A state prediction-based method and device for fast synchronization and power recovery after disconnection

Abstract

The application discloses a kind of quick synchronization and power recovery method and device based on state prediction after disconnection.Flight vehicle and candidate control terminal are pre-injected dynamic security root through one-time security channel, candidate control terminal independently runs protocol security state machine, and there is no need for any communication association before switching.During disconnection, both sides estimate crystal oscillator drift using historical calibration record, correct logic counter silently, and estimate physical anchor deviation and distance by predicting relative motion, and set initial transmission power accordingly.At recovery time, both sides directly communicate based on pre-known resource intention and pilot sequence;In multi-user scenario, resource collision is avoided by two-step deterministic arbitration.Candidate control terminal can take over without feeling after detecting that original control terminal stops launching, on the same logical resource.Small data packets can be modulated on orthogonal pilot, enabling same-frequency concurrent collision-free transmission.The application realizes sub-millisecond switching and recovery latency, significantly reduces electromagnetic exposure, and is suitable for strong confrontation and high dynamic scenarios such as unmanned aerial vehicle swarm and missile cluster.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, specifically relating to a synchronization and power recovery method for re-establishing a connection after two communicating parties have lost contact. It is particularly suitable for multi-user, multi-control terminal scenarios such as UAV swarms, missile clusters, and satellite communications, which have high requirements for link reliability, rapid recovery, low probability of interception, and seamless transfer of control. Background Technology

[0002] In mobile communication scenarios involving drones, missiles, etc., the loss of contact between communicating parties due to terrain obstruction, long-distance flight, or electromagnetic interference (hereinafter referred to as "loss of contact") is a common engineering problem. Re-establishing the communication link after loss of contact (hereinafter referred to as "recovery") faces the following core issues: 1. Synchronization Recovery Issues: Traditional solutions rely on random access or signaling handshakes (such as RRC reconstruction), requiring multiple signaling interactions, resulting in recovery delays of tens to hundreds of milliseconds. In highly dynamic scenarios, the relative positions of both parties continuously change during recovery, further increasing the difficulty of synchronization.

[0003] 2. Power control issue: The transmitter needs to select an appropriate transmission power. Existing graded scanning methods have slow convergence speeds, predictable power variation patterns, and high electromagnetic exposure.

[0004] 3. Multi-user conflict issue: When multiple aircraft lose contact and attempt to recover simultaneously, resource collisions lead to an increased recovery failure rate.

[0005] 4. Control transfer problem: Traditional control transfer requires re-pairing frequencies and rebuilding links, resulting in long downtime and low battlefield survivability.

[0006] The applicant's prior patents (application number: 2026100015014, invention title "Method, System and Device for Generating Wireless Communication Parameters Based on Dynamic Security Foundation"; application number: 2026101229059, invention title "Hard Deterministic Wireless Communication Method and System Based on Unified Anchor Time and Dual Fixed Offset"; application number: 2026105584990, invention title "A Wireless Communication Method and System Based on Dual-Level Synchronization and State Feedback"; application number: 2026105931074, invention title "A Logical Layer Autonomous Synchronization and Parameter Adaptation Method and Device Based on Two-Layer Synchronization Architecture") proposed a dual-level synchronization framework based on Dynamic Security Foundation (DSF) and Protocol Security State Machine (PSSM), which realizes continuous synchronization of logical states and physical anchor point alignment during disconnection. Among them, patent application number 2026105931074 discloses "same-period difference mode," "pure local sequence mode," and "downlink pilot measurement mode," which are used by the UE to autonomously adjust the logic step size using local real-time measurements during normal communication (i.e., when there is uplink and downlink interaction or at least the other party's signal can be monitored), without relying on network broadcast signaling. However, the above patents do not solve the following problems: - Autonomous calibration problem of logic counter deviation caused by crystal oscillator drift during disconnection: Although the prior patent (2026105931074) solved the autonomous synchronization of the logic layer during normal communication, in the extreme scenario of complete disconnection (i.e., even the downlink signal of the other party cannot be received), the UE cannot perform any real-time measurement and still lacks a mechanism to calculate the drift and silently correct it using the historical calibration records stored before the disconnection.

[0007] - The problem of predicting transmit power when reconnecting after loss of contact (current recovery relies on blind scanning); - Resource arbitration issue when multiple users resume simultaneously (the existing two-step arbitration is not deeply integrated with the scenario of recovery from loss of connection). - The issue of seamless transfer of control between different handheld terminals (existing PSSM mirroring patents rely on context transfer between base stations, which is not suitable for point-to-point control terminal handover and requires the establishment of communication association before handover).

[0008] In particular, in point-to-point control transfer scenarios, existing PSSM mirroring patents require the target control end to obtain the PSSM state context of the source control end via signaling before the handover, meaning that the two parties must have a communication association before the handover. The "pre-synchronization zero association" mechanism proposed in this invention is completely different: all candidate control ends inject DSF once before takeoff and synchronize independently with the aircraft, without any communication association or state transfer before the handover.

[0009] Furthermore, the applicant's prior patent (application number: 2026105584990) discloses a method for deriving individual DSFs based on the group DSF root key, enabling each node within the group to run PSSM independently and maintain logical synchronization. However, this patent mainly addresses the issue of new nodes dynamically joining the group (by listening to silent calibration tokens, parsing counters, and calculating start anchors), without revealing the inherent orthogonality of physical and pilot resources of multiple nodes under logical synchronization after the group is formed, nor does it propose a communication method to achieve collision-free concurrent transmission using this characteristic. This invention further reveals and utilizes the following key characteristics based on group DSF derivation: all nodes share a globally unified Rule_ID, thus achieving complete logical time synchronization; however, due to the differences in K_sec and Init_Anchor in individual DSFs, the pilot sequences and encryption keys generated by each node are inherently orthogonal. Although their resource intentions (time-frequency positions) may collide due to modulo operations, the pilots themselves can remain mutually non-interfering. This characteristic can be summarized as "logical synchronization, physical separation." Summary of the Invention

[0010] Purpose of the invention The purpose of this invention is to provide a method for rapid synchronization and power recovery after disconnection based on state prediction, to solve problems in existing technologies such as reliance on signaling for disconnection recovery, blind power control, multi-user resource conflicts, and the need for pre-communication association for control transfer. This achieves the following technical effects: during disconnection, the logic counter drift is autonomously corrected using historical calibration records, without relying on network signaling; after disconnection, rapid synchronization and power prediction are based on predicted distance and timing errors, reducing the number of attempts and lowering electromagnetic exposure; when multiple users recover simultaneously, collisions are avoided through anticipated resource intentions and two-step arbitration; control is seamlessly transferred between different handheld terminals without any prior communication association; pilot orthogonality derived from group DSF is used to achieve collision-free concurrent transmission of small data packets; the handover interruption time can be shortened to sub-millisecond level, and it is seamlessly compatible with existing DSF patent systems.

