention, the timing reference unit also includes a digital configuration device, which includes a digital configuration interface, adjustable resistor-capacitor components, and a digital potentiometer. The adjustable resistor-capacitor components and the digital potentiometer are electrically connected to the digital configuration interface. The frequency configuration network is composed of fixed resistor-capacitor components and adjustable resistor-capacitor components. The adjustable resistor-capacitor components are programmable capacitors or digitally adjustable resistors. The coupling device of the node coupling link adopts a digital potentiometer. The digital configuration device can receive external digital configuration commands and configure the coupling coefficient, the first initial oscillation frequency, the second initial oscillation frequency, and the phase difference according to the external digital configuration commands.

[0014] As a further improvement of the present invention, it also includes a multi-layer coupling network, which includes node coupling links disposed in the bottom cell, intermediate cell and convergent cell. Multilayer coupling networks also include intralayer coupling links between different bottom-layer cells within the bottom coupling sensing layer; The multilayer coupled network also includes cross-layer coupling links between the bottom coupled perception layer and the intermediate dimension classification layer, and between the intermediate dimension classification layer and the top feature convergence layer.

[0015] As a further improvement of the present invention, both the first inner ring oscillation component and the first outer ring oscillation component are formed by cascading three oscillation nodes; The bottom-level coupled sensing layer is constructed from six bottom-level unit cells, and the intermediate-dimensional classification layer is constructed from six intermediate unit cells, so as to form six output signals representing six dimensions.

[0016] As a further improvement of the present invention, the bottom coupling perception layer, the intermediate dimension classification layer and the top feature convergence layer are all multiple, and each intermediate unit cell has a fixed dimensional interface. Based on the preset numbering rules, each intermediate cell under the same intermediate dimension classification layer is assigned a unique dimension number corresponding to the dimension interface, and adjacent intermediate dimension classification layers are connected through the dimension interface signal corresponding to the same unique dimension number.

[0017] As a further improvement of the present invention, the output interface includes a working output interface and a phase difference output interface. The reference voltage is output through the working output interface, and the phase difference output interface is used to output multiple standardized timing phase signals.

[0018] As a further improvement of the present invention, an array node can be formed by a bottom coupling perception layer, an intermediate dimension classification layer and a top feature convergence layer, and adjacent array nodes follow the same dimension point-to-point docking rule.

[0019] This invention also proposes a distributed array synchronization method based on multi-layer dual-ring coupling, applied to the above-mentioned distributed array synchronization system based on multi-layer dual-ring coupling, comprising the following steps: S1. Start the self-synchronization module in the bottom cell, intermediate cell and convergent cell to achieve autonomous oscillation and enter the initial steady state; S2. Initiate node coupling links within the bottom cell, intermediate cell, and convergent cell to achieve self-stabilization and self-locking of the bottom cell, intermediate cell, and convergent cell; S3. Start the intra-layer coupling link of the bottom coupling perception layer to achieve single-layer global self-synchronization within the bottom coupling perception layer; S4. Activate the cross-layer coupling link between the bottom coupling perception layer and the intermediate dimension classification layer so that the output signal of the bottom coupling perception layer can be transmitted to the corresponding intermediate dimension classification layer. S5. Activate the cross-layer coupling link between the intermediate dimension classification layer and the top-level feature convergence layer so that the output signal of the intermediate dimension classification layer is transmitted to the corresponding top-level feature convergence layer, and the top-level feature convergence layer outputs a global fusion feature signal.

[0020] The beneficial effects of the present invention include at least the following: 1. Full-domain crystal-free self-synchronization, completely eliminating dependence on external clocks. Relying on the bottom layer cell, intermediate cell, convergent cell and multi-layer coupling network to achieve full-domain timing self-locking, eliminating the need for external crystal oscillators, clock trees, and central timing circuits, the hardware power consumption, size and material cost are significantly reduced. The synchronization stability under strong electromagnetic interference, weak network and complex working conditions is significantly improved, fundamentally solving problems such as video imaging frame tearing, data packet loss and timing offset.

[0021] 2. Layered dimensional isolation provides extremely strong interference localization capabilities. The decoupled design of the intermediate dimensional classification layer achieves six-dimensional physical isolation, ensuring that local interference is limited to the corresponding dimension and spatial unit and does not spread globally. Combined with the two-dimensional address of "spatial coordinates + dimensional coordinates", interference source tracing can be completed quickly, significantly reducing the difficulty of operating and maintaining large arrays.

[0022] 3. Pure hardware-based system with fast response and high reliability. The entire system uses analog hardware circuitry, eliminating digital computation and software latency, achieving microsecond-level response; it supports continuous 24 / 7 industrial operation and has superior anti-interference capabilities compared to the traditional "sensor + MCU + algorithm" architecture. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a schematic diagram of the three-layer interconnection architecture of the present invention; Figure 2 This is a simplified diagram of the three-layer interconnection architecture of the present invention; Figure 3 This is a schematic diagram of the ring topology of the present invention; Figure 4 This is a schematic diagram of the structure of the top-level feature convergence layer and the intermediate dimension classification layer of the present invention; Figure 5 This is a schematic diagram of a graphene-like array arrangement according to another embodiment of the present invention; Figure 6 This is a simplified view of the extended connection according to another embodiment of the present invention. Detailed Implementation

[0025] The accompanying drawings are for illustrative purposes only and should not be construed as limiting the scope of this patent. To better illustrate this embodiment, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual dimensions of the product.

