A DSF cold start two-dimensional synchronization method and device based on pre-cached backward matching

Abstract

The application discloses a DSF cold start two-dimensional synchronization method and device based on pre-cached backward matching. After the same DSF is pre-injected, both communication parties independently start PSSM, and unknown offset Delta n exists in logical state due to the fact that physical clocks are not synchronized. The transmitting end periodically sends a safe pilot; the receiving end pre-caches signals, performs offline traversal of candidate logical state values to perform sliding correlation matching, and simultaneously determines logical state offset Delta n and physical time offset Delta, according to which the local logical state and the receiving window are adjusted, and two-level initial synchronization of the logical layer and the physical layer is completed at one time. For Delta n>0, direct alignment is adopted; for Delta n<0, waiting alignment is adopted. The application adopts a hash chain to drive forward / backward safety, is an initialization protocol of a DSF deterministic communication platform, and constitutes a complete closed loop with the existing patent.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, specifically relating to a method and apparatus for achieving initial synchronization between the logical and physical layers of communicating parties based on Dynamic Security Foundation (DSF) and Protocol Security State Machine (PSSM) without any prior synchronization information. This invention is an initialization protocol for a DSF / PSSM deterministic communication platform. The core features of this platform are zero signaling coordination, intrinsic security (forward / backward security), and deterministic latency guarantees. This invention is particularly suitable for infrastructure-free scenarios such as satellite communication, IoT device deployment, emergency communication, and high-speed aircraft data links. Background Technology

[0002] In fifth-generation mobile communication systems (5G) and earlier technologies, the initial synchronization between communicating parties was essentially only physical layer synchronization. Taking 5G as an example, the User Equipment (UE) obtains time slot and frame boundary synchronization with the base station by detecting the Primary Synchronization Signal (PSS) and Secondary Synchronization Signal (SSS), and reads the Master Information Block (MIB) through the Physical Broadcast Channel (PBCH), thereby establishing coarse synchronization with the base station. The core of this process is solving the physical timing alignment problem of "when to receive and when to transmit." Subsequently, the UE further refines synchronization through a random access procedure and establishes a Radio Resource Control (RRC) connection.

[0003] However, 5G and its evolved systems completely lack the concept of "logical layer synchronization." In other words, the communicating parties do not have a logical state machine that operates independently of the physical clock, nor do they have "logical state values"—synchronization states decoupled from the physical layer—used to generate communication parameters. All higher-level protocol states (such as RRC states and discontinuous reception DRX cycles) and parameter generation directly or indirectly depend on the synchronized physical layer timing (such as the system frame number (SFN) and time slot number). Once the physical layer loses synchronization, the entire protocol stack will "fly blind," requiring a complex signaling reconstruction process, during which determinism and security of communication cannot be guaranteed.

[0004] The applicant's prior series of patent applications (application numbers: 2026100015014, 2026100463580, 2026101229059, etc.) proposed a deterministic communication framework based on Dynamic Security Foundation (DSF) and Protocol Security State Machine (PSSM). This framework introduces for the first time the concepts of a logical timeline and a logical state machine decoupled from the physical clock, enabling secure collaboration between communicating parties under zero signaling conditions. Through strict synchronization of the time-varying state S(t) output by the PSSM, both communicating parties can independently generate consistent communication parameters, fundamentally eliminating the dependence on the continuous accuracy of the physical clock. It should be noted that the PSSM state value S(t) can have various specific implementations, including but not limited to: physical layer timing information (SFN / Slot) in rule A, the secret hash chain state T_State[n] in rule B, the public logical state (Epoch_ID, Counter) in rule C, and the logical counter n used in this application for simplification. Among them, rule B (hash chain driven rule) provides the most core security attributes of this invention—forward security and backward security.

[0005] Building upon this foundation, the applicant further proposed a two-level synchronization maintenance method (application number: 2026105584990) and a logic-level autonomous synchronization and parameter adaptation method (application number: 2026105931074), solving the problems of synchronization maintenance and short-term drift compensation in a synchronized state. Simultaneously, the applicant also proposed a rapid synchronization method after disconnection (application number: 2026106011905), addressing the problem of rapid recovery after a short-term disconnection.

[0006] However, the aforementioned patents all assume that the communicating parties have established initial synchronization through some means (such as RRC signaling or NTP synchronization), that is, they assume that the logical states of both parties are aligned (or have only minor deviations), and the physical layer is in a synchronized state. The aforementioned patents are not applicable to the following scenarios: First, the cold start scenario: Both parties pre-inject the same DSF through an offline secure channel, but each starts the PSSM independently. Because the physical clocks of both parties are not synchronized (no GPS, no NTP, no base station broadcast), the start time of the PSSM has an unknown and potentially large time offset, resulting in a fixed and potentially large difference Δ between the logical states of both parties. In extreme cases, the number of logic ticks corresponding to Δ can reach tens of thousands or even more. Existing patented synchronization maintenance and short-term recovery mechanisms assume Δ≈0 or only require ±1 to ±3 steps of silent calibration, which cannot handle such a large initial offset.

[0007] Second, in the scenario of prolonged disconnection: if the two parties are disconnected for too long (hours or even days), the crystal oscillator drift has rendered the previously stored synchronization information completely invalid, and the search range of Δ far exceeds the small window (K=1~100) defined in existing patents. At this point, the two parties have actually reverted to a cold start state of "never being synchronized".