[0011] Core concept definition To facilitate understanding of this invention, the following key concepts are first clarified. Some of these concepts are based on the definitions in the applicant's prior patents (2026100015014, 2026101229059, 2026105584990, 2026105931074), while others are new or further defined in this invention.

[0012] Concept 1: Dynamic Security Foundation (DSF) A shared cryptographically secure triple (K_sec, Init_Anchor, Rule_ID) is used for the generation of all deterministic parameters and state synchronization between the communicating parties. Specifically: K_sec is the shared key, distributed securely in a pre-shared manner; Init_Anchor is the initial anchor point, defining the start point of the logical timeline and the initial logical state S(0); and Rule_ID is the state transition rule identifier, containing parameters such as the logical tick length T_step, the state update function, and the calibration period.

[0013] Concept 2: Protocol Security State Machine (PSSM) A deterministic state machine based on the DSF evolves independently for each communicating party, outputting a time-varying logic state S(t). The evolution of the PSSM is independent of the physical clock and is driven only by the logical tick defined by the Rule_ID. When both parties share the same DSF, the PSSM remains synchronized on the logical timeline.

[0014] Concept 3: The Logical Decision Moment Discrete moments on the logical timeline defined by Rule_ID. At each logical decision moment, PSSM outputs the current state S(t) and triggers the calculation of communication parameters (such as pilot sequences and resource indices) and physical layer operations.

[0015] Concept 4: Unified Anchor Point Time and Fixed Offset At the logical decision-making moment, the communication node instantaneously reads its local physical clock, and the resulting value is recorded as the unified anchor point moment T_anchor. Further, the physical layer operation moment is determined based on the pre-configured fixed offsets Δ_dl (downlink offset) and Δ_ul (uplink offset): - Downlink listening / transmission time: T_DL = T_anchor + Δ_dl; - Uplink transmission time: T_UL = T_anchor + Δ_ul.

[0016] Where Δ_ul > Δ_dl, and Δ_ul - Δ_dl is not less than the sum of the maximum processing delay, propagation delay, and guard interval.

[0017] Concept 5: Loss of contact This refers to a state in which both communicating parties fail to successfully receive a valid signal from the other party within a series of logical decision moments (typically 3 to 5). During the period of disconnection, the logical state (first-level synchronization) of the PSSM is still maintained and evolved independently by both parties, but the physical anchor alignment (second-level synchronization) may fail.

[0018] Concept 6: Historical Calibration Records This refers to historical data stored in local memory before the loss of connection, used to characterize the crystal oscillator's drift characteristics. This includes, but is not limited to: sequences of correction values ​​from each silent calibration; sequences of drift trends from each same-period difference mode measurement; sequences of capture bias changes from each downlink pilot measurement mode measurement; and sequences of historical readings from the local temperature sensor. These records are used to estimate the long-term drift rate and drift trend of the crystal oscillator during the period of loss of connection.

[0019] Concept 7: Window before loss of contact This refers to a continuous logical decision-making time interval preceding the loss of contact. Within this window, both communicating parties are in normal communication status, continuously performing two-way ranging, timing observation error measurement, and accumulating historical calibration records.

[0020] Concept 8: Synchronized Search Window This refers to the time-codeword joint search range set by the receiver during reception recovery to tolerate potential residual deviations in the logic counter. Centered on the predicted recovery logic counter value n0, the range expands K steps in both positive and negative directions, generating 2K+1 candidate pilot sequences. Sliding correlation detection is then performed within the corresponding physical time window. The half-width K of the window is determined based on the duration of disconnection, the maximum uncertainty of the crystal oscillator, and a safety margin.

[0021] Concept 9: Two-Step Deterministic Arbitration When multiple user devices' uplink resource intentions are mapped to the same physical resource, the network side performs the following two steps to resolve the conflict: - Step 1 (Inter-user conflict arbitration): Select a winner from the conflict set according to a preset strategy (static priority, round-robin fairness, weighted fair queue, earliest deadline priority, etc.); - Step 2 (Resource Availability Check): Query the global resource calendar to confirm whether the physical resources intended by the winner are idle and meet the spatial / beam isolation conditions at the corresponding time.

[0022] After a successful arbitration, the network side notifies the winner via downlink implicit authorization.

[0023] Concept 10: Downlink Implicit Authorization This refers to the network side sending a valid downlink signal (data frame, control frame, or empty frame) on the downlink listening resource R_d known to the user equipment. The physical event of the signal's "existence and successful decoding by the target user" itself constitutes authorization for the user to use the paired uplink resource R_u, without requiring any explicit authorization signaling bits. Technical solution

[0024] To achieve the above objectives, this invention provides a method for rapid synchronization and power recovery after disconnection based on state prediction, comprising the following steps: Part 1: Pre-synchronization Configuration Before the Loss of Contact Step P0: One-time DSF pre-injection Before the aircraft takes off or the mission begins, complete the following configuration in one go via a secure channel (wired connection, short-range wireless, or physical contact): - Point-to-point scenario: The aircraft shares the exact same DSF triple (K_sec, Init_Anchor, Rule_ID) with all candidate control terminals (pilot terminals, backup ground stations, etc.). Subsequently, each candidate control terminal runs the PSSM independently, and its logical state is automatically synchronized with the aircraft. There is no communication between the candidate control terminal and the aircraft before the handover—that is, the candidate control terminal does not need to receive the aircraft's status information, synchronization signaling, or context passing; logical synchronization is achieved solely through the pre-injected DSF and the local physical clock (used only to read the T_anchor).

[0025] - One-to-many scenario: Multiple aircraft and multiple control terminals (e.g., ground stations, multiple pilot terminals) share the same group DSF root key K_sec_root and a globally unified Rule_ID (denoted as Rule_ID_global). Each node (whether an aircraft or a control terminal) has a unique identifier ID_node, and derives its individual key and initial anchor point from the root key based on its own ID: K_sec_node = PRF(K_sec_root, ID_node, "NODE") Init_Anchor_node = PRF(Init_Anchor_root, ID_node, "ANCHOR") Each node's complete individual DSF triple is (K_sec_node, Init_Anchor_node, Rule_ID_global). All nodes run PSSM independently based on their respective individual DSFs. Because the Rule_ID is globally unified and its derivation rules are consistent, the logical states of all nodes remain aligned at the same logical moment (i.e., the logical counter n increments synchronously). However, because K_sec_node and Init_Anchor_node are different, the pilot sequences and encryption keys generated by each node are naturally orthogonal. Although resource intentions R_u may collide due to modulo operations, the pilots themselves can remain independent. This characteristic can be summarized as "logical synchronization, physical separation".