[0026] It will be understood by those skilled in the art that certain well-known structures and their descriptions may be omitted in the accompanying drawings. The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0027] This invention is the system-level networking and large-scale implementation of the parent application 2026107263443 "A phase encoding method and system based on cyclic phase change topology", the sub-applications 2026110201654 "A self-synchronization module and phase control method based on double-ring coupled oscillation", and the sub-application 2026111656892 "An interference detection system and control method based on double-ring coupled oscillation".

[0028] The three prior applications mentioned above form an inseparable and complete technical protection chain with this invention: Sub-application "A Self-Synchronization Module and Phase Control Method Based on Dual-Ring Coupled Oscillation" provides a single-node dual-ring self-synchronization timing source, which is the principle basis of cell self-synchronization; Sub-application "An Interference Detection System and Control Method Based on Dual-Ring Coupled Oscillation" provides a standardized interference detection cell, which is the bottom cell, intermediate cell and convergence cell of this invention; The parent application provides a six-dimensional phase encoding and analysis system, which is the reason why the preferred bottom coupling sensing layer of this invention is formed by constructing six bottom cells and the intermediate dimension classification layer is formed by constructing six intermediate cells.

[0029] This invention realizes multi-cell hierarchical networking, global cellular interconnection, and three-dimensional stacking expansion. The four inventions are linked together to cover the entire chain of "time source (self-synchronization module) - smallest layer unit (cell, which is the superordinate description of the bottom cell, intermediate cell and convergent cell of this invention) - hierarchical array - global large-scale network - phase feature analysis (parent application)".

[0030] This system uses the interference detection system (same as the interference detection cell, i.e., the bottom cell, intermediate cell, and convergence cell of this invention) from the sub-application "An Interference Detection System and Control Method Based on Dual-Ring Coupled Oscillation" as the smallest hardware unit, relying on the self-synchronization module built into the interference detection cell to achieve a self-excited timing reference. The three-layer architecture of this invention can interface upwards with phase sampling, phase encoding, and cyclic phase transition analysis systems; downwards, it can be adapted to various sensor arrays such as visible light imaging, infrared sensing, depth sensing, and industrial condition monitoring. Its applications extend to panoramic live-streaming visual acquisition, robot visual perception, distributed interference identification in large-scale areas, full-cycle condition monitoring of industrial ring equipment, and graphene-like topology IoT sensor networks. Combining the above information provides a better understanding of the technical solution and beneficial effects of this invention.

[0031] Reference Figures 1 to 6 As can be seen, the present invention proposes an array distributed synchronization system based on multi-layer dual-ring coupling, including multiple bottom cells and intermediate cells capable of receiving signals, outputting signals and self-synchronizing, and at least one convergent cell capable of receiving signals, outputting signals and self-synchronizing. The bottom cells, intermediate cells and convergent cells all have the same ring topology. Based on the ring topology, multiple bottom cells are constructed to form at least one bottom coupled sensing layer. The bottom cells within the same bottom coupled sensing layer have a weak coupling relationship to form a single-layer global self-synchronization. At least one intermediate dimension classification layer is formed by constructing multiple intermediate unit cells. The intermediate dimension classification layer is used to receive the output signal of the corresponding lower coupling sensing layer. The intermediate unit cells in the same intermediate dimension classification layer are independent of each other. At least one top-level feature convergence layer is formed by constructing at least one convergence cell. The top-level feature convergence layer is used to receive the output signal of the corresponding intermediate dimension classification layer and output the global fusion feature signal.

[0032] By connecting the output signal of the intermediate dimension classification layer to the convergent unit cell, the physical attention mechanism can be implemented natively in hardware to achieve multi-dimensional feature weighted convergence.

[0033] This system enables the formation of a purely physical, universal, and shared feature base, serving as a unified underlying platform spanning visual imaging, industrial sensing, and robotic perception. Signal formats, coupling rules, and timing standards are unified across all scenarios, compatible with various sensor arrays including visible light, infrared, Time-of-Flight (TOF), and industrial sensors. The front end can interface with any type of sensor (visible light image sensors, fisheye panoramic sensors, infrared sensors, industrial vibration sensors, etc.), while the back end outputs a unified feature package format. Backend semantic encoding, phase resolution systems, stitching, and algorithm modules require no secondary modification. The entire array outputs a standardized six-dimensional feature package, seamlessly integrating with the parent application's six-dimensional phase encoding system. Furthermore, the system adapts to the transitional engineering form of existing coupling boards, meeting the requirements for mass production; hardware reserves pins, traces, and coupling points, allowing for smooth upgrades to a fully homogeneous topology. The system's bottom-layer coupling perception layer, intermediate-dimensional classification layer, and top-layer feature convergence layer can be integrated within a single analog chip.

[0034] For ring topology, please refer to Figure 3 As shown, the ring topology of the bottom coupled sensing layer is consistent with the self-synchronization modules of the bottom cell, intermediate cell and convergent cell.