[0008] Third, in scenarios without external time synchronization: the two parties cannot rely on GPS, NTP, or base station broadcasts for physical layer synchronization, and can only maintain time by relying on their respective independent local crystal oscillators. 5G and existing technologies are completely ineffective in such scenarios.

[0009] In the above scenario, the two communicating parties are in a state of "never being synchronized," requiring the establishment of a two-level synchronization system at both the logical and physical layers from scratch. Neither existing technologies nor the applicant's existing patents have solved this problem.

[0010] The essential difference from existing technologies It is important to note that while existing drone self-organizing network technologies (such as Mesh networks, AODV / OLSR routing protocols, DJI OcuSync, etc.) also achieve communication without relying on base stations and GPS, they are essentially still within the scope of physical layer networking and data transmission, which is fundamentally different from this invention. First, existing ad hoc network technologies still heavily rely on physical layer synchronization. Whether it's a TDD system based on GPS timing or timestamp alignment achieved through ranging, its routing protocols and higher-layer coordination are based on the physical clock. Once physical layer synchronization is disrupted or cannot be established, the entire network will be unable to coordinate transmit and receive time slots.

[0011] Second, the "logical synchronization" in existing technologies is actually a high-level protocol state machine (such as routing table maintenance and handshake signaling). These state machines are tightly coupled to the physical layer synchronization and lack universality. Each application needs to design its own state machine independently, making it impossible to form a unified logical benchmark.

[0012] Third, the PSSM proposed in this invention is an independent logical layer inserted between the physical layer and higher-layer protocols, decoupled from the physical clock. This logical layer provides a unified "logical clock" benchmark for all higher-layer services, making the operation of higher-layer protocols completely independent of the physical clock. This "two-dimensional synchronous" architecture (the logical layer evolves independently, and the physical layer only acts as an actuator) is not found in any existing ad hoc networking technology.

[0013] Fourth, this invention employs a hash chain-driven PSSM (Rule B), where the pilot sequence and encryption key for each logical cycle are unidirectionally derived from the previous state. Even if the key for a certain cycle is cracked, it is impossible to deduce subsequent cycles (forward security); even if the device is physically captured, historical communications cannot be decrypted (backward security). This is a cryptographic-level security property that existing fixed or periodic pilot schemes cannot provide.

[0014] In summary, while existing ad hoc network technologies can achieve infrastructure-free communication, they cannot simultaneously meet the requirements of "zero signaling, intrinsic security, and determinism." The DSF / PSSM framework is a unified communication platform designed to address this "impossible triangle." Existing infrastructure-free communication solutions face mutual constraints among these three dimensions: zero signaling requires no interaction, but traditional security mechanisms rely on interaction; intrinsic security requires unpredictable states, but traditional synchronization relies on fixed signals; determinism requires bounded delays, but traditional protocols suffer from random backoff. Summary of the Invention

[0015] Key concept definition To facilitate understanding of this invention, the following core concepts will first be clarified: Dynamic Security Foundation (DSF): A cryptographically secure triple shared by both communicating parties, consisting of a security key K_sec, an initial anchor Init_Anchor, and a state transition rule Rule_ID. The DSF is the initialization parameter of the Protocol Security State Machine (PSSM), determining the starting point and evolution rules of the logical timeline.

[0016] Protocol Security State Machine (PSSM): A deterministic state machine based on DSF, evolving independently between the communicating parties and outputting a time-varying logic state S(t). The evolution of the PSSM is decoupled from the physical clock and driven only by the logical tick defined by Rule_ID. The PSSM state value S(t) can have various specific implementations, including but not limited to: physical layer timing information (SFN / Slot) in rule A, secret hash chain state T_State[n] in rule B, public logic state (Epoch_ID, Counter) in rule C, and the logic counter n used in this application for simplification.

[0017] Logical decision moment: The discrete time point at which the PSSM outputs the logical state S(t) according to a predetermined logical cycle. Each logical decision moment can be associated with a logical state value. In this invention, all communication events use the logical decision moment as a unified time index.

[0018] Logical timeline: A discrete sequence of logical decision moments arranged in chronological order, decoupled from the local physical clock and evolving independently.

[0019] Logical state offset (Δn): The difference between the logical state values ​​of the transmitter and receiver PSSM, Δn = n_tx - n_rx, where n_tx is the current logical state value of the transmitter and n_rx is the current logical state value of the receiver. Since there is an unknown time offset between the start-up times of the PSSMs of both parties, Δn is a fixed unknown value that needs to be determined by the method of this invention.

[0020] Physical time deviation (Δ): The fixed time deviation between the actual arrival time of the signal measured at the receiver and the theoretical expected time, Δ = T_actual - T_expected(n_match). This deviation includes propagation delay, processing delay, and physical time offset caused by the difference in the start times of the PSSM between the two parties.

[0021] Two-dimensional synchronization refers to a two-level synchronization process that simultaneously determines the logical state offset Δn and the physical time deviation Δ. Logic layer synchronization addresses the question of "which beat is it now," while physical layer synchronization addresses the question of "when did the signal arrive?"

[0022] Cold start: refers to the process by which two communicating parties establish a two-level synchronization for the first time without any prior synchronization information. Its characteristics include: the offset Δn between the logical states of the two parties can be very large (tens of thousands or even more), and there are no historical records available for reference.

