A method for distributed event causal trustworthy transmission, observation and verification
Patent Information
- Application Number
- CN202611199122.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-09
- Publication Date
- 2026-10-16
AI Technical Summary
[0002]在保序传输领域，现有方案主要分为全局时钟同步、混合逻辑时钟、可靠性传输、逻辑时钟四大类：全局时钟同步方案依赖原子钟、GPS或中心化授时服务建立全局时间轴，存在运维成本高、跨地域延迟大、时钟可篡改的缺陷；混合逻辑时钟以物理时钟为逻辑时钟天花板，仍需依赖NTP维持同步误差上界，无法解决单链保序传输的超时裁决与审计污染问题；TCP等可靠性传输协议通过序列号与重传机制保障有序交付，不具备场景化降级与审计特化能力；传统逻辑时钟仅依靠消息传递建立因果偏序，缺少工程化的超时裁决、降级策略与审计合规支持
[0045]1.本发明摆脱了对NTP、中心化授时服务等跨节点时钟同步方案的依赖，仅通过接收端单边本地单调钟维护的MJD窗口完成因果序裁决与有序交付，大幅降低了分布式系统的时钟同步运维成本与硬件投入；同时发送端协议层不携带跨端可比时间戳，从根源上消除了协议层时间篡改的攻击面，配合可配置的超时降级策略与主副双链审计模式，既能适配不同业务的可用性需求，又能保障审计证据不受协议降级策略的污染。
Smart Images
Figure CN122932874A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of distributed system communication technology, observability technology and time tracing technology, and in particular to a method for reliable transmission, observation and verification of causal events in a distributed manner. Background Technology
[0002] In the field of ordered transmission, existing solutions mainly fall into four categories: global clock synchronization, hybrid logical clocks, reliable transmission, and logical clocks. Global clock synchronization solutions rely on atomic clocks, GPS, or centralized time synchronization services to establish a global timeline, resulting in high maintenance costs, large cross-regional latency, and the possibility of clock tampering. Hybrid logical clocks use physical clocks as the ceiling for logical clocks and still rely on NTP to maintain the upper bound of synchronization errors, failing to solve the timeout adjudication and audit pollution problems in single-chain ordered transmission. Reliable transmission protocols such as TCP ensure orderly delivery through sequence numbers and retransmission mechanisms, but lack scenario-based degradation and audit specialization capabilities. Traditional logical clocks rely solely on message passing to establish causal partial order, lacking engineered timeout adjudication, degradation strategies, and audit compliance support. In the field of multi-chain observation and auditing, from OpenTelemetry observability data models, multi-stream sequencing of financial transactions, and industrial control accident liability determination to cross-border audit reconciliation, existing solutions all rely on global wall clock alignment to timestamp events. Cross-node clock drift can lead to timing discrepancies, reconciliation disputes, and difficulties in accident liability determination.
[0003] In the field of time traceability, existing distributed timestamp schemes all suffer from fundamental flaws: NTP timestamps rely on network synchronization, are susceptible to attack surfaces such as drift, rollback, and hijacking, and cannot provide a traceability chain with the International System of Units (SI); satellite time sources such as GPS / BeiDou are inaccessible in indoor or underground environments, posing a single point of failure risk; while Google TrueTime reduces the time uncertainty window through GPS and atomic clocks, its hardware costs are high, and it does not provide an explicit traceability chain with the SI second. Overall, none of the existing schemes utilize the insurmountable upper bound of the speed of light, the speed of causal propagation, nor do they rely on the SI2019 framework established by the 26th General Conference on Weights and Measures—an international convention that all base units are defined based on physical constants—and therefore cannot provide a credible boundary verification and ultimate anchoring at the physical law level for distributed events.
[0004] However, existing technologies still have the following problems:
[0005] 1. How to achieve strictly causal order transmission of point-to-point communication links without relying on cross-node clock synchronization, without acknowledging the upper bound ε of clock synchronization error, and without eliminating the attack surface of protocol layer time tampering, and provide scenario-based degradation mechanisms and audit compliance support.
[0006] 2. In scenarios where multiple senders communicate with the same receiver, how can we provide the receiver with a reliable local timeline that is separate from causal order, without relying on a global wall clock, for the merging of multi-chain events and audit traceability?
[0007] 3. How to utilize the physical constant c (speed of light, SI 2019 fixed value 299792458 m / s) and the instantiation of the International System of Units (SI) second to provide insurmountable physical boundary verification for distributed events—superluminal forgery detection, light cone incomparability decision, and causal irreversible anchoring. Summary of the Invention
[0008] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.
[0009] In view of the problems existing in the above and / or existing methods for trusted transmission, observation and verification of causal events of distributed events, this invention is proposed.