[0035] Intermediate cells within the same intermediate dimension classification layer are independent, achieving complete physical isolation, and single-channel disturbances do not propagate across dimensions. Specific signal rules, docking rules, and connection methods will be described in detail in the self-synchronization module later. It should be noted that independence specifically refers to physical isolation without electrical coupling, not just logical independence.

[0036] Furthermore, it should be noted that this invention is not limited to three layers; see references... Figure 6 Alternatively, it can be based on a three-layer stack, for example, the output of the top feature convergence layer is coupled to another bottom layer coupled to the perception layer. (See reference...) Figure 5 It can also be a horizontal connection and stacking, such as connecting intermediate dimension classification layers to other intermediate dimension classification layers.

[0037] It should be noted that signal transmission between layers uses a unidirectional bus.

[0038] The invention offers specific advantages in different scenarios as follows: Live streaming / panoramic visual scenarios: multi-channel acquisition time sequence is naturally aligned, eliminating problems such as stitching gaps, uneven brightness, and dynamic misalignment, and resulting in a precipitous drop in computing power and bandwidth costs for both the terminal and the cloud. Industrial ring equipment monitoring: The array perfectly matches the shape of the ring equipment, enabling distributed collaborative detection with full coverage; Robot vision: Cellular array adapted to curved eye sensor, irregular curved surface layout with no blind spots; Internet of Things (IoT) sensor network: The hexagonal dense cellular topology array has strong topology expansion capabilities and is suitable for distributed sensing in large-area fields.

[0039] As a further improvement of the present invention, the bottom cell, the intermediate cell and the convergent cell all include a timing reference unit, an interference injection unit and a comparison and judgment unit; The timing reference unit includes a self-synchronization module, which can output a self-stabilized frequency and self-locked phase reference voltage. The self-synchronization module has a preset sensitive node. The input of the interference injection unit is used to receive external interference signals so that the bottom cell and intermediate cell can receive signals. The output of the interference injection unit is used to couple the external interference signals to the sensitive node to change the voltage of the sensitive node and form the disturbed voltage. The comparison and judgment unit includes a voltage comparator and a voltage divider potentiometer. The input terminal of the voltage divider potentiometer is electrically connected to the self-synchronization module, and the output terminal of the voltage divider potentiometer is electrically connected to the inverting terminal of the voltage comparator to provide an adjustable reference voltage to the voltage comparator. The non-inverting terminal of the voltage comparator is electrically connected to the sensitive node to compare the disturbed voltage of the sensitive node with the adjustable reference voltage to form a node interference indication signal.

[0040] As a further improvement of the present invention, the output interface includes a working output interface and a phase difference output interface. The reference voltage is output through the working output interface, and the phase difference output interface is used to output multiple standardized timing phase signals.

[0041] This technical solution has three core functions. First, interference detection eliminates the need for an external crystal oscillator, reducing cost and size. It offers high stability, low drift, and a purely hardware-based detection mechanism with fast response and strong anti-interference capabilities. Second, it outputs multiple standardized timing phase signals, as well as high-level or low-level alarm signals, as node interference indication signals and singularity marker signals to the back-end encoding system via a phase difference output interface. This enables the back-end encoding system to perform a series of operations, including cyclic phase transition analysis, phase encoding, chaotic transition state marking, timing reconstruction, and anomaly warning. Cyclic phase transition analysis, chaotic transition state marking, and phase encoding are the core concepts and operations in the parent application 2026107263443, "A Phase Encoding Method and System Based on Cyclic Phase Transition Topology." Third, it features array-based deployment, achieving distributed collaborative detection through weakly coupled synchronization, and enabling independent configuration of array partition parameters through a digital configuration device.

[0042] Specifically, the sensitive node in this invention is the CV pin, which does not refer to a standard pin of a particular chip. It is named using the semantics of ControlVoltage. The output of the interference injection unit uses a capacitor or resistor to couple the external interference signal to the sensitive node.

[0043] The formation of a node interference indication signal specifically refers to outputting a high-level or low-level alarm signal as a node interference indication signal.

[0044] As a further improvement of the present invention, the self-synchronization module includes a first inner loop oscillation component, a first outer loop oscillation component, a node coupling link, and an output interface to form a ring topology. Both the first inner ring oscillation component and the first outer ring oscillation component are formed by cascading at least two oscillation nodes; The node coupling link is located between the first inner loop oscillation component and the first outer loop oscillation component. It can inject the output signal of the first inner loop oscillation component with a frequency of the first initial oscillation frequency into the first outer loop oscillation component with a preset coupling coefficient, and inject the output signal of the first outer loop oscillation component with a frequency of the second initial oscillation frequency into the first inner loop oscillation component with a preset coupling coefficient, so that the oscillation frequencies of the first inner loop oscillation component and the first outer loop oscillation component tend to be consistent and the phase difference is locked. The oscillation node is equipped with a frequency configuration network, which is used to determine the first initial oscillation frequency and the second initial oscillation frequency. The output interface is used to output a reference voltage.

[0045] The oscillation node can be selected from any one of the following: 555 timer, operational amplifier, MEMS resonator, ceramic resonator, or quartz resonator. Replacing the oscillation device without changing the dual-ring coupled core topology is within the protection scope of this invention.