[0026] In this configuration, any control terminal that needs to communicate with a specific aircraft can use the aircraft's individual DSF (i.e., K_sec_node and Init_Anchor_node) to generate communication parameters (resource intention, pilot sequence, encryption key) at any logical decision point. The aircraft uses its own individual DSF to generate reception parameters, and can only be correctly demodulated if the received signal uses parameters that match its individual DSF. Because the pilots of different nodes are orthogonal, multiple aircraft can concurrently send small data packets on the same physical resources without interfering with each other, or coordinate the transmission of large data blocks through two-step arbitration.

[0027] Part Two: Autonomous Synchronization Maintenance During the Period of Disconnection Step S1: First-level synchronization maintenance (independent evolution of logic counters) During the period of disconnection, both communicating parties continue to run their PSSMs independently. The logic counter n increments normally according to the predetermined rhythm T_step, and the logic state S(t) is updated normally. Neither party needs to receive signals from the other, and theoretically, they can maintain logical state synchronization.

[0028] Step S2: Autonomous Drift Calibration During Disconnection Due to crystal oscillator frequency errors, the logic counters of both parties will accumulate deviations during actual operation. During the period of disconnection, each communicating party independently performs the following autonomous drift calibration procedure, requiring no network-side signaling: Sub-step S2-1: Extraction of historical drift trends Read the historical calibration records stored before the loss of contact, including: - A sequence of correction values ​​from each silent calibration (if there was previous network coverage). - Drift trend sequence of previous same-period difference pattern measurements (if there is up-down interaction, see prior patent 2026105931074). - Sequence of acquisition bias changes in each downlink pilot measurement mode (if only downlink can be listened to, see prior patent 2026105931074). - Local temperature sensor historical reading sequence (used for temperature drift compensation).

[0029] Filter the historical drift trend (weighted moving average, Kalman filter, or linear regression) to estimate the current crystal oscillator's long-term drift rate γ (unit: ppm) and drift rate change trend dγ / dt.

[0030] Sub-step S2-2: Drift estimation during the period of loss of contact Let the duration of the disconnection be T_loss, and the nominal period of the logic clock be T_step. Then, the estimated value of the accumulated logic counter deviation during the disconnection period is: Δn_est = γ · T_loss / T_step + 0.5 · (dγ / dt) · (T_loss / T_step)² The second item is an optional higher-order correction item, which is suitable for scenarios where the connection is lost for a long time.

[0031] Sub-step S2-3: Temperature compensation correction When the transmitter is equipped with a temperature sensor, it reads the temperature change records during the period of disconnection (or the temperature difference before and after disconnection), and calculates the temperature-induced drift contribution based on the pre-stored crystal oscillator temperature characteristic curve γ_temp(T): Δn_temp = ∫ γ_temp(T(t)) dt / T_step The temperature-induced drift can be separated from the total drift or directly used as a superposition correction term.

[0032] Sub-step S2-4: Multi-source fusion correction When multiple information sources are available simultaneously, the fusion drift rate γ_fused is estimated through weighted fusion. Specifically, the long-term drift rate of historical calibration records, the recent rate of change extracted from the ranging sequence before the loss of connection, and the temperature sensor readings are each assigned a weight, with each weight dynamically adjusted according to the confidence level of the corresponding information source. Then, the weighted average is calculated as the fusion drift rate.

[0033] Sub-step S2-5: Silent correction of local counters Based on the calculated Δn_est (or the fused correction amount), a silent correction is performed on the local logic counter: n_local_new = n_local_old - round(Δn_est) The correction process requires no external signals and does not generate air interface signaling. The magnitude of a single correction is limited to a preset safety range (e.g., ±N_max steps, typically 10 steps). If the magnitude exceeds the range, it is marked as "requiring network-side auxiliary calibration" and a lightweight status query is triggered after recovery.

[0034] Step S3: Second-level synchronization preparation (physical anchor point prediction) During the period of lost contact, although the logic counters of both communicating parties have been synchronized after the autonomous drift calibration in step S2 (i.e., both parties have a consistent understanding of the current logic time n), the timing alignment of the physical layer may still fail. There are two reasons for this: first, changes in the relative positions of both parties lead to changes in propagation delay; second, residual frequency errors in the crystal oscillator cause a cumulative offset of the local physical clock relative to the nominal time. Therefore, before sending the recovery signal, the transmitting end needs to predict the physical layer timing deviation at the end of the lost contact and adjust the timing advance (TA) in advance to ensure that the recovery signal falls accurately within the receiving end's time window.

[0035] Define ε(k) as the difference between the arrival time of the k-th signal measured by the receiver before the connection loss and the theoretically expected time (this difference comprehensively reflects propagation delay, crystal oscillator drift, and processing delay). Based on the ε(k) sequence recorded within a time window before the connection loss, the relative velocity v_ε and acceleration a_ε are estimated using a uniform motion model or a uniformly accelerated motion model. Then, the predicted value of the physical anchor point deviation after the connection loss duration T_loss is: ε_pred = ε_last + v_ε · T_loss + 0.5 · a_ε · T_loss² Where ε_last is the deviation value measured last before the loss of contact.

[0036] The physical meaning of ε_pred is: if the transmitter still uses the timing advance before the loss of connection to send the signal, the predicted deviation between the actual arrival time measured by the receiver and the theoretically expected time. The transmitter adjusts its local timing advance accordingly: TA_new = TA_old - ε_pred (or a more refined correction algorithm). In this way, when the signal is resumed, its actual arrival time will be basically aligned with the theoretically expected time of the receiver, thus achieving rapid recovery of the physical anchor point.

[0037] Part 3: Post-Loss of Contact Recovery Preparation (Synchronization and Power Prediction) Step S4: Recovery Timing and Resource Prediction Due to the maintenance of first-level synchronization (via autonomous drift calibration), the transmitter accurately predicts the following information from the receiver at the recovery time: - The logical decision time t_restore (determined by the logical counter n and T_step); - Uplink resource intention R_u(t_restore) = F(K_sec, S(t_restore), "UL") mod M; - Downlink monitoring resource R_d(t_restore) = Pair(R_u, K_sec, S(t_restore)); - Safety pilot sequence P(t_restore) = F(K_sec, S(t_restore), "PILOT").