[0010] Therefore, the problem to be solved by this invention is how to address the issue that existing distributed event order-preserving transmission, multi-chain observation, and time tracing all rely on cross-node clock synchronization, which leads to high operation and maintenance costs, the existence of protocol layer time tampering attack surfaces, frequent disputes over multi-stream time sequence reconciliation, and the lack of ultimate trusted anchor points for timestamps. Without acknowledging the upper bound of cross-end clock synchronization error, this invention aims to achieve causal order-preserving transmission and scenario-based degradation of point-to-point links, dual-axis decoupled trusted observation of multi-chain events, and physical boundary-level causal trusted verification based on the SI2019 framework and the speed of light constant.
[0011] To solve the above technical problems, the present invention provides the following technical solution: a method for order-preserving transmission in a point-to-point communication link, comprising a sending end and a receiving end, wherein the sending end assigns a binary identifier (epoch, seq) consisting of a lifecycle version number epoch and a logical sequence number seq to each event to be transmitted, and the binary identifier is sent to the receiving end along with the event payload;
[0012] The epoch is incremented with each restart of the sending end's lifecycle and is persistently stored locally on the sending end;
[0013] The receiver maintains a maximum decision delay (MJD) window, the length of which is independently measured by the receiver's local monotonic clock.
[0014] The receiving end sorts the received events in the MJD window according to the binary identifier (epoch, seq). When there is a continuous (epoch, seq) prefix that matches the expected sequence in the window, the event corresponding to the continuous prefix is delivered to the upper layer application. When the expected missing event in the window stays in the window for a duration that reaches the length of the MJD window, a timeout decision is triggered, and the event sequence is processed according to a preset decision strategy.
[0015] Preferably, the sending end also carries the local monotonic clock reading t_send along with the event, and the t_send is used by the receiving end for differential verification and auditing records.
[0016] Preferably, the adjudication strategy includes at least one of the following:
[0017] Strategy A: The receiving end skips the missing expected events, continues delivery from the subsequent consecutive events at the timeout position, and generates a missing order alarm to notify the upper layer application;
[0018] Strategy B: The receiving end does not skip missing events, the main delivery path remains blocked, subsequent cached events within the window are discarded, and a security alarm is triggered;
[0019] Strategy C: The receiving end skips the missing event and continues delivery in the main delivery path, while delivering the missing event and subsequent sequences to a separate order completion audit port.
[0020] Preferably, the length of the MJD window is set to infinite, so that the receiving end never triggers a timeout decision, all received events are fully recorded in the audit log, and the delivery path is permanently blocked in case of missing events.
[0021] Preferably, the same sending end and the same receiving end maintain two independent communication chains in parallel: a main business chain and an audit sub-chain. The main business chain is configured with a limited-length MJD window and a business-level adjudication strategy, while the audit sub-chain is configured with an unlimited-length MJD window. The two communication chains share the binary identifier (epoch, seq) of the sending end, and the receiving end maintains an independent MJD window respectively.
[0022] A method for observing and outputting multi-chain events, applied to a receiving end that has established communication links with multiple sending ends, includes:
[0023] The receiving end establishes independent communication links with multiple sending ends, and records a local arrival timestamp (arrival_time) for each received event. The arrival_time is generated by reading the local monotonic clock of the receiving end.
[0024] The receiving end maintains a tuple identifier (epoch, seq) assigned to the corresponding sending end for each event;
[0025] The receiving end maintains a causality axis based on the tuple identifier (epoch, seq) of each event and the reference association field between events, which is used to express the causal partial order relationship between events;
[0026] The receiving end maintains a proper time axis based on the arrival_time of each event, which is used to express the arrival sequence of events from the local perspective of the receiving end;
[0027] The causality axis and the proper time axis are stored and recorded independently, and data is provided to upper-layer applications or audit systems through independent interfaces.
[0028] Preferably, comparable events with causal association on the causality axis are marked with corresponding causal relationships; concurrent incomparable events without causal association on the causality axis are clearly marked as concurrent and incomparable when sorted by arrival time on the proper time axis, and are prohibited from being directly used as a causal order.
[0029] A distributed event causality trusted verification method includes configuring a local clock source at the receiving end, and the local clock source is traceable to the definition of the International System of Units second;
[0030] The receiving end records a local arrival timestamp arrival_si for each received event, where the arrival_si is the instantiated reading of the International System of Units second at the receiving end;
[0031] The receiving end preconfigures or obtains in real time the geometric distance lower bound d_min with the transmitting end;
[0032] The receiving end extracts the transmission timestamp t_send recorded by the transmitting end from the event;
[0033] For each event, calculate the physical minimum delay t_min=d_min / c, where c is the speed of light constant defined by the International System of Units, with a value of 299792458m / s;
[0034] If arrival_si-t_send<t_min-δ is satisfied, where δ is the sum of the distance measurement error and the clock traceability error, the event is determined as a physically impossible event, marked as non-compliant, and subjected to abnormal processing.