[0023] Purpose of the invention The purpose of this invention is to provide a method and apparatus for achieving initial synchronization between the logical and physical layers of communicating parties through pre-caching and backward matching without any prior synchronization information, filling the gap of "cold start" in the DSF framework, and together with existing patents, forming a complete closed loop from initial synchronization to long-term maintenance and then to recovery from disconnection.

[0024] Technical solution To achieve the above objectives, this invention provides a two-dimensional synchronization method for DSF cold start based on pre-cached backward matching.

[0025] Figure 1 This is a flowchart illustrating the overall process of cold start two-dimensional synchronization in this invention. Figure 1 As shown, the method includes the following core steps: Step 1: DSF pre-injection Both communicating parties inject the same Dynamic Security Foundation (DSF) once via an offline secure channel (such as a wired connection, physical contact, or pre-configuration). The DSF contains at least a security key K_sec, an initial anchor (Init_Anchor), and a state transition rule (Rule_ID). Each party stores its DSF independently, requiring no interaction before communication begins. The initial anchor (Init_Anchor) defines the starting point of the logical timeline, which can be an absolute physical moment or a relative moment (e.g., "the 1000th local clock tick after receiving configuration").

[0026] Step 2: The transmitter autonomously starts the PSSM After the local clock reaches the start time specified by Init_Anchor, the first communicating party (referred to as the transmitter) independently starts the protocol security state machine PSSM. Starting from the initial state, it independently evolves the logical state S(n) according to the logical beat T_step defined by Rule_ID. The evolution of PSSM is completely decoupled from the physical clock and is driven only by the local timer.

[0027] When rule B (hash chain driven rule) is used, the logical state is a secret hash chain value, and its evolution follows: S(0) = H(K_sec || Init_Anchor || "INIT") S(n) = H(K_sec || S(n-1) || "CHAIN"), n ≥ 1 Where H is a collision-resistant hash function (such as SHA-256). This mechanism ensures that the state of each logical cycle is irreversible, providing forward and backward security. It should be noted that the pilot signal in this scheme is only used for synchronization and does not carry a payload. Even if a quantum computer can crack the hash function in the future, it can only know the synchronization time and cannot decrypt historical communication data, thus ensuring backward security.

[0028] At each logical decision point, the transmitter sends a safety pilot sequence derived from the current logical state S(n): P_tx(n) = PRF(K_sec, S(n) || "PILOT") Where PRF stands for pseudo-random cryptographic function (such as HMAC-SHA256), and "PILOT" is a context label that distinguishes the purpose of the pilot. This pilot sequence is time-varying, unpredictable, and resistant to replay attacks.

[0029] The transmitter continuously sends pilot signals for a duration T_tx ≥ N × T_step, where N is the preset number of transmission cycles (typically 1000~100000).

[0030] Step 3: Receiver pre-buffering The second communicating party (referred to as the receiving end) also independently starts its local PSSM after the local clock reaches the start time specified by Init_Anchor, starting from the same initial state and evolving independently at the same T_step.

[0031] Because the physical clocks of both parties are not synchronized, there is an unknown time offset Δt_start between the start time of the receiver's PSSM and that of the transmitter, resulting in a fixed unknown difference Δn = Δt_start / T_step between the logical states of both parties.

[0032] The receiver opens its receiving window near the estimated start time T_start_est, continuously capturing signals and storing them in a buffer. The capture duration T_capture ≥ 2 × T_step × N_search, where N_search is a preset search cycle number, calculated by dividing the estimated maximum initial time offset by T_step, rounding up, and adding a safety margin (typically 10~100). Typical values ​​for N_search are 1000~100000.

[0033] The receiver does not perform real-time matching during the acquisition phase, but only performs signal storage to avoid real-time processing pressure and ensure the integrity of the acquisition.

[0034] Step 4: Offline sliding match (backward match) After acquisition, the receiver generates a local candidate pilot sequence set: P_local(n) = PRF(K_sec, n || "PILOT"), n = 0, 1, 2, ..., N_max Where N_max = T_capture / T_step.

[0035] For each candidate state value, the receiver calculates the theoretical arrival time: T_expected(n) = T_start_est + n×T_step Within a time window of T_expected(n) ± ε (ε is a preset time search margin, typically T_step / 10 to T_step / 20), the local pilot and the buffer signal are subjected to sliding correlation operation, and the maximum correlation value C(n) is recorded.

[0036] Figure 2 This is a timing diagram illustrating pre-caching and backward matching. For example... Figure 2As shown, the transmitter sends the pilot sequence periodically according to the logical beat. The receiver opens the acquisition window near the estimated start time and stores the signal in the buffer. After acquisition, the receiver traverses the candidate logic state values ​​n, generates the corresponding local candidate pilot, and performs sliding matching near the theoretical arrival time to find the relevant peak value, thereby determining the actual logic state of the transmitter.

[0037] Step 5: Determine the matching results and calculate the double bias. The receiver determines n_match = argmax C(n). If C(n_match) exceeds the preset threshold, the match is successful, and the corresponding actual arrival time T_actual is recorded, where T_actual = capture window start time + offset / sampling rate.

[0038] At this point, the receiver calculates two key deviations: (1) Logical state offset Δn: Δn = n_match - n_rx, where n_rx is the local logical state value of the receiver when a match is made. This offset is used for logical layer synchronization.