[0046] The ring topology of the bottom coupled sensing layer is consistent with the self-synchronization module of the bottom unit cell, intermediate unit cell, and convergent unit cell. Using three bottom unit cells as the second inner ring oscillation component and another three bottom unit cells as the second outer ring oscillation component, the coupling structure principle is consistent with the self-synchronization module, thereby realizing single-layer global self-synchronization of the bottom coupled sensing layer.

[0047] For details, please refer to Figure 1 The bottom layer is the low-level coupling perception layer, the middle layer is the intermediate dimension classification layer, and the top layer is the top-level feature convergence layer. The yellow links represent the first inner-loop or first outer-loop oscillation component formed by the coupling and cascading of three oscillation nodes. The blue links represent the coupling between the first inner-loop and first outer-loop oscillation components. The green lines in the low-level coupling perception layer represent the second inner-loop or second outer-loop oscillation component formed by the coupling and cascading of three low-level cells. The red lines in the low-level coupling perception layer represent the coupling between the second inner-loop and second outer-loop oscillation components. The middle red line represents the coupling between the intermediate dimension classification layer and the low-level coupling perception layer. Specifically, node 1 of the first low-level cell in the low-level coupling perception layer is coupled to node 1 of the first intermediate cell; node 2 of the same low-level cell is coupled to node 1 of the second intermediate cell; node 3 is coupled to node 1 of the third intermediate cell, and so on, until finally node 6 of the first low-level cell is coupled to node 1 of the sixth intermediate cell. The node 1 of the second bottom-level unit cell is coupled to the node 2 of the first intermediate unit cell. In other words, the six oscillation nodes of the first intermediate unit cell are connected to the six oscillation nodes 1 of the bottom-level unit cells. The green line above represents the coupling between the intermediate-dimensional classification layer and the top-level feature convergence layer. Specifically, the nodes 1 of the intermediate unit cells are coupled sequentially.

[0048] Furthermore, it should be noted that, referring to Figure 1 The bottom cell, intermediate cell and convergent cell are divided into nodes 1 to 6. The signal rule of the bottom coupling sensing layer of this invention is: independent output buses are led out according to the oscillation node number of each bottom cell. All bottom cells with the same number node buses are summarized to obtain a global dimension bus, forming six global dimension buses, which is consistent with the specific connection method mentioned above.

[0049] The connection rules for the intermediate dimension classification layer are as follows: Global Dimension Bus Node 1 → Intermediate Cell 1 signal input, Global Dimension Bus Node 2 → Intermediate Cell 2 signal input… Global Dimension Bus Node 6 → Intermediate Cell 6 signal input, a one-to-one physical direct connection with no crosstalk or parallel interference. Core function: Only completes the global aggregation of signals in the same dimension; single-channel amplitude perturbations only affect the cell of that channel and will not interfere with the other five channels, nor will they negatively affect the underlying timing frequency. (See reference here.) Figure 2The cyan and purple lines represent bus lines, while the red and dark blue lines represent intra-layer coupling. Since the intermediate unit cells are independent, the red and dark blue lines are dashed lines, and finally they are coupled to the convergent unit cell.

[0050] The convergence cell in the top-level feature convergence layer receives all the output signals from the six intermediate cells. Its core function is to achieve six-dimensional feature weighted fusion through steady-state evolution of the hardware circuit, natively implement a physical attention mechanism, automatically enhance highly perturbable features, suppress static background noise, and output a globally standardized feature package.

[0051] In addition, it should be noted that signal transmission between layers uses a unidirectional bus.

[0052] This section briefly explains the working principle of the self-synchronization module. For more detailed information, please refer to sub-application 2026110201654, "A Self-Synchronization Module and Phase Control Method Based on Dual-Ring Coupled Oscillation": (a) Frequency convergence mechanism to achieve self-stabilization of frequency.

[0053] Let the first initial oscillation frequency of the first inner loop oscillation component be... The second initial oscillation frequency of the first outer ring oscillation component is Both are set by the frequency configuration network, and can exist in the initial state. .

[0054] The cross-coupling network will convert the output signal of the first inner loop oscillation component into a cross-coupled network. The coupling coefficient k is injected into the first outer loop oscillation component, and the output signal of the first outer loop oscillation component is simultaneously... The same coupling coefficient k is injected into the first inner-loop oscillating component, forming a bidirectional symmetrical constraint. Under the coupling effect, the actual frequency of the first inner-loop oscillating component is pulled towards the first outer-loop oscillating component. The first outer ring oscillation component is simultaneously pulled towards the first inner ring oscillation component by the offset. Offset. The system satisfies the following condition when it reaches steady state: = = , in The common locked oscillation frequency after convergence is between the original... and The convergence process is entirely automated by the cross-coupled network, requiring no external reference or digital calibration.