[0035] Preferably, the method further comprises, for a first event and a second event from different transmitting ends, respectively acquiring the geometric distance lower bounds d_A1B, d_A2B between the corresponding transmitting end and the receiving end, and the arrival timestamps arrival_1, arrival_2 of the two events;
[0036] If |d_A1B-d_A2B|>c*|arrival_1-arrival_2|+δ', where δ' is the comprehensive error margin, then the two events are determined to be physically incomparable, and any total order relation claim provided by a non-protocol causal chain is rejected.
[0037] Preferably, the method is applied to a point-to-point communication link, where the cause-to-effect causal chain established by the binary identifier (epoch, seq) is anchored by the SI tracing sequence of the arrival_si. Any external evidence that claims to reverse the causal order must first overturn the SI tracing chain of the local clock source.
[0038] Preferably, the sending timestamp t_send of the sending end is generated by the sending end's local clock source traceable to the International System of Units (SI) second definition.
[0039] Preferably, the arrival_si is jointly bound and stored with the SI traceability calibration certificate identifier si_cert_id of the local clock source to form an SI-level verifiable record of the event arrival time.
[0040] A transmitting device for implementing the steps executed by the transmitting end in the method includes a processor, a memory, and a network interface. The memory stores a computer program, and when the processor executes instructions, it implements the steps executed by the transmitting end in the method.
[0041] A receiving device for implementing the steps executed by the receiving end in the method is characterized by comprising a processor, a memory, and a network interface, wherein the memory stores a computer program, and when the processor executes the computer program, it implements the steps executed by the receiving end in the method.
[0042] A distributed event communication system includes at least one sending end device and at least one receiving end device.
[0043] A non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method.
[0044] The beneficial effects of this invention are:
[0045] 1. This invention eliminates the reliance on cross-node clock synchronization solutions such as NTP and centralized time synchronization services. It completes causal order determination and ordered delivery solely through the MJD window maintained by the receiver's local monotonic clock, significantly reducing the clock synchronization operation and maintenance costs and hardware investment of distributed systems. At the same time, the sending end protocol layer does not carry cross-end comparable timestamps, eliminating the attack surface of protocol layer time tampering at the root. Combined with configurable timeout degradation strategies and a main-secondary dual-chain audit mode, it can adapt to the availability requirements of different businesses and ensure that audit evidence is not contaminated by protocol degradation strategies.
[0046] 2. This invention, through the dual-axis separation of the causal axis and the original time axis, preserves the strict causal order of events within the chain and related events to support the execution of business logic, and provides a unified arrival time sequence from the local perspective of the receiving end to support multi-stream convergence observation and audit traceability, without relying on global wall clock alignment.
[0047] Meanwhile, it clearly marks concurrent events that are causally incomparable, thus avoiding the logical error of misusing arrival order as causal order from a mechanism perspective. This can effectively solve timing disputes caused by cross-node clock drift in scenarios such as financial transaction reconciliation, industrial control accident liability determination, and cross-border judicial audit.
[0048] 3. This invention uses the SI 2019 International System of Units (SI) framework as the ultimate anchor point, tracing the local clock back to the standard second defined by the transition frequency of cesium atoms. Combined with the physical insurmountable property of the speed of light constant, it achieves accurate detection of superluminal forgery events, physical-level judgment of the incomparability of light cones in cross-chain events, and SI tracing and anchoring of causal order. This elevates the credibility of distributed events from the protocol agreement level to the physical law level, providing time and causal evidence with ultimate credibility for scenarios with high compliance and high security requirements. Attached Figure Description
[0049] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0050] Figure 1 This is a flowchart of the transmission process in Example 1.
[0051] Figure 2 This is a diagram of the MJD window state machine in Example 1.
[0052] Figure 3 This is a branch diagram of the adjudication strategy in Example 2.
[0053] Figure 4 This is a flowchart of the dual-chain companion audit mode in Example 3.
[0054] Figure 5 This is a schematic diagram of the multi-chain observation dual-axis output of Example 5.
[0055] Figure 6 This is a schematic diagram of the SI traceability clock and the lower bound of the speed of light verification in Example 6.
[0056] Figure 7 This is a schematic diagram of the incomparability decision of the optical cone in Example 6. Detailed Implementation
[0057] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.
[0058] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.
[0059] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.
[0060] Example 1, such as Figure 1 and Figure 2 As shown, this is the first embodiment of the present invention, a method for pre-ordered transmission in a point-to-point communication link, including a sender and a receiver. The sender assigns a binary identifier (epoch, seq) consisting of a lifecycle version number epoch and a logical sequence number seq to each event to be transmitted. The binary identifier is sent to the receiver along with the event payload.
[0061] Among them, the epoch is incremented with the life cycle of the sending end and persistently stored locally on the sending end. The protocol layer of the sending end does not carry cross-end comparable clock readings or timestamps.
[0062] The receiver maintains a maximum decision delay (MJD) window. The length of the MJD window is independently measured by the receiver's local monotonic clock, without relying on clock synchronization between the transmitter and receiver, or on any upper bound ε of clock synchronization error.