[0039] (2) Physical time deviation Δ: Δ = T_actual - T_expected(n_match), where T_expected(n_match) = T_start_est + n_match × T_step. This deviation comprehensively reflects the physical time offset caused by the propagation delay, processing delay, and the difference in the start time of the PSSM between the two parties, and is used for physical layer synchronization.

[0040] It should be noted that the logical state offset Δn and the physical time deviation Δ satisfy the following relationship: Δ = Δn×T_step + δ Where δ represents the residual deviations such as propagation delay and processing delay. Since Δn is an offset of an integer number of beats, and Δ includes fine time deviations at the sub-beat level, the two complement each other to form complete two-dimensional synchronization information.

[0041] Step 6: Establish two-level synchronous operation (execute in real time) After completing the offline matching in step 5 and successfully calculating the logical state offset Δn and physical time deviation Δ, the present invention immediately enters the real-time synchronous execution phase. This phase consists of two independent but coordinated parts.

[0042] Figure 3 This is a diagram illustrating a logical state catch-up / wait mechanism. (Example:) Figure 3 As shown, the logic layer synchronization controller at the receiving end performs the following real-time operations based on Δn calculated in step 5: - If Δn > 0: The receiver's local logical state lags behind the transmitter. The controller immediately sets the local PSSM's logical state directly to align with the transmitter, skipping the missing |Δn| logical ticks. For state machines with forward safety (such as rule B driven by hash chains), this jump operation does not reveal information about the skipped states.

[0043] - If Δn < 0: The receiver's local logic state is ahead of the transmitter's. The receiver cannot "go back," and the controller will control the local PSSM to enter a waiting state, pausing the evolution of the logic state and waiting for |Δn| logic ticks. For extreme scenarios where Δn < 0, the receiver can choose to enter deep sleep, setting the wake-up time to the end of the waiting period to reduce power consumption. After the waiting time ends, the transmitter's logic state will naturally evolve to align with the receiver's current position.

[0044] - If Δn = 0: The logic layers of both sides are naturally synchronized and no adjustment is needed.

[0045] Simultaneously, the physical layer controller at the receiving end immediately adjusts the position of the subsequent physical layer receiving window based on the physical time deviation Δ calculated in step 5. At each subsequent logical decision time n, the physical layer receiving window at the receiving end will be precisely set to: T_window(n) = T_expected(n) + Δ ± ε Where T_expected(n) = T_start_est + n×T_step, and ε is the preset search margin. This window position ensures that the receiver can accurately capture all signals subsequently transmitted by the transmitter.

[0046] It should be noted that the physical time deviation Δ and the logical state offset Δn are two independent and complementary deviations: Δn solves the "logical round alignment" problem, while Δ solves the "physical time alignment" problem. Together, they constitute a complete two-dimensional synchronization.

[0047] Step 7: Communication and Mobility Management After Synchronization After the initial two-level synchronization is established, both communicating parties have the following basic requirements: - Logic layer: The PSSM logic states of both parties are aligned. - Physical layer: The receiver has obtained a fixed time deviation Δ, which can accurately predict the arrival time of the other party's signal. Based on this, both parties can enable the DSF framework for normal deterministic communication, including zero signaling parameter generation, bidirectional synchronization and mobility tracking, integration with existing patents, and closed-loop self-optimization.

[0048] Figure 4 This diagram illustrates the relationship between the present invention and the existing patent system. For example...Figure 4 As shown, this invention fills the gap in the "cold start" concept within the DSF framework, and together with existing patents (synchronization maintenance, disconnection recovery), forms a complete technical loop from initial synchronization to long-term maintenance, and then to disconnection recovery.

[0049] Beneficial effects Compared with existing technologies and the applicant's existing patents, the present invention has the following beneficial effects: First, zero-signaling coordination. Both communicating parties independently evolve their PSSM based on pre-injected DSFs, with all communication parameters derived from the current logical state, requiring no real-time signaling interaction. In environments with strong interference or signal interception, the survivability of this invention is significantly superior to existing solutions that require handshake negotiation.

[0050] Second, inherent security. This invention employs a hash-chain driven PSSM (Rule B), where the pilot sequence and encryption key for each logical cycle are unidirectionally derived from the previous state. Even if the key for a certain cycle is cracked, subsequent cycles cannot be deduced (forward security); even if the device is physically captured, historical communications cannot be decrypted (backward security). Furthermore, the pilots in this scheme are only used for synchronization and do not carry a payload. Even if a quantum computer can crack the hash function in the future, it can only know the synchronization time and cannot decrypt historical communication data, further ensuring backward security. Existing fixed or periodic pilot schemes do not possess this characteristic; their pilots are predictable, vulnerable to replay attacks, and lack forward security.

[0051] Third, deterministic communication is guaranteed. The logical tick of PSSM provides a deterministic time base for all communication events, and the end-to-end delay has a strict upper bound. Unlike existing schemes that rely on real-time scheduling, the deterministic delay of this invention can be mathematically proven.

[0052] Fourth, it fills the gap in cold start technology. For the first time, it solves the initial synchronization problem "from zero to synchronization" within the DSF framework. This invention, together with existing patents (synchronization maintenance, disconnection recovery), forms a complete closed loop from initial synchronization to long-term maintenance, and then to disconnection recovery.