[0055] (II) Phase Difference Locking Mechanism After the frequencies of the first inner-loop oscillation component and the first outer-loop oscillation component converge to the same value, the cross-coupling network applies a rigid constraint to the instantaneous phase. Let the phase of the first inner-loop oscillation component be... The phase of the first outer ring oscillation component is The phase constraint established by the cross-coupled network is: , in It is determined by the coupling coefficient, injection strength, and nonlinearity of the oscillation node. Under symmetrical coupling, It is a constant value and does not drift with time, temperature, or voltage. When it is necessary to adjust the phase difference, it can be approximated by gradually adjusting the coupling coefficient.

[0056] The essence of this phase-locked loop (PLL) is that a closed-loop constraint path is formed by two loops, and any phase deviation will be automatically corrected, eventually converging to a unique stable phase difference. It does not rely on a phase detector or loop filter and belongs to a pure topology-constrained PLL.

[0057] (III) Drift Self-Inhibition Mechanism Temperature and voltage fluctuations will synchronously change the time constants of the first inner loop oscillation component and the first outer loop oscillation component, causing... and Drifting in the same direction.

[0058] In a single-ring oscillator, drift is directly manifested as frequency shift; In this module's dual-ring coupling structure, drift is canceled out by differential coupling: phase constraint. Almost unaffected by co-directional drift; locked frequency The sensitivity to single-loop drift is significantly reduced. In engineering, the long-term frequency stability of dual-loop coupling is more than an order of magnitude better than that of single-loop coupling, and the stability depends on the coupling coefficient k.

[0059] As a further improvement of the present invention, the timing reference unit also includes a digital configuration device, which includes a digital configuration interface, adjustable resistor-capacitor components, and a digital potentiometer. The adjustable resistor-capacitor components and the digital potentiometer are electrically connected to the digital configuration interface. The frequency configuration network is composed of fixed resistor-capacitor components and adjustable resistor-capacitor components. The adjustable resistor-capacitor components are programmable capacitors or digitally adjustable resistors. The coupling device of the node coupling link adopts a digital potentiometer. The digital configuration device can receive external digital configuration commands and configure the coupling coefficient, the first initial oscillation frequency, the second initial oscillation frequency, and the phase difference according to the external digital configuration commands.

[0060] This invention retains programmable timing capabilities, supporting online frequency and phase modulation, and parameter reconfiguration for individual devices, partitions, and entire arrays. Parameter updates do not interrupt sensing and synchronization services. It is suitable for large-scale distributed IoT and industrial monitoring networks.

[0061] It should be noted that when there are no different parameter requirements, the system can run according to the factory-set parameters.

[0062] During normal phase-locked output, the coupling coefficient can be configured. When the phase difference needs to be adjusted, it can be gradually adjusted by adjusting the coupling coefficient to approximate the desired result. The system automatically re-enters the coupling traction and dynamic phase-locked process, updates the working parameters, and maintains uninterrupted output and timing synchronization throughout the entire process. It has strong operational continuity and is suitable for 24 / 7 uninterrupted monitoring scenarios in industrial settings.

[0063] As a further improvement of the present invention, it also includes a multi-layer coupling network, which includes node coupling links disposed in the bottom cell, intermediate cell and convergent cell. Multilayer coupling networks also include intralayer coupling links between different bottom-layer cells within the bottom coupling sensing layer; The multilayer coupled network also includes cross-layer coupling links between the bottom coupled perception layer and the intermediate dimension classification layer, and between the intermediate dimension classification layer and the top feature convergence layer.

[0064] The routing between layers and between units is the shortest theoretical distance, minimizing parasitic capacitance, signal delay, and crosstalk; the hexagonal tiling adapts to all layout forms, including planar, circular, arc, and irregular curved surfaces, with no invalid redundant units.

[0065] It should be noted that the signal transmission between layers in this invention is unidirectional, and the upper layer signal cannot control the timing reference of the lower layer in reverse. Small-scale prototypes do not require additional isolation circuits, and large-scale arrays can be further strengthened with only simple unidirectional protection, resulting in low engineering modification costs.

[0066] As a further improvement of the present invention, both the first inner ring oscillation component and the first outer ring oscillation component are formed by cascading three oscillation nodes; The bottom-level coupled sensing layer is constructed from six bottom-level unit cells, and the intermediate-dimensional classification layer is constructed from six intermediate unit cells, so as to form six output signals representing six dimensions.

[0067] It should be noted that even numbers such as 4, 6, and 8 can also be used to construct a double-ring coupled topology. As long as the self-similar double-ring restraint logic of the oscillation nodes inside a single cell is retained, it is an equivalent underlying array structure of this invention. However, if a simplified topology of 4 / 8 cells is used, it can only output 4 sets of phase references, supports only rectangular array arrangements, cannot build hexagonal curved surface arrays, cannot adapt to the six-dimensional phase encoding system of the parent application "A Phase Encoding Method and System Based on Cyclic Phase Transition Topology", loses high-order core functions such as six-way independent dimensional isolation and hierarchical precise traceability, and is only suitable for low-cost single-point simple monitoring devices. It does not belong to the standard implementation architecture for vision, brain-like computing, and carbon-silicon co-fiber computing power. Nor does it belong to the standard commercial solution for robots, panoramic vision, and six-dimensional phase encoding. The six-cell dual topology adopted in this invention is the optimal technical solution to achieve all functions.

[0068] Of course, there can be more than two oscillation nodes, and it is not limited to three.