[0038] The above information requires no signaling negotiation and is calculated independently by both parties.

[0039] Step S5: Predicted Power Calculation Estimate the path loss and set the initial transmit power based on the predicted physical anchor point deviation ε_pred or the converted predicted distance d_pred: P_tx_init = P_rx_target + PL(d_pred) + G_loss Where P_rx_target is the target power at the receiver, PL() is the path loss function (e.g., free space model PL(dB)=20log10(d)+20log10(f)+32.4), and G_loss is the system margin (including shadow fading, multipath, etc.). If the transmitter has a pre-stored environmental attenuation map, a more accurate loss estimate can be obtained by querying the map in conjunction with the predicted distance.

[0040] Part 4: Multi-user resource arbitration (applicable to large data blocks or high-load scenarios) Step M1: Multi-user disconnection detection The network side (ground station or LPC) maintains a synchronized copy of the PSSM for all aircraft and detects multiple aircraft losing contact simultaneously or sequentially.

[0041] Step M2: Restore Time-of-Flight Resource Prediction The network side predicts the resource information of each aircraft at its respective recovery time: t_restore_i, R_u(t_restore_i) and the corresponding physical resource mapping Map(R_u(t_restore_i)).

[0042] Step M3: Resource Conflict Prediction Determine if the uplink resource intentions of multiple aircraft map to the same physical resource. Conflict condition: Map(R_u(t_restore_i)) = Map(R_u(t_restore_j)) (for i ≠ j) Step M4: Two-Step Deterministic Arbitration For each physical resource involved in a conflict, execute: - Step 1: User Conflict Arbitration: From the conflicting user set U_conflict, a winner is deterministically selected according to a preset strategy. The arbitration strategy includes at least one of the following: static priority strategy, round-robin fairness strategy, weighted fairness queue strategy, and earliest deadline priority strategy; the strategy can be dynamically configured.

[0043] Step 2: Resource Availability Check: Query the global resource calendar to confirm that the winner's physical resources simultaneously meet the following requirements at the corresponding uplink transmission time: not occupied by high-priority services; not locked by other aircraft at the same time; and spatial / beam isolation meets the requirements (if applicable). If these conditions are met, the arbitration is successful; otherwise, the arbitration fails, and the process proceeds to the next round according to preset rules or triggers the coordination mechanism.

[0044] Step M5: Downlink Implicit Authorization Execution For aircraft that successfully arbitrate, the network side sends a recovery authorization signal (bearing a security pilot frame) on its known downlink listening resource R_d(t_restore_i); for aircraft that fail to arbitrate, no signal is sent or a silence instruction is sent.

[0045] Step M6: Retreat and Retry of Failed Aircraft If an aircraft that fails to arbitrate does not detect a valid authorization signal on the predicted downlink resources, it determines that the recovery has failed and retains the data packets to be sent in the local queue, waiting to re-participate in arbitration at the next logical decision moment. If consecutive failures exceed a preset threshold (e.g., 3 times), it will initiate power increment or fall back to the traditional reconstruction process (e.g., random access).

[0046] Part Five: Resumption of Execution and Closed-Loop Correction Step S6: Resume signal transmission At the predicted recovery time t_restore, the ground station (or the control unit of the winning aircraft) uses the predicted safety pilot sequence P(t_restore) and the set initial transmit power P_tx_init to transmit a recovery signal on the predicted downlink listening resource R_d(t_restore). The recovery signal can carry a synchronization probe (for physical anchor alignment measurements) or a complete data frame.

[0047] Step S7: Receiver processing At the same logical time t_restore, the aircraft (receiver) uses the locally synchronized pilot copy P(t_restore) to perform sliding correlation detection on the received signal: - If the relevant peak value exceeds the threshold, the recovery is considered successful; - Measure the actual arrival time t_actual, and calculate ε = t_actual - t_expected; - Optionally, the signal strength indicator RSSI is measured; - Optionally, the residual deviation between the local logic counter and the transmitter (such as the counter value carried in the demodulated recovered signal) is measured. If a synchronous search window is used, the receiver uses multiple sets of candidate pilots for detection within the window and extracts the correction amount.

[0048] Step S8: Closed-loop correction The aircraft transmits a simplified response frame on a predetermined uplink feedback resource, containing at least one of the following information: - Measured ε value (used for TA correction); - Measured RSSI value (used for power correction); - Measured counter deviation (used for secondary calibration of logic counters).

[0049] After the ground station receives the response: - Update TA based on ε: TA_new = TA_old + α·ε; - Update the power reference based on RSSI: P_ref_new = P_ref_old + β·(RSSI_target -RSSI_measured); - Calibrate the local logic counter based on the counter deviation (if there is residual error in the self-calibration).

[0050] Step S9: Recovery complete When multiple consecutive successful transmissions and receptions occur and |ε| ≤ Δ_sync, the link is considered fully restored, the recovery mode is exited, and the normal communication mode is entered.

[0051] Part 6: Seamless Transfer of Control (Synchronization and Power Framework Based on the Same DSF Pre-Injection) Step H1: Inject before takeoff All candidate control terminals (pilots A, B, C, etc.) and the aircraft are injected with the same DSF (point-to-point scenario) or group DSF root key (one-to-many scenario) via a secure channel before takeoff. Each control terminal derives its individual DSF based on its own ID (if it is a group), and runs the PSSM independently with the aircraft, maintaining logical synchronization. This one-time injected DSF is also the foundational safety parameter for autonomous drift calibration, recovery time resource prediction, predicted power calculation, and closed-loop correction during the loss of contact between the aircraft and the ground station, as described in Parts 1 to 5 above. Therefore, the candidate control terminals obtain the exact same synchronization foundation as the aircraft before takeoff, can independently maintain logical synchronization without any subsequent signaling, and autonomously execute the same synchronization and power recovery procedures as the aircraft when contact loss occurs.

[0052] Step H2: Normal operation Pilot A's terminal acts as the current active control terminal, sending uplink control signals at each logical decision point. Pilots B and C's terminals are in "listen-synchronize" mode: the PSSM continuously evolves and synchronizes with the aircraft (this synchronization is guaranteed by the aforementioned independent evolution of the PSSM and the autonomous drift calibration in step S2), but does not send any uplink signals. During this period, pilots B and C can simultaneously perform the disconnected autonomous drift calibration in step S2 (listening to signals emitted by the aircraft, or relying solely on local historical records) to ensure that their logic counters are strictly aligned with the aircraft.