[0063] The reasoning section that does not acknowledge ε: The MJD window length of this technical solution is measured by the receiver's unilateral local clock. The setting of the MJD length only needs to satisfy: MJD ≥ receiver local clock duration corresponding to the maximum out-of-order depth of this chain network + receiver local clock drift margin.
[0064] Since this length is measured by the local clock on one side of the receiver and is unrelated to the clock on the transmitter, the clock difference between the two ends, and the NTP accuracy, it is neither necessary nor possible to accept any upper bound ε of cross-end clock synchronization error. Therefore, those skilled in the art can set MJD according to the characteristics of this chain network without the participation of ε. This point constitutes an architectural divergence from HLC's |hlc.l-pt|≤ε, which is not a parametric optimization.
[0065] The receiving end sorts the received events according to the tuple identifier (epoch, seq) within the MJD window. When there is a consecutive (epoch, seq) prefix that matches the expected sequence within the window, the event corresponding to the consecutive prefix is delivered to the upper layer application. When the expected missing event in the window stays in the window for a duration that reaches the length of the MJD window, a timeout decision is triggered, and the event sequence is processed according to the preset decision strategy.
[0066] Specifically, sender A and receiver B establish a communication link.
[0067] A maintains (epoch=1, seq=0), increments seq for each event, and sends a packet (1, seq, payload) to B. B maintains an MJD window Δ=5s (measured by B's local monotonic clock, such as LinuxCLOCK_MONOTONIC), expected=(1,1).
[0068] B receives (1,1) → (1,1) is a continuous prefix in the window, deliver, expected jumps to (1,2);
[0069] B receives (1,3) → inserts (1,3) into the window, and waits for (1,2);
[0070] B receives (1,2) → (1,2) and (1,3) are consecutive in the window, deliver (1,2), (1,3), expected jumps to (1,4);
[0071] If (1,2) remains in the window for more than 5 seconds (B local clock reading difference) and the timeout is not reached, a timeout decision will be triggered, and the downgrade will be performed according to the configuration.
[0072] The pseudocode steps for the MJD window are as follows:
[0073] on_receive(packet):
[0074] window.insert(packet.epoch,packet.seq,packet.payload,
[0075] arrival_time=B.local_mono())
[0076] #Delivery consecutive prefix
[0077] whilewindow.has_consecutive_prefix(expected.epoch,expected.seq):
[0078] deliver(window.pop_prefix())
[0079] expected.seq+=1
[0080] #Timeout Check
[0081] for(ep,sq)inwindow.pending(expected.epoch,expected.seq):
[0082] ifB.local_mono()-window.arrival(ep,sq)>=MJD_length:
[0083] trigger_degradation(ep,sq)
[0084] It should be noted that the MJD window length is set to infinity, ensuring that the receiving end never triggers a timeout decision. All received events are fully recorded in the audit log, and the delivery path is permanently blocked in case of missing events. This is the MJD=∞ audit mode.
[0085] When receiver B establishes a link with sender A, MJD on B's side is ∞ (infinite). Expected events do not skip, while out-of-order events cause a permanent deadlock.
[0086] A sends (epoch=1, seq=1..10), B receives 1, 2, 3, 4, 5, 7, 8, 9, 10, missing 6;
[0087] In window B, 6 never arrives → the chain is stuck, and an alarm is issued to the operator.
[0088] All packets are entered into the audit log, which contains (epoch, seq, arrival_time, payload_hash);
[0089] During the dispute, the auditors traced back the chain log and found that "seq=6 was never reached," indicating that the protocol layer did not perform any automatic skipping or downgrading, and the evidence was clean.
[0090] It should be noted that the sending end also carries the local monotonic clock reading t_send along with the event. t_send is used by the receiving end for differential verification and auditing records.
[0091] Specifically, in addition to (epoch, seq, payload), sender A can optionally carry the local monotonic clock reading t_send.
[0092] Receiver B records t_send in the audit log and performs the following differential review (without entering the adjudication layer):
[0093] Monotonicity: If t_send is not decreasing within the same epoch, then mark clock_anomaly as t_send_non_monotonic;
[0094] Logical consistency: Δt_send is normal when seq moves forward → otherwise, mark logic_mismatch;
[0095] Drift warning: Long-term Δt_send / Δarrival slope deviation → clock_drift_warning;
[0096] Partition recovery identification: t_send value range dispersed, arrival value range squeezed → mark batch_recovery.
[0097] Example 2, as Figure 3 As shown, this is the second embodiment of the present invention, which is based on the previous embodiment. The adjudication strategy includes at least one of the following:
[0098] Strategy A: The receiving end skips the missing expected events, continues delivery from the subsequent consecutive events at the timeout position, and generates a missing order alarm to notify the upper layer application;
[0099] Specifically, receiver B is configured with degradation strategy A: when a timeout occurs, the expected path jumps to the timeout sequence, the main path continues delivery, and the gap is transparent to the upper layer.