[0069] The intermediate-dimensional classification layer is constructed from six independent, uncoupled intermediate cells. It classifies and aggregates signals across the entire domain based on node dimensions, enabling the construction of two-dimensional coordinates combining physical and spatial addresses for precise interference localization. For easier understanding, please refer to... Figure 1 The connection architecture diagram shows six nodes forming a unit cell. At the bottom is the underlying coupled sensing layer. Corresponding to the six bottom unit cells, there are six groups of nodes, each group consisting of nodes 1 through 6 arranged sequentially. The underlying coupled sensing layer is coupled to the intermediate dimensional classification layer via cross-layer coupling links, which is node-dimensional coupling. See the diagram for details. Figure 1 As can be seen from the connection in the middle, node 1 of the first bottom cell of the bottom coupling sensing layer is coupled to node 1 of the first intermediate cell, node 2 of the same bottom cell is coupled to node 1 of the second intermediate cell, node 3 is coupled to node 1 of the third intermediate cell, and so on, until finally node 6 of the first bottom cell is coupled to node 1 of the sixth intermediate cell.

[0070] As a further improvement of the present invention, the bottom coupling perception layer, the intermediate dimension classification layer and the top feature convergence layer are all multiple, and each intermediate unit cell has a fixed dimensional interface. Based on the preset numbering rules, each intermediate cell under the same intermediate dimension classification layer is assigned a unique dimension number corresponding to the dimension interface, and adjacent intermediate dimension classification layers are connected through the dimension interface signal corresponding to the same unique dimension number.

[0071] As a further improvement of the present invention, an array node can be formed by a bottom coupling perception layer, an intermediate dimension classification layer and a top feature convergence layer, and adjacent array nodes follow the same dimension point-to-point docking rule.

[0072] It's important to note that points with the same unique dimension number are points in the same dimension. The default numbering rule is to use a fixed node numbering rule, either clockwise or counterclockwise. Based on this fixed numbering, a polar coordinate-like layered addressing system is constructed, enabling array units to simultaneously possess two-dimensional positioning capabilities in both spatial and dimensional coordinates. Of course, this also means that corresponding numbering must be done according to the dimension, such as... Figure 5 The numbers 1, 2, 3, 4, 5, and 6 shown correspond to the six internal dimensional channels, respectively.

[0073] Furthermore, an externally interconnected array can be formed through a bottom-level coupling sensing layer, creating a graphene-like array. A six-channel bus completes the cellular network, resulting in stronger synchronization constraints across all cells. This allows for batch integration with front-end acquisition hardware such as CMOS image sensors and infrared arrays, significantly improving the timing synchronization performance of multi-channel images and sensor signals. This is suitable for panoramic vision and large-area distributed monitoring scenarios. The invention can also be extended to construct flexible biomimetic sensing arrays, applicable to cutting-edge applications such as embodied intelligent full-body tactile sensing and synchronous thermal sensing.

[0074] Alternatively, external interconnection can be achieved through a top-level feature convergence layer, with the output channel of the top-level feature convergence layer completing the cellular network. Implementing cellular networking through an intermediate dimension classification layer is merely a preferred embodiment of this invention.

[0075] At this time, refer to Figure 5 This is a schematic diagram of a graphene-like array arrangement, representing a specific embodiment of the present invention. Each array node represents a three-layer structure: a bottom-layer coupling sensing layer, a middle-layer classification layer, and a top-layer feature convergence layer. A hexagonal honeycomb topology is adopted, with array nodes following a point-to-point docking rule within the same dimension, achieving smooth expansion of the entire array and time-series self-locking. This point-to-point docking rule includes not only point-to-point docking within the same dimension but also no intersection, no redundancy, and no reuse. Regardless of array size or layout (planar, arc-shaped, irregular curved surface, hexagonal, or ring-shaped honeycomb), all array units adopt the aforementioned point-to-point docking rule. Specifically, signal 1 of this unit connects to neighboring unit 1, signal 2 of this unit connects to neighboring unit 2, signal 3 of this unit connects to neighboring unit 3, signal 4 of this unit connects to neighboring unit 4, signal 5 of this unit connects to neighboring unit 5, and signal 6 of this unit connects to neighboring unit 6. Each of the six dimensional channels forms a globally continuous coupling network, and the entire array shares a set of endogenous timing references, completely eliminating the need for an external global clock.

[0076] This invention also proposes a distributed array synchronization method based on multi-layer dual-ring coupling, applied to the above-mentioned distributed array synchronization system based on multi-layer dual-ring coupling, comprising the following steps: S1. Start the self-synchronization module in the bottom cell, intermediate cell and convergent cell to achieve autonomous oscillation and enter the initial steady state; S2. Initiate node coupling links within the bottom cell, intermediate cell, and convergent cell to achieve self-stabilization and self-locking of the bottom cell, intermediate cell, and convergent cell; S3. Start the intra-layer coupling link of the bottom coupling perception layer to achieve single-layer global self-synchronization within the bottom coupling perception layer; S4. Activate the cross-layer coupling link between the bottom coupling perception layer and the intermediate dimension classification layer so that the output signal of the bottom coupling perception layer can be transmitted to the corresponding intermediate dimension classification layer. S5. Activate the cross-layer coupling link between the intermediate dimension classification layer and the top-level feature convergence layer so that the output signal of the intermediate dimension classification layer is transmitted to the corresponding top-level feature convergence layer, and the top-level feature convergence layer outputs a global fusion feature signal.