[0053] Fifth, breaking the "impossible triangle." Existing infrastructure-free communication solutions face mutual constraints across three dimensions: zero signaling, intrinsic security, and determinism. Zero signaling requires no interaction, but traditional security mechanisms rely on interaction; intrinsic security requires unpredictable states, but traditional synchronization relies on fixed signals; determinism requires bounded delays, but traditional protocols suffer from random backoff. This invention, based on the DSF / PSSM framework, breaks through these three limitations simultaneously for the first time through pre-injected shared foundations and independently evolving state machines.

[0054] Sixth, two-dimensional synchronization capability. Simultaneously determine the logical state offset Δn and the physical time deviation Δ, and complete the initial synchronization of the logic layer and the physical layer in one go.

[0055] Seventh, no prior information is required. Synchronization is achieved through a purely passive listening method, without relying on historical records, external time synchronization, or network signaling.

[0056] Eighth, wide-range search capability. The search range for state offsets can reach tens of thousands or even more, enabling it to handle cold start scenarios where both parties have never synchronized before.

[0057] Ninth, zero transmitter modifications. The transmitter sends pilot signals according to the normal procedure, requiring no adjustments and remaining completely transparent to the transmitter.

[0058] Tenth, combining offline and real-time processing. The pre-caching phase is handled offline to avoid real-time processing pressure; the catch-up / waiting phase is executed in real-time to ensure the immediacy of synchronization establishment. For extreme scenarios where Δn < 0, the receiver can choose to enter deep sleep, setting the wake-up time to the end of the wait time to reduce energy consumption.

[0059] Eleventh, construct a unified deterministic communication foundation platform. This invention, together with existing patents, constitutes a complete zero-signaling, intrinsically secure, deterministic communication platform, enabling the upper-layer protocol stack to complete deterministic communication without needing to concern itself with the specific implementation and clock skew of the physical layer, only needing to follow the logical beat of PSSM. Attached Figure Description

[0060] Figure 1 This is the overall flowchart of the cold start two-dimensional synchronization of the present invention.

[0061] Figure 2 This is a timing diagram illustrating pre-caching and backward matching.

[0062] Figure 3 This is a diagram illustrating the logical state catch-up / wait mechanism.

[0063] Figure 4 This diagram illustrates the relationship between the present invention and the existing patent system. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and embodiments. The described embodiments are some, but not all, of the embodiments of this invention. For the sake of simplicity, the following embodiments use a logical counter n as an example of a PSSM state value. It should be noted that although the embodiments use a counter as an example, the claims cover all S(t) forms (such as hash chain values, logical epoch states, and physical layer timing information).

[0065] Example 1: Cold start of data link for high-speed aircraft Before launch, a high-speed aircraft completed DSF pre-injection via a ground test cable to the launch platform. The DSF parameters were configured as follows: K_sec is a 256-bit shared key, Init_Anchor is set to the absolute physical time, Rule_ID adopts a hash chain-driven rule (Rule B), and logical tick T_step = 1ms.

[0066] After launch, the physical connection between the two systems is broken. The aircraft (transmitter) and the launch platform (receiver) independently start their PSSM at the moment their local clocks reach the time specified by Init_Anchor, both evolving from their initial states. Due to the inherent discrepancy between their physical clocks (assuming the aircraft's clock is 500ms slower than the launch platform's), a fixed offset of Δn = 500 exists between their logical states.

[0067] The launch platform executes the synchronization method of this invention: it initiates a 2-second pre-buffered signal reception window and iterates through candidate n values ​​for backward matching. When n=500, the correlation peak exceeds the threshold, indicating a successful match. The launch platform determines Δn=500 and calculates the physical time deviation Δ. Subsequently, the launch platform directly jumps its local PSSM state to alignment with the transmitter and adjusts the reception window position according to Δ. The entire process is completed within approximately 2 seconds after the spacecraft launches, and due to the hash chain-driven approach, the communication has forward security.

[0068] Example 2: Synchronization Maintenance and Closed-Loop Tracking in High-Speed ​​Aircraft Flight Following Example 1, after the spacecraft and the launch platform complete two-level initial synchronization, the spacecraft enters the flight phase. The PSSM logic counters of both sides are aligned, and the receiver has obtained the initial Δ.

[0069] (1) Normal communication phase: At each logical decision moment (T_step=1ms), the aircraft sends a safety pilot frame derived from the current PSSM state, which carries flight status data. The transmitting platform captures the signal within a precise receiving window T_window = T_expected + Δ ± ε, demodulates it using a locally generated pilot copy, and simultaneously measures the actual arrival time T_actual, calculating the timing observation error ε = T_actual - T_expected.

[0070] (2) Closed-loop tracking: The transmitting platform uses exponential smoothing to update Δ: Δ_new = 0.8·ε + 0.2·Δ_old. Due to the high speed of the aircraft, the relative distance changes rapidly, causing the propagation delay to change rapidly. Closed-loop tracking can adaptively adjust Δ to ensure that the receiving window is always aligned with the signal.

[0071] (3) Two-way ranging and positioning: The launching platform simultaneously calculates the real-time distance to the aircraft based on ε and the known TA value: d = c × (TA + ε / 2), which is used for ballistic monitoring and trajectory prediction. The aircraft can also measure the distance using the same method by receiving the downlink pilot signal from the launching platform, thus achieving two-way closed-loop mutual calibration.