[0100] B receives (1,1) and (1,3) from A, and waits for (1,2) to reach MJD=5s timeout;
[0101] B triggers a downgrade to A: expected jumps to (1,3), deliver(1,3), and alerts missing_seq=(1,2) to notify the upper layer;
[0102] The upper level handles the gap automatically after receiving the alarm;
[0103] If (1,2) arrives late, →B will mark it as late_arrival and add it to the audit log.
[0104] Strategy B: The receiving end does not skip missing events, the main delivery path remains blocked, subsequent cached events within the window are discarded, and a security alert is triggered;
[0105] Specifically, receiver B is configured with a degradation strategy B: when timeout occurs, the expected path will not jump, the main path will be blocked, all packets within the window will be lost, and the system will enter a safe state.
[0106] B receives (1,1) and (1,3) from A, and waits for (1,2) to reach MJD=5s timeout;
[0107] B triggers a downgrade: expected remains (1,2), discards (1,3) and subsequent window packets, enters safe state, and triggers an alarm;
[0108] All data is entered into the audit log, including discarded packets;
[0109] After the operations and maintenance team intervenes and investigates, manually RESET the chain or reset the expected chain.
[0110] Strategy C: The receiving end skips the missing event and continues delivery on the main delivery path, while delivering the missing event and subsequent sequences to a separate completion audit port.
[0111] Example 3, as Figure 4 As shown, this is the third embodiment of the present invention. This embodiment is based on the first two embodiments. Two independent communication chains, the main business chain and the audit sub-chain, are maintained in parallel between the same sending end and the same receiving end. The main business chain is configured with a limited-length MJD window and a business-level adjudication strategy, while the audit sub-chain is configured with an unlimited-length MJD window. The two communication chains share the binary identifier (epoch, seq) of the sending end, and the receiving end maintains an independent MJD window respectively.
[0112] Specifically, two chains are maintained in parallel between the same pair of endpoints:
[0113] Main chain: MJD=100ms, downgrade strategy A (continue + alarm), business traffic goes through the main chain;
[0114] Audit sub-chain: MJD=∞, one mirror of the same batch of business packages (or only header + payload_hash), full log, never downgraded;
[0115] The two chains share the sending end (epoch, seq) space, while the receiving end has two independent windows;
[0116] The auditors only look at the secondary chain logs to ensure that the evidence is not contaminated by the main chain downgrade strategy.
[0117] Example 4 is the fourth embodiment of the present invention. This embodiment is based on the previous two embodiments. This embodiment corresponds to the engineering-oriented concrete expansion of the lifecycle version number epoch. Specifically, the sending end A persists the epoch in local storage (such as a file or KV), and the initial epoch=1.
[0118] A is running normally, sending packets (1, seq=1..1000);
[0119] A crashes and restarts, reads epoch=1 from the persistent storage, increments it by 1, then epoch=2, and after fsync, starts sending (2, seq=1..);
[0120] B receives (2,1), finds that epoch=2>expected.epoch=1→determines that the old epoch packet has expired, resets expected=(2,2), and continues delivery;
[0121] A late packet of the old epoch (e.g., (1,900) arrives later)→is discarded because epoch=1<expected.epoch=2.
[0122] Embodiment 5, such as Figure 5 shown in the figure, which is the fifth embodiment of the present invention. This embodiment is based on Embodiment 1 to Embodiment 4, and relates to a method for observing and outputting multi-chain events, applied to a receiver that separately establishes communication links with a plurality of transmitters, comprising:
[0123] The receiver separately establishes independent communication links with a plurality of transmitters, and records a local arrival timestamp arrival_time for each received event, where arrival_time is generated by reading from the receiver's local monotonic clock;
[0124] The receiver maintains the tuple identifier (epoch, seq) assigned by the corresponding transmitter for each event;
[0125] The receiver maintains a causality axis based on the tuple identifier (epoch, seq) of each event and the reference association field between events, for expressing the causal partial order relationship between events;
[0126] The receiver maintains a proper time axis based on the arrival_time of each event, for expressing the arrival order of events from the local perspective of the receiver;
[0127] The causality axis and the proper time axis are stored and recorded independently, and data is provided to an upper-layer application or an audit system through an independent interface.
[0128] Comparable events with causal association on the causality axis are marked with corresponding causal relationships; concurrent incomparable events without causal association on the causality axis are explicitly marked as concurrent and incomparable when sorted by arrival time on the proper time axis, and are prohibited from being directly used as a causal order.
[0129] Specifically, receiver B separately establishes independent communication links with a plurality of transmitters A1, A2, and A3.