[0077] Each underlying unit cell acquires front-end sensor signals in real time and completes synchronous sampling based on a unified timing sequence. Local interference only appears in the corresponding dimension and corresponding spatial unit, and the fault location and dimension can be quickly located by relying on two-dimensional coordinates.

[0078] The first specific embodiment of the present invention is the three-layer structure basic scheme of the present invention, which is suitable for the following scenarios: small single-point sensing, ordinary planar image acquisition, and small equipment status monitoring; the perforated plate prototype and the standard coupling plate mass production can be directly implemented.

[0079] The second specific embodiment introduces a digital configuration device into the basic solution of this invention, enabling independent configuration of a single node, batch configuration of partitions, and global configuration of the entire array. The following parameters are configured: first initial oscillation frequency, second initial oscillation frequency, phase difference, steady-state threshold, and interference detection sensitivity. After issuing external digital configuration commands, it can automatically achieve secondary phase locking, ensuring uninterrupted acquisition, synchronization, and feature output throughout the entire process. Suitable scenarios include: distributed IoT sensing in large-scale sites, multi-zone industrial equipment cluster monitoring, and large-area visual arrays; it supports remote operation and maintenance and multi-condition switching.

[0080] The third specific embodiment is a panoramic multi-view sensor array, configured with four independent three-layer standard array nodes (each array node includes the three-layer structure of this invention), each connecting to four fisheye sensors. Each of the four array nodes completes intra-layer synchronization, dimensional classification, and feature convergence. The groups are interconnected according to cellular same-dimensional docking rules, ensuring unified temporal sequence across the entire array. It outputs four standardized six-dimensional feature packages, requiring only geometric image stitching at the backend, without the need for temporal alignment or feature compensation algorithms. The core effect is that the panoramic live streaming experience completely matches that of planar live streaming, with no stitching flaws, no stuttering, and no additional computing power overhead.

[0081] The fourth specific embodiment integrates the first embodiment of the present invention into a single-chip analog ASIC; the chip externally only exposes six standardized dimensional interfaces, power supply pins, and a digital configuration interface; there are no external discrete coupling resistors. Suitable scenarios: lightweight sensing front-end for consumer-grade XR and small humanoid robot heads, enabling mass production with standardized specifications and reduced hardware size.

[0082] In the fifth specific embodiment, two specific embodiments of the present invention are vertically stacked and connected, forming an upper array node and a lower array node (each array node includes the three-layer structure of the present invention). The lower array node collects the original sensing signals, completes the six-dimensional spatial aggregation, and after fusion, transmits them upward to the upper array node. The upper array node then completes cross-regional feature classification and secondary convergence again. The double-layer stacking automatically completes two-level feature distillation, and the waveform layering phenomenon is clearly observable. It can adapt to multimodal composite sensing (vibration + temperature + vision) and bionic robot multi-layer tactile cortex hardware prototypes.

[0083] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. Those skilled in the art can make other variations or modifications based on the above description. It is neither necessary nor possible to exhaustively describe all embodiments here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A distributed synchronization system for an array based on multi-layer dual-ring coupling, characterized in that, It includes multiple bottom cells and intermediate cells capable of receiving signals, outputting signals, and self-synchronizing, and also includes at least one convergent cell capable of receiving signals, outputting signals, and self-synchronizing, wherein the bottom cells, intermediate cells, and convergent cells all have the same ring topology. Based on the ring topology, multiple bottom cells are constructed to form at least one bottom coupling sensing layer. The bottom cells within the same bottom coupling sensing layer have a weak coupling relationship to form a single-layer global self-synchronization. At least one intermediate dimension classification layer is formed by constructing multiple intermediate unit cells. The intermediate dimension classification layer is used to receive the output signal of the bottom coupled sensing layer that corresponds to it. The intermediate unit cells in the same intermediate dimension classification layer are independent of each other. At least one top-level feature convergence layer is formed by constructing at least one convergence cell. The top-level feature convergence layer is used to receive the output signal of the intermediate dimension classification layer and output the global fusion feature signal.

2. The array-based distributed synchronization system based on multi-layer dual-ring coupling according to claim 1, characterized in that, The bottom cell, intermediate cell and convergent cell each include a timing reference unit, an interference injection unit and a comparison and judgment unit; The timing reference unit includes a self-synchronization module, which can output a reference voltage that is self-stabilized and self-locked. The self-synchronization module has a preset sensitive node. The input terminal of the interference injection unit is used to receive external interference signals so that the bottom cell and intermediate cell can receive signals. The output terminal of the interference injection unit is used to couple the external interference signals to the sensitive node so as to change the voltage of the sensitive node to form a disturbed voltage. The comparison and judgment unit includes a voltage comparator and a voltage divider potentiometer. The input terminal of the voltage divider potentiometer is electrically connected to the self-synchronization module, and the output terminal of the voltage divider potentiometer is electrically connected to the inverting terminal of the voltage comparator to provide an adjustable reference voltage to the voltage comparator. The non-inverting terminal of the voltage comparator is electrically connected to the sensitive node to compare the disturbed voltage of the sensitive node with the adjustable reference voltage to form a node interference indication signal.