[0053] Step H3: Switch Trigger Pilot A stops transmitting (due to voluntary withdrawal, terminal failure, being hit, or tactical need for silence). Switching conditions include, but are not limited to: Pilot A actively pressing the "transfer" button, signal strength falling below a threshold, external command (ground station notification), or Pilot B detecting that Pilot A's uplink signal has disappeared for more than a preset number of consecutive logic cycles.

[0054] Step H4: Seamless Takeover (Method 1) Pilot B detects that Pilot A has stopped transmitting (either by monitoring the uplink signal disappearance of Pilot A through its own receiving link until it exceeds a preset threshold, or by receiving a handover notification through the inter-pilot communication link). Pilot B's terminal directly transmits control signals on the anticipated uplink resources. Since the PSSM states of the two terminals have been synchronized through the aforementioned autonomous drift calibration in step S2, the generated pilot signals, resource locations, and encryption parameters are identical. The aircraft receives the signal on the expected resource, demodulates it successfully, and continues to execute commands, unable to distinguish the signal source, resulting in a seamless handover. The interruption time depends only on the time difference between Pilot B detecting that A has stopped transmitting and starting to transmit again, typically less than one logic cycle (e.g., 0.125ms).

[0055] Step H5: Authentication Takeover (Method 2, Optional) In scenarios requiring audit logs, permission changes, or security policy switching, pilot B sends a takeover request frame to the aircraft. This frame is generated using pilot B's individual DSF derived parameters and includes: requester identifier (pilot B's ID), target aircraft identifier, timestamp, priority (if there are multiple levels of control), and optional permission parameters (such as "listen only," "send commands," or "modify tasks"). The aircraft verifies whether the requester ID is in the authorized list. If so, it replies with a takeover confirmation frame, indicating that it is ready to accept the switchover. The original pilot A stops sending after receiving the confirmation (or the aircraft ignores A's signal at a specified time). Pilot B begins sending at the first logical decision moment after confirmation. This method supports audit trails, hierarchical permissions, and anti-preemption (the successor can only take over after the current controller releases or loses contact).

[0056] Step H6: One-to-many communication capability Based on the aforementioned group DSF configuration for one-to-many scenarios, the same control unit can select any aircraft's individual DSF to generate a signal and communicate with that aircraft at any logical decision moment, without needing to establish a connection or negotiate parameters beforehand. Each aircraft uses its own individual DSF to receive signals and only responds to matched signals. This capability also relies on the aforementioned synchronization and power prediction framework to ensure that aircraft can quickly restore communication with the control unit after losing contact.

[0057] Part 7: Collision-free concurrent transmission based on orthogonal pilots (preferred mode for small data packets) The prerequisites upon which this section relies are the same as those for all the preceding parts of this patent: the aircraft and the control unit have completed DSF pre-injection through a one-time secure channel (see Part 1), both communicating parties operate PSSM independently, their logical states remain synchronized (see steps S1 and S2), and the physical anchor points have been basically aligned through TA closed-loop or predictive methods (see step S3). Under these premises, this section further utilizes the characteristics of "logical synchronization and physical separation" to provide a collision-free concurrent transmission mode for small data packets. This mode can be seamlessly integrated with the aforementioned loss-of-connection recovery process: when the aircraft recovers from loss of connection, if only small data packets (such as status reports and sensor readings) need to be sent, this mode can be used directly; if large data blocks need to be sent, it reverts to the two-step arbitration mode in Part 4.

[0058] Step C1: Sender preparation The transmitting end (control unit or aircraft) determines the individual DSF of the target receiver (if communication with a specific aircraft is required) or uses its own individual DSF (if broadcasting). The small data packet to be transmitted (length L bits, where L is less than a preset threshold, e.g., 64 bytes) is directly modulated onto a safety pilot sequence derived from the current logic state S(t) and the individual DSF. Specifically, the reference pilot sequence is set to P_base = F(K_sec_node, S(t), "PILOT"), with a length of N_chip. The transmitting end spreads and scrambles the data bits b and then superimposes them onto P_base, or directly selects different orthogonal sequences using different data bits (e.g., using sequence selection modulation). Finally, the transmission waveform is generated. The logic state S(t) used here is completely consistent with the resource prediction and safety pilot sequence generation in the aforementioned loss-of-connection recovery step S4, both derived from the same PSSM evolution and DSF derivation.

[0059] Step C2: Concurrent Sending Multiple transmitters can simultaneously send their respective security pilot frames on the same time-frequency physical resources because the pilot sequences of each node are approximately orthogonal due to differences in individual DSFs. The receiver does not need to know in advance which nodes are transmitting, which does not contradict the aforementioned "predictive recovery time resource information" of this patent, but is a natural extension of orthogonality. If the receiver has just recovered from a loss of connection, it can first perform a synchronization confirmation according to step S7 in Part 5, and then start concurrent reception in this mode.

[0060] Step C3: Multi-user detection at the receiver The receiver uses locally generated pilot copies of each node (calculated from the node ID and PSSM state) to perform parallel matched filtering or multi-user detection (such as matched filter banks, continuous interference cancellation) on the received signal. For each target node i, the correlation value ρ_i = |∑ r(t)·P_i(t)|² is calculated. If ρ_i exceeds the threshold, the data packet of that node is considered to have been successfully received, and the data bits are demodulated.

[0061] Step C4: Collision Handling and Blending Mode If detection ambiguity occurs due to non-orthogonal channels or the number of nodes exceeding orthogonal capacity, the receiver can fall back to the two-step arbitration mode in Part 4 or request a retransmission. The system can dynamically select the transmission mode according to the service type: small data packets use pilot concurrent mode, and large data blocks use two-step arbitration mode.

[0062] The core advantage of this mode lies in leveraging the inherent pilot orthogonality of the group DSF to enable simultaneous, same-frequency concurrency among multiple users. The recovery latency is only one logical cycle (synchronous reception), and resource collision probability is completely eliminated. This mode shares the same DSF, PSSM state evolution, and physical anchor alignment mechanism as the aforementioned disconnection recovery framework, requiring no additional signaling or configuration. It can be directly used as the default time-saving transmission method after disconnection recovery. It is particularly suitable for scenarios such as massive terminal reporting in the Internet of Things, cellular broadcast short messages, and distributed sensing data fusion.

[0063] Part 8: Synchronized Search Window (Enhanced Recovery Robustness) To avoid single-point synchronization failure due to residual deviations in the logic counter or crystal oscillator temperature drift after a prolonged period of disconnection, the receiver can set a synchronization search window when resuming reception. The specific steps are as follows: Sub-step W1: Determine the window half-width K. The receiver calculates the window half-width K = ceil(Δf_max · T_loss / T_step) + δ based on the disconnection duration T_loss, the maximum uncertainty of the crystal oscillator Δf_max (ppm), and the system's preset safety margin δ. Typically, δ is taken as 1 to 3.