[0130] B records arrival_time for each event;
[0131] B maintains the causality axis based on (epoch, seq), and maintains the proper time axis based on arrival_time;
[0132] Events that are comparable on the causality axis are marked with causal relationships, and events that are incomparable are marked as "concurrent and incomparable" on the proper time axis;
[0133] The causal axis is used for business logic, and the physical time axis is used for observation and auditing.
[0134] Embodiment 6, as Figure 6 shown, is the sixth embodiment of the present invention. This embodiment is based on the previous embodiment, and relates to a distributed event causal credible verification method, which includes configuring a local clock source at the receiving end, and the local clock source is traceable to the definition of the International System of Units second;
[0135] The receiving end records a local arrival timestamp arrival_si for each received event, where arrival_si is the instantiated reading of the International System of Units second at the receiving end;
[0136] The receiving end preconfigures or acquires in real time the geometric distance lower bound d_min with the transmitting end;
[0137] The receiving end extracts the sending timestamp t_send recorded by the transmitting end from the event;
[0138] For each event, calculate the physical minimum delay t_min=d_min / c, where c is the speed of light constant defined by the International System of Units, with a value of 299792458m / s;
[0139] If arrival_si-t_send<t_min-δ is satisfied, where δ is the sum of the distance measurement error and the clock traceability error, the event is determined as a physically impossible event, marked as non-compliant and subjected to abnormal processing.
[0140] Moreover, the sending timestamp t_send of the transmitting end is generated by the local clock source of the transmitting end that is traceable to the definition of the International System of Units second.
[0141] Specifically, the distance between B and A is d=1000km. The local clock on the B side is a Rb atomic clock, which is traceable to the SI second.
[0142] t_min=d / c=10 6 / 299792458≈3.335ms;
[0143] The calibration method of δ: The distance measurement error δ_d depends on the acquisition method of d_min — if d_min is obtained through optical fiber routing engineering ranging, δ_d can take the uncertainty of the optical fiber routing (typical value 0.1~0.5ms); if d_min is obtained through GPS / Beidou ranging, δ_d can take the satellite ranging uncertainty (typical value 10~100ns). The clock traceability error δ_t depends on the type of the local clock source and the calibration cycle — the annual drift of the Rb atomic clock is less than 1μs, the annual drift of the Cs atomic clock is less than 0.1μs, and δ_t within the calibration cycle can take the upper limit of the drift. δ=δ_d+δ_t, which can be calibrated by those skilled in the art according to engineering practices.
[0144] In this example, δ_d = 0.5ms (fiber optic routing), δ_t = 0.1ms (Rb clock annual drift is 1μs, calculated according to the calibration period), and δ = 0.6ms;
[0145] Legal lower bound: t_min-δ=2.735ms.
[0146] B receives a packet from A (epoch=1, seq=42, t_send_si=1000000ns, payload), arrival_si=1003335ns:
[0147] arrival_si-t_send_si=3.335ms≥2.735ms→Valid.
[0148] If the attacker forges t_send_si to be 1002000ns:
[0149] arrival_si-t_send_si_fake=1.335ms<2.735ms→mark si_violation:superluminal, rejected.
[0150] In addition, such as Figure 7 As shown, the distributed event causal trusted verification method also includes obtaining the lower bounds of the geometric distance d_A1B and d_A2B between the corresponding sending end and receiving end for the first event and the second event from different sending ends, as well as the arrival timestamps arrival_1 and arrival_2 of the two events respectively.
[0151] If |d_A1B-d_A2B|>c*|arrival_1-arrival_2|+δ', where δ' is the comprehensive error margin, then the two events are determined to be physically incomparable, and any total order relation claim provided by a non-protocol causal chain is rejected.
[0152] Specifically, B simultaneously receives events A1 (d1≈0km) and A2 (d2=11000km).
[0153] B receives A1: arrival1 = t0 + 1μs;
[0154] B receives A2: arrival2=t0+36.71ms.
[0155] |d1-d2|=11000km, c·|arrival2-arrival1|≈11000km→comparable.
[0156] If A2's arrival2 is forged as t0+35.00ms:
[0157] c·|arrival2_fake-arrival1|≈10492km<11000km+δ'→ Determines physical incomparability, marked si_violation:spacelike_fabricated.
[0158] Furthermore, the distributed event causal trust verification method is applied to the point-to-point communication link in Example 1. The cause-to-effect causal chain established by the binary identifier (epoch, seq) within the chain is anchored by the SI tracing sequence of arrival_si. Any external evidence that claims to reverse the causal order must first overturn the SI tracing chain of the local clock source.
[0159] arrival_si is bound and stored together with the SI traceability calibration certificate identifier si_cert_id of the local clock source, forming an SI-level verifiable record of the event arrival time.
[0160] Specifically, the intra-chain causal relationship between B and A is seq=5→seq=6.
[0161] Audit log for side B:
[0162] (epoch=1,seq=5,arrival_si=t+1000ms,causal_ref=none,si_cert_id=Rb20260101);
[0163] (epoch=1,seq=6,arrival_si=t+1050ms,causal_ref=(1,5),si_cert_id=Rb20260101).