3. The array-based distributed synchronization system based on multi-layer dual-ring coupling according to claim 2, characterized in that, The self-synchronization module includes a first inner loop oscillation component, a first outer loop oscillation component, a node coupling link, and an output interface to form the ring topology. Both the first inner ring oscillation component and the first outer ring oscillation component are formed by cascading at least two oscillation nodes; The node coupling link is located between the first inner loop oscillation component and the first outer loop oscillation component. It can inject the output signal of the first inner loop oscillation component with a frequency of the first initial oscillation frequency into the first outer loop oscillation component with a preset coupling coefficient, and inject the output signal of the first outer loop oscillation component with a frequency of the second initial oscillation frequency into the first inner loop oscillation component with a preset coupling coefficient, so that the oscillation frequencies of the first inner loop oscillation component and the first outer loop oscillation component tend to be consistent and the phase difference is locked. The oscillation node is equipped with a frequency configuration network, which is used to determine the first initial oscillation frequency and the second initial oscillation frequency. The output interface is used to output the reference voltage.

4. The array-based distributed synchronization system based on multi-layer dual-ring coupling according to claim 3, characterized in that, The timing reference unit further includes a digital configuration device, which includes a digital configuration interface, adjustable resistor-capacitor components, and a digital potentiometer. The adjustable resistor-capacitor components and the digital potentiometer are electrically connected to the digital configuration interface. The frequency configuration network consists of fixed resistor-capacitor components and the adjustable resistor-capacitor components. The adjustable resistor-capacitor components are programmable capacitors or digitally adjustable resistors. The coupling device of the node coupling link adopts the digital potentiometer. The digital configuration device can receive external digital configuration commands and configure the coupling coefficient, the first initial oscillation frequency, the second initial oscillation frequency, and the phase difference according to the external digital configuration commands.

5. The array-based distributed synchronization system based on multi-layer dual-ring coupling according to claim 1, characterized in that, It also includes a multi-layer coupling network, which includes node coupling links disposed in the bottom cell, intermediate cell and convergent cell; The multilayer coupling network also includes intralayer coupling links between different bottom-layer unit cells within the bottom-layer coupling sensing layer; The multi-layer coupled network also includes cross-layer coupling links between the bottom coupled perception layer and the intermediate dimension classification layer, and between the intermediate dimension classification layer and the top feature convergence layer.

6. The array-based distributed synchronization system based on multi-layer dual-ring coupling according to claim 2, characterized in that, Both the first inner ring oscillation component and the first outer ring oscillation component are formed by cascading three oscillation nodes; The bottom coupling sensing layer is constructed from six bottom unit cells, and the intermediate dimension classification layer is constructed from six intermediate unit cells to form six output signals representing six dimensions.

7. The array-based distributed synchronization system based on multi-layer dual-ring coupling according to claim 6, characterized in that, The bottom coupling perception layer, the intermediate dimension classification layer, and the top feature convergence layer are all multiple, and each intermediate unit cell has a fixed dimensional interface. Based on the preset numbering rules, each intermediate cell under the same intermediate dimension classification layer is assigned a unique dimension number corresponding to the dimension interface, and adjacent intermediate dimension classification layers are connected through the dimension interface signal corresponding to the same unique dimension number.

8. The array-based distributed synchronization system based on multi-layer dual-ring coupling according to claim 3, characterized in that, The output interface includes a working output interface and a phase difference output interface. The reference voltage is output through the working output interface, and the phase difference output interface is used to output multiple standardized timing phase signals.

9. The array-based distributed synchronization system based on multi-layer dual-ring coupling according to claim 1, characterized in that, An array node can be formed by a bottom-level coupled perception layer, an intermediate-dimensional classification layer, and a top-level feature convergence layer. The connection between adjacent array nodes follows the same-dimensional point-to-point docking rule.

10. A distributed synchronization method for an array based on multi-layer dual-ring coupling, characterized in that, An array-based distributed synchronization system based on multi-layer dual-ring coupling, as described in any one of claims 1 to 9, comprises the following steps: S1. Start the self-synchronization module in the bottom cell, intermediate cell and convergent cell to achieve autonomous oscillation and enter the initial steady state; S2. Activate the node coupling links within the bottom cell, intermediate cell, and convergent cell to achieve self-stabilization and self-locking of the bottom cell, intermediate cell, and convergent cell; S3. Start the intra-layer coupling link of the underlying coupled sensing layer to achieve single-layer global self-synchronization within the underlying coupled sensing layer; S4. Activate the cross-layer coupling link between the bottom coupling perception layer and the intermediate dimension classification layer, so that the output signal of the bottom coupling perception layer is transmitted to the corresponding intermediate dimension classification layer; S5. Activate the cross-layer coupling link between the intermediate dimension classification layer and the top-level feature convergence layer, so that the output signal of the intermediate dimension classification layer is transmitted to the corresponding top-level feature convergence layer, and the top-level feature convergence layer outputs a global fusion feature signal.