[0072] (4) Connection with Loss of Communication Recovery: If the aircraft enters the signal shielded area, causing communication interruption, the PSSMs of both parties continue to evolve independently. After leaving the shielded area, due to the short loss of communication time (about a few seconds), the Δn change caused by crystal oscillator drift is within ±10. At this time, the small window search mechanism (K=10) in patent 2026106011905 is activated to quickly restore synchronization without having to re-execute the cold start process of this invention.

[0073] This embodiment demonstrates that the initial synchronization established by the present invention provides a solid foundation for subsequent deterministic communication, mobility tracking, and recovery from loss of connection, forming a complete technical closed loop together with existing patents.

[0074] Example 3: Initialization of UAV swarms in GPS-free areas A drone swarm consists of 10 drones deployed in a canyon area with no GPS signal and no ground base station. Before the mission begins, all drones are injected with the same group DSF root key K_sec_root through a ground planning station and configured with the same Init_Anchor ("10 seconds after takeoff") and Rule_ID (logical epoch-driven rule, rule C, T_step=100ms).

[0075] After takeoff, each drone acts as the "first communicator," independently activating its PSSM according to its internal clock and beginning to transmit safety pilot signals from its initial logical state (Epoch_ID=0, Counter=0). Due to differences in the takeoff time and internal clocks of each drone, their logical state activation times are different, resulting in varying offsets between them.

[0076] One of the drones designated as the "lead drone" (as the receiver) executes the synchronization method of the present invention: it opens a 10-second capture window near a preset start anchor point (10 seconds after takeoff) to pre-buffer the signals of all other 9 drones.

[0077] After capture, the lead drone iterates through candidate logic state values ​​(corresponding to Counter values ​​from 0 to 100), generates candidate pilots, and performs backward matching on the buffered signals. Due to the different startup times of each drone, their pilots will show correlation peaks at different time positions in the buffer sequence. Through matching, the lead drone can simultaneously calculate the logic state offset Δn_i with each drone (e.g., Δn=5 with drone 2, Δn=12 with drone 3, etc.) and the physical time deviation Δi (including propagation delay caused by distance).

[0078] Subsequently, based on these results, the lead drone adjusts its local logical state (for drones with Δn>0, it jumps directly; for drones with Δn<0, it waits for the corresponding beat) and receiver window position when communicating with each drone, completing the initial synchronization of the logical and physical layers with the entire swarm in one go. Afterward, the swarm can then conduct cooperative formation flight and information sharing under zero signaling conditions, maintaining its formation through continuous two-way ranging.

[0079] Example 4: Long-term sleep wake-up of satellite IoT terminals A low-Earth orbit satellite IoT terminal is deployed in a remote area and has completed its initial configuration with the ground gateway via DSF pre-injection. The terminal enters a deep sleep mode for 30 days. Due to long-term drift of the crystal oscillators of the terminal and the ground gateway (assuming a drift of ±10ppm per day), after 30 days, the cumulative deviation of the logic counters of both parties can reach ±10ppm × 30 days × 24 hours × 3600 seconds / T_step. Assuming T_step = 1 second, then Δn can reach ±25.9 seconds, or approximately ±26 ticks.

[0080] After the terminal is woken up, due to the prolonged loss of connection, the previously stored synchronization information has become completely invalid, and it has actually entered a cold start state. The terminal (as the receiving end) executes the synchronization method of this invention: opening a capture window with a length of 60 seconds (covering the expected Δn range) and pre-buffering the pilot signal sent by the gateway.

[0081] After capture, the terminal iterates through candidate n values ​​(n=0~60) and performs backward matching. When n=26, the relevant peak value exceeds the threshold, and the match is successful. The terminal determines Δn=26 and calculates the physical time deviation Δ. Subsequently, the terminal jumps its local logic counter directly from 0 to 26 and adjusts the receive window position according to Δ, quickly restoring the communication link with the gateway. The entire process does not rely on any external time synchronization or historical records.

[0082] Example 5: Receiver logic counter leading scenario (Δn<0) Following Example 1, let's assume the opposite scenario: the aircraft's clock is 500ms faster than the transmitting platform's clock, causing the logic counter n_tx at the transmitting end (aircraft) to be 500ms smaller than the n_rx at the receiving end (transmitting platform) (i.e., Δn = -500).

[0083] During the offline matching phase, when the receiver iterates through the n values, it will find a relevant peak near n=0 (because the pilot signal from the transmitter arrives earlier than the receiver expects). After a successful match, the receiver determines Δn = -500.

[0084] During the real-time synchronization phase, since the receiver cannot "reverse" the logic counter, the receiver's logic layer synchronization controller will control the local PSSM to enter a waiting state, pausing the increment of the logic counter and waiting for 500 logic ticks (i.e., 500ms). During the waiting period, the receiver can remain silent or send empty frames to maintain basic synchronization. For scenarios with a large negative value of Δn, the receiver can choose to enter deep sleep, setting the wake-up time to the end of the waiting period to reduce power consumption. After the waiting time ends, the transmitter's logic counter will naturally evolve to align with the receiver's current value. After this, logic layer synchronization between the two parties is established, and physical layer synchronization has also been calibrated via Δ.