[0164] During the dispute, A claimed that "seq=6 precedes seq=5";
[0165] Party B's evidence: The arrival_si sequence (SI tracing) shows that seq=6 is later than seq=550ms, and the causal axis ref=(1,5) indicates that seq=6 depends on seq=5;
[0166] If A is to be overturned, the SI traceability chain of the Rb clock on side B must first be overturned.
[0167] Example 7 is the seventh embodiment of the present invention. This embodiment is based on Examples 1 to 6. A transmitting end device implements the steps executed by the transmitting end, including a processor, a memory and a network interface. The memory stores a computer program. When the processor executes instructions, it implements the steps executed by the transmitting end in the method.
[0168] Example 8 is the eighth embodiment of the present invention. This embodiment is based on Examples 1 to 6 and implements the steps executed by the receiving end, including a processor, a memory and a network interface. The memory stores a computer program. When the processor executes the computer program, it implements the steps executed by the receiving end in the method.
[0169] Example 9 is the ninth embodiment of the present invention. This embodiment is based on Examples 1 to 6. It is a distributed event communication system, including at least one transmitting end device as claimed in claim 13 and at least one receiving end device as claimed in claim 14.
[0170] Example 10 is the tenth embodiment of the present invention. This embodiment is based on Examples 1 to 6. A non-transitory computer-readable storage medium is provided on which a computer program is stored. When the computer program is executed by a processor, it implements the method of any one of Examples 1 to 6.
[0171] Example 11 is the eleventh embodiment of the present invention. This embodiment is an optional extension of the present invention, and those skilled in the art can implement it based on the core principles described in the foregoing embodiments. The following are descriptions of each embodiment:
[0172] The multi-chain observation method can be applied to multiple heterogeneous data streams, such as the ordering of order streams, market data streams, and risk control streams, or distributed observable data model scenarios, or industrial control multi-sensor convergence scenarios, or cross-border multi-jurisdictional reconciliation scenarios, or blockchain peer-to-peer transaction broadcast scenarios, where the rule of separating the causal axis from the original time axis remains unchanged.
[0173] In the verification method based on SI and physical constants, `arrival_si` and `si_cert_id` can constitute an SI-level verifiable record of the event arrival time, allowing third parties to verify the physical consistency of the event arrival time. The payload can include the measured value `v` and the calibration certificate identifier `cal_cert_id`, where the unit value of `v` is traceable to at least one of the seven fundamental constants of the International System of Units (SI). The transmitting end can be a quantum measurement device, whose emitted events include quantum measurement results. The `arrival_si` record serves as the physical entry anchor point for the quantum event to enter the classical causal network.
[0174] The method may also include a third-party verification node, which maintains a local clock source and traces back to the definition of the International System of Units (SI) second, receives the hash or metadata of events, and forms a multi-party verifiable audit record. It may also include a second receiver, which performs event consistency reconciliation with the first receiver based on the same SI traceability chain. Events may also be accompanied by a hardware-unclonable fingerprint based on the elementary charge e, forming a dual-anchored evidence with arrival_si. The payload may also contain a temperature measurement traced back to the Boltzmann constant k_B, output coaxially with the causal order.
[0175] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.
Claims
1. A method for order-preserving transmission in a point-to-point communication link, characterized in that: It includes a sending end and a receiving end. The sending end assigns a binary identifier (epoch, seq) consisting of a lifecycle version number epoch and a logical sequence number seq to each event to be transmitted. The binary identifier is sent to the receiving end along with the event payload. The epoch is incremented with each restart of the sending end's lifecycle and is persistently stored locally on the sending end; The receiver maintains a maximum decision delay (MJD) window, the length of which is independently measured by the receiver's local monotonic clock. The receiving end sorts the received events in the MJD window according to the binary identifier (epoch, seq). When there is a continuous (epoch, seq) prefix that matches the expected sequence in the window, the event corresponding to the continuous prefix is delivered to the upper layer application. When the expected missing event in the window stays in the window for a duration that reaches the length of the MJD window, a timeout decision is triggered, and the event sequence is processed according to a preset decision strategy.
2. The method for preserving order transmission in a point-to-point communication link as described in claim 1, characterized in that: The sending end also carries the local monotonic clock reading t_send along with the event, and the t_send is used by the receiving end for differential verification and auditing records.
3. The method for preserving order transmission in a point-to-point communication link as described in claim 1, characterized in that: The adjudication strategy includes at least one of the following: Strategy A: The receiving end skips the missing expected events, continues delivery from the subsequent consecutive events at the timeout position, and generates a missing order alarm to notify the upper layer application; Strategy B: The receiving end does not skip missing events, the main delivery path remains blocked, subsequent cached events within the window are discarded, and a security alarm is triggered; Strategy C: The receiving end skips the missing event and continues delivery in the main delivery path, while delivering the missing event and subsequent sequences to a separate order completion audit port.