[0064] Sub-step W2: Generate multiple sets of candidate pilot copies. The receiver uses the local PSSM to generate 2K+1 sets of candidate pilot sequences corresponding to n0 - K, ..., n0, ..., n0 + K, centered on the logical counter value n0 corresponding to the known recovery logic time t_restore.

[0065] Sub-step W3: Sliding correlation detection. The receiver performs continuous sliding correlation detection on the received signal within the physical time window [t_restore - K·T_step - T_win, t_restore + K·T_step + T_win], where T_win is the system-preset arrival time uncertainty window (usually 2 to 3 times the length of the cyclic prefix).

[0066] Sub-step W4: Joint decision. If a candidate pilot causes the relevant peak value to exceed the threshold and the arrival time deviation is within the preset range, then synchronization is determined to be successful, and the logic counter correction amount and physical anchor point deviation are extracted.

[0067] This search window can be used as a supplement to the aforementioned Parts 1 through 6, or in conjunction with Part 7 (Pilot Concurrency Mode).

[0068] Beneficial effects Compared with the prior art and the applicant's previous patents, the present invention has the following beneficial effects: 1. Pre-synchronization zero communication association: The candidate control terminal and the aircraft do not need any communication association before the handover. The logic state is independently synchronized only by the DSF injected once, without the need for state transmission or signaling interaction.

[0069] 2. Logically synchronized, physically separate group communication architecture: All nodes in the group have completely synchronized logical timelines, but the pilot sequences and encryption keys are naturally orthogonal due to the differences in individual DSFs. Physical resources can be coordinated through two-step arbitration, or concurrency can be achieved directly by utilizing the orthogonality of pilots.

[0070] 3. Autonomous drift calibration during disconnection: The system independently calculates and corrects crystal oscillator drift during disconnection by utilizing historical calibration records, without the need for network signaling, thus ensuring that the logic counter maintains high accuracy even after a long period of disconnection.

[0071] 4. Synchronization and power prediction integration: Based on the maintained logical state, the recovery resources are predicted. At the same time, the physical anchor point deviation and distance are predicted by using the ranging sequence before the loss of connection. This achieves dual prediction of synchronization and power at the time of recovery, avoiding blind scanning and significantly reducing electromagnetic exposure.

[0072] 5. Multi-user resource arbitration and disconnection recovery collaboration: Two-step deterministic arbitration is applied to the scenario of simultaneous recovery of multiple aircraft. By arbitrating in advance by knowing the resource intentions, collisions are avoided, and a failure backoff and retry mechanism is designed.

[0073] 6. Seamless transfer of control: Through a pre-synchronization mechanism, the pilot switching time is less than one logical cycle (e.g., 0.125ms), and the aircraft is unaware of the change throughout the entire process, greatly improving battlefield survivability.

[0074] 7. Closed-loop correction ensures link quality: By feeding back ε, RSSI and counter deviation, TA and power are continuously fine-tuned to enable the link to quickly converge to the optimal operating point.

[0075] 8. Extremely short switching interruption time: The recovery process usually only requires one logic cycle, with a typical measured value of 0.125ms, which is far superior to traditional solutions.

[0076] 9. Seamlessly compatible with existing DSF patent system: Only the prediction module, autonomous drift calibration module, sliding window synchronization module, and multi-user arbitration module need to be added, which can be achieved through software upgrade.

[0077] 10. Enhanced recovery robustness with sliding synchronization window: It can tolerate the residual deviation of the logic counter ±K steps after disconnection, significantly improving the recovery success rate in scenarios of long-term disconnection, large crystal oscillator uncertainty, or drastic temperature changes.

[0078] 11. Collision-free concurrent transmission based on pilot orthogonality: Utilizing the pilot orthogonality granted by the group DSF, multiple nodes can concurrently send small data packets on the same time-frequency resources without the need for two-step arbitration, resulting in high resource utilization and a recovery delay of only one logical cycle. Attached Figure Description

[0079] Figure 1 This is the overall flowchart of the present invention.

[0080] Figure 2 The flowchart for the autonomous drift calibration sub-process during the period of loss of contact.

[0081] Figure 3 The flowchart for the two-step arbitration process for multiple users is shown below.

[0082] Figure 4 A sequence diagram for seamless transfer of control.

[0083] Figure 5 This is a diagram illustrating the synchronized search window.

[0084] Figure 6 This is a schematic diagram of collision-free concurrent transmission based on orthogonal pilot signals.

[0085] Figure 7 This is a diagram of the overall system structure. Detailed Implementation

[0086] Example 1: Resumption of communication between UAV and ground station (including autonomous drift calibration) A drone performing a power line inspection mission has established a Direct Flow Sequence (DSF) with a ground station, with a logical tick time T_step = 0.125 ms. The ground station stores the most recent 50 silent calibration records, showing a long-term crystal oscillator drift rate γ = 2.3 ppm. The drone flies into a valley, and communication is lost for 3 seconds. During the loss of communication, the ground station performs autonomous drift calibration: the cumulative counter deviation Δn_est = 2.3e-6 × 3 / 0.000125 = 0.0552 steps, which is less than 0.5 steps and requires no correction. Based on the ranging sequence of the last 20 logical moments before the loss of communication (relative distance gradually changing from 850m to 862m), the relative radial velocity v_rel ≈ 8.5m / s is estimated, and the distance d_pred = 887.5m is predicted after 3 seconds. With a carrier frequency of f=2.4GHz, target received power P_rx_target=-90dBm, system margin G_loss=5dB, calculated path loss PL≈99.1dB, and initial transmit power P_tx_init=14.1dBm, after the UAV flies out of the valley, the ground station sends a recovery signal on the known downlink resources. The UAV uses a synchronous search window (window half-width K=1) for sliding correlation detection, achieving a successful response on the first attempt, with a recovery latency of 2.8ms. Traditional hierarchical scanning schemes require 11 attempts, with a recovery latency of approximately 30ms.

[0087] Example 2: Resource Arbitration for Simultaneous Recovery of Missile Clusters Three hypersonic missiles simultaneously fly towards their target, encountering interference en route and losing contact for 1.5 seconds. The ground station anticipates resource recovery: Missile A (high priority, R_u=PRB5), Missile B (medium priority, R_u=PRB5), and Missile C (medium priority, R_u=PRB8). The conflict set PRB5 includes missiles A and B. According to static priority arbitration, missile A wins. The ground station sends an authorization signal on missile A's R_d (PRB5), and missile A recovers successfully; missile B does not receive authorization, backs off to the next logical time step, recalculates its resource index mapping to PRB7, and then recovers successfully; missile C recovers directly without conflict. The longest recovery delay is less than 3ms. Without arbitration, simultaneous use of PRB5 by A and B would lead to a collision, significantly increasing the probability of recovery failure.