[0085] Example 6: State Synchronization under Rule B (Hash Chain) This embodiment demonstrates the application of the present invention under rule B (hash chain driven rule). In rule B, the logical state is a secret hash chain value T_State[n], whose evolution follows: T_State[0] = H(K_sec || Init_Anchor || "INIT") T_State[n] = H(K_sec || T_State[n-1] || "CHAIN"), n ≥ 1 The transmitter sends a security pilot sequence derived from T_State[n] at each logical decision point. After the receiver pre-buffers the signal, it needs to traverse the candidate n values ​​(i.e., the candidate hash chain index) to generate the corresponding candidate pilot P_local(n) =PRF(K_sec, T_State[n] || "PILOT"), where T_State[n] is calculated independently by the receiver based on the candidate n.

[0086] After a successful match, the receiver determines the transmitter's current hash chain index n_tx and its local index n_rx, and calculates the offset Δn = n_tx - n_rx. Subsequently, the receiver adjusts its local hash chain index based on Δn (if Δn > 0, it jumps directly to n_tx; if Δn < 0, it waits for |Δn| ticks). Because hash chains have forward safety, the jump operation does not reveal the skipped chain values.

[0087] Example 7: Industrial Robot Production Line Debugging Based on Relative Time Anchor Points On an industrial robot production line, 10 robots need to quickly establish communication after power-on. During DSF pre-injection, Init_Anchor is set to "the 5000th local clock tick after system power-on" (assuming a clock frequency of 1kHz, i.e., 5 seconds later). The start time of each robot's internal clock varies due to differences in power-on timing (maximum difference of 2 seconds). Each robot starts PSSM when its local clock reaches "the 5000th tick after power-on". One robot, acting as the coordinator, opens a capture window near its own start time, executes the pre-buffering and backward matching of this invention, successfully calculates the logical state offset and physical time deviation with other robots, and completes the communication network of the production line robots in one go. This embodiment demonstrates that the present invention does not rely on an absolute external time reference (such as UTC) and can work only based on relative internal events.

[0088] Example 8: Complexity Analysis The computational complexity of the pre-buffered backward matching method of this invention is O(N × M), where N is the number of candidate state values ​​(i.e., the search range), and M is the sliding correlation computation cost for each candidate. After accelerating the correlation computation with FFT, the complexity of a single correlation operation is O(L log L), where L is the pilot sequence length. The overall complexity is O(N × L log L). For typical parameters (N=2000, L=100000), matching can be completed in milliseconds.

[0089] Industrial applicability This invention has broad industrial applicability and can be directly applied to the following fields: (1) Satellite communication and Internet of Things In scenarios such as low-Earth orbit satellite IoT and deep space exploration, communication terminals and ground gateways often experience issues such as long latency, Doppler shift, and lack of GPS signal. The cold-start two-dimensional synchronization method of this invention enables terminals to quickly resume synchronization with the gateway after prolonged sleep periods, without relying on external time synchronization, significantly reducing terminal power consumption and extending battery life.

[0090] (2) Unmanned Aerial Vehicle (UAV) Swarms and Ad hoc Networks In complex environments without GPS signals or ground base stations, drone swarms need to quickly establish secure and reliable communication links. This invention enables each drone to achieve initial synchronization within the swarm through pre-caching and backward matching after independently starting PSSM, providing a deterministic communication foundation for subsequent collaborative formation and task distribution.

[0091] (3) High-speed aircraft data link In the initial link establishment scenario after launch of a high-speed aircraft (such as a missile, drone, or hypersonic vehicle) and the launch platform, there is an unknown offset between the physical clocks of both parties, making external time synchronization unreliable. This invention can complete two-level synchronization within hundreds of milliseconds after launch, and is transparent to the launch end, requiring no modification to the existing launch process.

[0092] (4) Industrial Internet and Autonomous Driving In areas without GNSS coverage, such as factory workshops, underground parking lots, and tunnels, industrial robots and autonomous vehicles need to quickly establish secure communication. This invention enables the device to independently start the PSSM after power-on, completing network synchronization through pre-buffering and backward matching, without the need for pre-deploying synchronization base stations or beacons.

[0093] (5) Military communications and emergency communications In highly contested environments, communication nodes need to avoid transmitting predictable fixed pilot signals to prevent detection and interference. This invention employs hash-chain-driven time-varying pilot signals, where the pilot sequence for each logical beat is unpredictable, providing forward security and resistance to replay attacks. This meets the stringent requirements of military communications for low probability of intercept (LPI) and low probability of detection (LPD). In post-disaster emergency communication scenarios, rescue equipment can rapidly self-organize a network and establish secure and reliable communication links even without any infrastructure.

[0094] (6) Compatibility with existing standards This invention can be directly integrated into the 3GPP 5G / 6G NTN (non-terrestrial network) standard framework as an initial synchronization enhancement scheme for terminals in the absence of GPS signals. Furthermore, this invention is fully compatible with the applicant's existing DSF series patents (2026100015014, 2026100463580, 2026101229059, 2026105584990, 2026105931074, 2026106011905, etc.), and together they can constitute a complete zero-signaling, intrinsically secure, deterministic communication solution.

[0095] (7) Commercial value This invention can be developed into an independent IP core or software module and integrated into products such as baseband chips, communication modules, UAV flight control systems, and satellite communication terminals. Especially in the defense industry, this technology can serve as a core synchronization solution for high-security data links, possessing significant strategic value.