4. The method for preserving order transmission in a point-to-point communication link as described in claim 1, characterized in that: The length of the MJD window is set to infinite so that the receiving end never triggers a timeout decision, all received events are fully recorded in the audit log, and the delivery path is permanently blocked in case of missing events.
5. The method for preserving order transmission in a point-to-point communication link as described in claim 1, characterized in that: Two independent communication chains, a main business chain and an audit sub-chain, are maintained in parallel between the same sending end and the same receiving end. The main business chain is configured with a limited-length MJD window and a business-level adjudication strategy, while the audit sub-chain is configured with an unlimited-length MJD window. The two communication chains share the binary identifier (epoch, seq) of the sending end, and the receiving end maintains an independent MJD window respectively.
6. A method for observing and outputting multi-chain events, applied to a receiving end that has established communication links with multiple sending ends, characterized in that, include: The receiving end establishes independent communication links with multiple sending ends, and records a local arrival timestamp (arrival_time) for each received event. The arrival_time is generated by reading the local monotonic clock of the receiving end. The receiving end maintains a tuple identifier (epoch, seq) assigned to the corresponding sending end for each event; The receiving end maintains a causal axis based on the binary identifier (epoch, seq) of each event and the reference association field between events, which is used to express the causal partial order relationship between events; The receiving end maintains the original timeline based on the arrival_time of each event, which is used to express the arrival order of events from the local perspective of the receiving end; The causal axis and the original time axis are stored and recorded independently, and data is provided to upper-layer applications or audit systems through independent interfaces.
7. The method for observing and outputting multi-chain events as described in claim 6, characterized in that: Comparable events with causal association on the causal axis are labeled with the corresponding causal relationship; concurrent incomparable events without causal association on the causal axis are explicitly marked as concurrent and incomparable when sorted by arrival time on the original time axis, and are prohibited from being directly used as a causal order.
8. A distributed event causal trust verification method, characterized in that, Comprising configuring a local clock source at the receiving end, wherein the local clock source is traceable to the definition of the International System of Units second; The receiving end records a local arrival timestamp arrival_si for each received event, where arrival_si is the instantiated reading of the International System of Units second at the receiving end; The receiving end preconfigures or acquires in real time the geometric distance lower bound d_min with the transmitting end; The receiving end extracts the transmission timestamp t_send recorded by the transmitting end from the event; For each event, calculate the physical minimum delay t_min=d_min / c, where c is the speed of light constant defined by the International System of Units, with a value of 299792458 m / s; If arrival_si-t_send<t_min-δ is satisfied, where δ is the sum of the distance measurement error and the clock traceability error, the event is determined as a physically impossible event, marked as non-compliant, and subjected to exception handling.
9. The distributed event causal trust verification method as described in claim 8, characterized in that: The method further comprises, for a first event and a second event from different transmitting ends, respectively acquiring the geometric distance lower bounds d_A1B and d_A2B between the corresponding transmitting ends and the receiving end, as well as the arrival timestamps arrival_1 and arrival_2 of the two events; If |d_A1B-d_A2B|>c*|arrival_1-arrival_2|+δ' is satisfied, where δ' is the comprehensive error margin, the two events are determined as physically incomparable events, and any total order claim provided by a non-protocol causal chain is rejected.
10. The distributed event causal trust verification method as described in claim 8, wherein the method is applied to the point-to-point communication link as described in claim 1, characterized in that: The causal chain from cause to effect established by the binary identifier (epoch, seq) within the chain is anchored by the SI traceability sequence of arrival_si, and any external evidence claiming to reverse the causal order must first overturn the SI traceability chain of the local clock source.
11. The distributed event causal trust verification method according to claim 8, characterized in that, The transmission timestamp t_send of the transmitting end is generated by the local clock source of the transmitting end that is traceable to the definition of the International System of Units second.
12. The distributed event causal trust verification method according to claim 10, characterized in that, The arrival_si and the SI traceability calibration certificate identifier si_cert_id of the local clock source are bound and stored together to form an SI-level verifiable record of the event arrival time.
13. A transmitting end device, implementing the steps performed by the transmitting end in the method according to any one of claims 1 to 5, characterized in that, Comprising a processor, a memory and a network interface, wherein the memory stores a computer program, and when the processor executes instructions, the steps performed by the transmitting end in the method are implemented.
14. A receiving end device that implements the steps performed by the receiving end in the method according to any one of claims 1 to 12, characterized in that, Comprising a processor, a memory and a network interface, wherein the memory stores a computer program, and when the processor executes the computer program, the steps performed by the receiving end in the method are implemented.
15. A distributed event communication system, characterized in that, Comprising at least one transmitting end apparatus as claimed in claim 13, and at least one receiving end apparatus as claimed in claim 14.
16. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the method according to any one of claims 1 to 12.