[0088] Example 3: Seamless handover between pilots (point-to-point zero-association) In a reconnaissance drone, pilots A and B inject identical DSF (Dynamic Sequence Message) into their respective handheld terminals and the drone via a secure cable before takeoff. After takeoff, pilot A controls the drone, while pilot B's terminal is in listening synchronization mode (PSSM synchronization, no uplink transmission). During flight, pilot A is threatened by enemy fire and needs to evacuate urgently, shutting down their terminal. Pilot B's terminal does not detect pilot A's uplink signal for two consecutive logic cycles (0.25ms) (i.e., cannot hear A's transmission), and automatically begins sending control signals on the anticipated uplink resource. Because the PSSM is fully synchronized, the drone receives the signal on the expected resource, demodulates successfully, and continues its mission without any awareness of the switch. Pilot B has never established any communication association with the drone before the switch, nor has it received any status information; it only relies on the pre-injected DSF for independent synchronization. In traditional solutions, the prospective pilot must pre-pair with the drone or obtain its status via signaling, making zero-association switching impossible.

[0089] Example 4: One pilot sequentially controls multiple drones (one-to-many) The swarm consists of three drones: UAV1, UAV2, and UAV3. Pilot A shares the group DSF root key with the three drones before takeoff. Pilot A's terminal derives individual DSFs based on each drone's ID and maintains PSSM synchronization with each of the three drones. During the mission, Pilot A controls the drones sequentially at different logical moments: at time t1, it sends a "forward" command using parameters derived from UAV1; at time t2, it sends a "hover" command using parameters derived from UAV2; and at time t3, it sends a "return" command using parameters derived from UAV3. Each drone only receives commands at its corresponding moment and remains silent at other times. Throughout the entire process, there is no need to establish separate connections with each drone or to negotiate and switch control targets using signaling. Traditional solutions require establishing three independent links, and switching targets necessitates re-handshaking or switching channels.

[0090] Example 5: Pilot-to-Pilot Authentication Switching (Backup Takeover) In high-value missions, pilot B sends a takeover request frame to the drone (containing B's ID, permission parameters, and timestamp, derived using B's individual DSF). The drone verifies that B's ID is in the authorized list, replies with a confirmation frame, and records the handover timestamp. Pilot A stops sending upon receiving confirmation, and pilot B begins sending. After the handover is complete, the drone updates the controller's ID record, and the mission log saves the complete handover record. If pilot A does not actively release the drone, the drone can be configured to allow B to take over only when A loses contact or the signal quality falls below a threshold, preventing unauthorized takeover.

[0091] Example 6: Passive Tracking (Mobile Survivability) During Launch Vehicle Mobility Transfer After missile launch, the launch vehicle immediately maneuvers away at high speed. The launch vehicle only needs to continue transmitting downlink pilot signals according to the PSSM cycle (transmitting empty frames even without user data). The missile passively receives these downlink pilot signals, continuously measuring changes in ε, and calculating the launch vehicle's trajectory and current position in real time. The launch vehicle does not need to send any additional positioning request or response signals, and its maneuvering does not affect the missile's tracking capability. Even in a GPS-denied environment, the missile can still lock onto the launch vehicle's position, achieving high survivability for the launch vehicle.

[0092] Example 7: Large-window sliding synchronization under long-term disconnection (high robustness) A low-Earth orbit satellite IoT terminal, using a standard crystal oscillator (accuracy ±10ppm), lost contact with the ground station for 10 minutes (600 seconds), with a logic counter T_step = 10ms. Residual deviations remained after autonomous drift calibration during the loss of contact. The ground station, knowing the predicted recovery time t_restore, calculated the window half-width K = ceil(10e-6 × 600 / 0.01) + 2 = ceil(0.6) + 2 = 1 + 2 = 3, and the window size 2K+1=7. Within the window, the ground station sequentially used 7 sets of candidate pilots for sliding correlation detection, successfully capturing the correlation peak at i=+2 (logic counter deviation + 2 steps), and simultaneously measured ε=0.5μs. The link was restored after one closed-loop correction, with a total recovery delay of 15ms. If single-point synchronization (without a window) were used, pilot mismatch due to logic counter deviation would likely lead to recovery failure.

[0093] Example 8: Concurrent small data packet transmission based on orthogonal pilots In a typical IoT scenario: 100 sensor nodes share a group DSF root key and a globally unified Rule_ID with a ground station, and each node derives its own individual DSF. Each sensor periodically reports temperature data (12 bytes). At a logical decision time t0, the ground station simultaneously receives concurrent reports from 50 sensors. These sensors use their respective individual DSFs to modulate the data onto a security pilot sequence and transmit it simultaneously on the same time-frequency resources. The ground station uses 50 pre-generated pilot copies locally for parallel matched filtering. All 50 relevant peak values ​​exceed the threshold, successfully demodulating all data packets. The entire process takes one logical cycle (10ms). Without this mechanism, the 50 sensors would need to compete for resources through two arbitration steps, requiring at least dozens of cycles to complete the reporting.

[0094] This invention can be widely applied to scenarios such as drone swarms, missile clusters, satellite communications, and vehicle-to-everything (V2X) communication, and is particularly suitable for military communication systems that are highly dynamic, highly contested, and require low probability of interception. This invention can be seamlessly integrated with the applicant's existing DSF patent system without requiring additional hardware and can be implemented through software upgrades.

[0095] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for rapid synchronization and power recovery after disconnection based on state prediction, characterized in that, include: Pre-synchronization configuration steps: The aircraft and one or more candidate control terminals inject the same dynamic security foundation or group dynamic security foundation root key through a one-time secure channel. The candidate control terminal independently runs the protocol security state machine, and its logical state is synchronized with that of the aircraft. There is no communication association between the candidate control terminal and the aircraft before the switchover. Synchronization maintenance steps during the loss of contact: During the loss of contact, both communicating parties continue to run the protocol security state machine independently, and the logic counter increments according to a predetermined cycle; at the same time, the crystal oscillator drift rate is estimated based on the historical calibration records stored before the loss of contact, the accumulated logic counter deviation during the loss of contact is calculated, and the local logic counter is silently corrected. Recovery preparation steps: The transmitter calculates the receiver's logical decision time, uplink resource intention, downlink listening resources, and security pilot sequence at the recovery time based on the maintained logical state; and predicts the physical anchor point deviation and relative distance after the loss of connection based on the ranging sequence or timing observation error sequence stored before the loss of connection. Power recovery steps: Estimate path loss based on predicted relative distance and set initial transmit power; Recovery signal transmission steps: At the anticipated recovery time, using the anticipated safety pilot sequence and the set initial transmit power, transmit the recovery signal on the anticipated downlink monitoring resources.