[0096] In summary, this invention has clear industrial applicability and can be widely applied in various wireless communication scenarios.

Claims

1. A two-dimensional synchronization method for DSF cold start based on pre-cached backward matching, characterized in that, The method, applicable to scenarios where both communicating parties have pre-injected the same dynamic security foundation DSF offline, includes: Step A: The first communicating party starts the Protocol Security State Machine (PSSM) and periodically sends security pilot signals; Step B: The second communicating party starts PSSM, and only performs signal acquisition and storage, without real-time matching; Step C: After capture is completed, the second communication party offline traverses the candidate logic state values ​​and performs sliding correlation matching with the stored cached signals to determine the candidate state value corresponding to the maximum correlation value; Step D: The second communicating party calculates the logical state offset Δn and the physical time deviation Δ based on the matching result; Step E: The second communication party adjusts its local logical state according to the logical state offset Δn, and at the same time adjusts the receiving window position according to the physical time deviation Δ, thus completing the initial synchronization of the logical layer and the physical layer.

2. The method according to claim 1, characterized in that, The step of adjusting the local logical state according to the logical state offset Δn includes: If Δn > 0, set the local logic state directly to align with the transmitter, skipping the middle |Δn| beats; If Δn < 0, wait for |Δn| logical ticks until the transmitter naturally evolves to local alignment; for extreme scenarios where Δn < 0, the receiver can choose to enter deep sleep and set the wake-up time to the end of the wait time. If Δn = 0, no adjustment is needed.

3. The method according to claim 1, characterized in that, The secure pilot sequence is generated by the formula P(n) = PRF(K_sec, S(n) || "PILOT"), where PRF is a cryptographic pseudo-random function and S(n) is the PSSM logic state at the nth logic decision time.

4. The method according to claim 1, characterized in that, The PSSM adopts a hash chain-driven rule, where the logical state S(n) is a secret hash chain value, and its evolution follows: S(0) = H(K_sec || Init_Anchor || "INIT") S(n) = H(K_sec || S(n-1) || "CHAIN"), n ≥ 1 H is a collision-resistant hash function; this mechanism ensures that the state of each logical cycle is irreversible, providing forward safety and backward safety.

5. The method according to claim 1, characterized in that, The capture duration T_capture ≥ 2×T_step×N_search, where N_search is the preset number of search steps and T_step is the logical step; N_search is determined by dividing the estimated maximum initial time offset by T_step, rounding up, and adding a safety margin, the value of which ranges from 10 to 100.

6. The method according to claim 1, characterized in that, After the initial two-level synchronization is established, at least one of the following follow-up operations is also included: (1) Generate communication parameters based on the synchronized PSSM state and perform zero-signaling deterministic communication; (2) Continuously measure the timing observation error and track the physical time deviation Δ in a closed loop; (3) Activate the small window search mechanism for rapid recovery during short-term loss of connection; (4) Calculate the relative distance based on the bidirectional measurement values ​​for mobility management.

7. The method according to claim 1, characterized in that, The logical state value is any of the following forms: logical counter n, hash chain state T_State[n], logical epoch state (Epoch_ID, Counter), or physical layer timing information.

8. A two-dimensional synchronization device for DSF cold start based on pre-buffered backward matching, characterized in that, include: The DSF pre-injection module is used to inject the same Dynamic Security Foundation (DSF) into both communicating parties through an offline secure channel. The transmitter PSSM module is used to independently start the PSSM at a specified start time and evolve it according to the logical tick T_step, periodically transmitting the safety pilot sequence; The receiver pre-buffer module is used to continuously capture signals and store them in a buffer near the estimated start time, while recording the local logic state value corresponding to each capture window. The candidate pilot generation module is used to generate a local candidate pilot sequence set; The sliding matching module is used to perform sliding correlation operations between each candidate state value and the buffer signal near the theoretical arrival time. The matching decision module is used to determine the candidate state value that maximizes the correlation value, and to calculate the logical state offset Δn and the physical time deviation Δ. The logic layer synchronization controller is used to adjust the local logic state according to Δn: if Δn > 0, it will be directly aligned; if Δn < 0, it will wait for the corresponding clock cycle. The physical layer synchronization controller is used to adjust the position of the receive window based on Δ.

9. The apparatus according to claim 8, characterized in that, The logic layer synchronization controller works in conjunction with the physical layer synchronization controller to perform real-time synchronization adjustment immediately after offline matching is completed.

10. A wireless communication device, characterized in that, It includes the DSF cold start two-dimensional synchronization device based on pre-cached backward matching as described in claim 8 or 9.

11. The wireless communication device according to claim 10, characterized in that, The device can be any of the following: high-speed aircraft data link terminal, UAV communication terminal, satellite communication terminal, Internet of Things device, autonomous driving vehicle terminal, industrial robot communication module, emergency communication terminal, or individual soldier radio terminal.

12. The method according to any one of claims 1 to 7 is applied in the initial synchronization of satellite communication terminals under conditions where there is no GPS signal, the batch deployment of Internet of Things devices, the rapid establishment of emergency communication networks, and the initialization of data links for high-speed aircraft.

13. A computer-readable storage medium, characterized in that, The device contains a computer program that, when executed by a processor, implements the method of any one of claims 1 to 7.

Images