2. The method according to claim 1, characterized in that, In the pre-synchronization configuration step, multiple aircraft and multiple control terminals share the same group dynamic security foundation root key and globally unified rule identifier; each node derives its own dynamic security foundation based on its own identifier, wherein the individual dynamic security foundation includes the derived individual key and the individual initial anchor point, while the rule identifier remains globally unified; all nodes independently run the protocol security state machine based on their respective individual dynamic security foundations, and the logical states remain aligned at the same logical moment, but the pilot sequences generated by each node are orthogonal due to different individual keys, and resource intentions may collide due to modulo operations; any control terminal can use the target aircraft's individual dynamic security foundation to generate communication parameters and communicate with the aircraft at any logical decision moment without prior connection establishment or parameter negotiation.

3. The method according to claim 1, characterized in that, The autonomous drift calibration in the synchronization maintenance step during the period of disconnection includes: - Read the historical calibration records stored before the loss of contact, the records including at least one of the following: silent calibration correction sequence, drift trend sequence of same-period difference mode measurement, capture deviation change sequence of downlink pilot measurement mode measurement, and local temperature sensor reading sequence; - Filter the historical drift trend to estimate the long-term drift rate of the crystal oscillator; - Calculate the cumulative counter deviation based on the duration of disconnection and the logical beat cycle; - Perform silent correction of logic counters locally.

4. The method according to claim 3, characterized in that, The autonomous drift calibration also includes temperature compensation: reading the temperature changes during the period of disconnection, calculating the temperature-induced drift contribution based on the pre-stored crystal oscillator temperature characteristic curve, and separating the temperature-induced drift from the total drift or using it as a superposition correction term.

5. The method according to claim 1, characterized in that, It also includes a multi-user resource arbitration step: - The network side predicts the uplink resource intentions of each user device at the recovery time based on the protocol security state machine state of each user device; - Determine whether the uplink resource intentions of multiple user devices are mapped to the same physical resource; - For conflicting physical resources, a winner is deterministically selected from the set of conflicting users according to a preset arbitration strategy; - Query the global resource calendar to check if the winner's intended resources are available at the corresponding time; - For the winning user, send an authorization signal on the downlink listening resources they are aware of; For failed users, no message is sent or a silent instruction is sent.

6. The method according to claim 1, characterized in that, It also includes the control transfer process: After the first control terminal stops transmitting, the second control terminal detects the switching conditions by listening for the disappearance of the uplink signal from the first control terminal or receiving external instructions. Once the conditions are met, it directly starts transmitting uplink signals. The aircraft automatically identifies the control terminal based on the legality of the signal, without the need for explicit takeover signaling. Alternatively: The second control terminal sends a takeover request frame to the aircraft. The takeover request frame contains the second control terminal identifier and authorization parameters. After the aircraft verifies the takeover, it replies with a takeover confirmation frame. The second control terminal receives the confirmation and then begins sending.

7. The method according to claim 1, characterized in that, It also includes a synchronization search window step: the receiving end sets a window half-width K with the predicted recovery logic time as the center, generates a total of 2K+1 candidate pilot sequences corresponding to the center logic counter value before and after, and performs sliding correlation detection within the physical time window; when the correlation peak of any candidate sequence exceeds the threshold and the arrival time deviation is within the preset range, the synchronization is determined to be successful, and the logic counter correction amount and physical anchor point deviation are extracted; the window half-width K is calculated based on the disconnection time and the maximum uncertainty of the crystal oscillator: K = ceil(Δf_max·T_loss / T_step) + δ, where δ is a preset safety margin.

8. The method according to claim 1, characterized in that, It also includes a closed-loop correction step: after receiving the response from the receiver, the transmitter updates the local timing advance, power control reference value and logic counter based on the timing observation error, signal strength indication and counter deviation carried in the response.

9. The method according to claim 1, characterized in that, It also includes concurrent transmission steps based on orthogonal pilots: - The sending end modulates the small data packets to be sent onto a security pilot sequence derived from the current logical state and the individual dynamic security foundation to generate a security pilot frame; - Multiple transmitters simultaneously transmit their respective security pilot frames on the same time-frequency physical resources; - The receiving end uses locally generated pilot copies of each node to perform parallel matched filtering or multi-user detection on the received signal, and determines which nodes' data packets were successfully received based on the relevant peak values ​​and demodulates the data bits.

10. A device for rapid synchronization and power recovery after disconnection based on state prediction, characterized in that, include: - The pre-synchronization configuration module is used to inject a dynamic security foundation into the aircraft and the candidate control terminal through a one-time security channel, and enable the candidate control terminal to run the protocol security state machine independently, keeping its logical state synchronized with the aircraft; - Autonomous drift calibration module, used to estimate crystal oscillator drift rate based on historical calibration records during disconnection, calculate cumulative counter deviation and perform silent correction on local logic counter; - The synchronous prediction module is used to calculate resource information at the recovery time based on the maintained logical state, and predict the physical anchor point deviation and relative distance based on the ranging sequence before the loss of contact; - Power prediction module, used to estimate path loss and set initial transmit power based on predicted distance; - A recovery signal transmission module for transmitting a recovery signal at a set initial power at a predicted recovery time; - Multi-user arbitration module, used to perform resource conflict prediction and two-step deterministic arbitration when multiple users resume simultaneously; - Control transfer module, used to support seamless takeover by the successor control unit without communication association; - The sliding window synchronization module is used to set the synchronization search window when resuming reception and generate multiple sets of candidate pilot sequences for sliding correlation detection; - Orthogonal concurrent transmission module, used to modulate small data packets onto a security pilot sequence and perform multi-user concurrent transmission and reception on the same physical resources.

11. The apparatus according to claim 10, characterized in that, The autonomous drift calibration module includes: a historical record reading unit, a drift rate estimation unit, a deviation calculation unit, a silent correction unit, and a temperature compensation unit.

12. A wireless communication terminal, characterized in that, It includes the state prediction-based rapid synchronization and power recovery device after disconnection as described in claim 10 or 11.

Images