A method and system for generating a perceptual pilot based on cryptographic primitives for integrated sensing and communication deterministic sensing

Abstract

The application discloses a kind of based on cryptography primitive generation perception pilot integrated sensing algorithm integration deterministic perception method and system, belong to 6G integrated sensing algorithm integration technical field.The application makes both sides of communication based on shared cryptography material and time-varying synchronization parameter, same perception pilot is independently generated by cryptography deterministic function, and multi-node zero signaling cooperative perception is realized.Synchronization parameter can use physical layer broadcast timing, the logical state of protocol security state machine output or unified anchor point time T_anchor;Wherein rule D directly uses T_anchor As state value, can generate various communication and perception parameters.The application provides three kinds of implementation paths: path A is compatible with 5G traditional architecture, path B introduces DSF and PSSM, path C uses LPC+PAN decoupling architecture, and introduces two layers of decoupling and centralized clock bias management, and physical layer uncertainty is transparent to node.Through two-step deterministic arbitration and downlink implicit authorization mechanism, multi-node resource conflict is solved in order.The application completely eliminates pilot signaling overhead, reduces external synchronization dependence, tolerates the inconsistency between nodes Physical parameters, applicable to ground network, non-ground network and hybrid networking scene, provides end-to-end deterministic guarantee for integrated sensing algorithm integration.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, specifically relating to Integrated Sensing and Communication (ISAC) technology in sixth-generation (6G) mobile communication systems. In particular, it is a method and system for generating sensing pilot sequences using cryptographic primitives to achieve multi-node zero-signaling cooperative sensing, covering implementation paths for various network architectures, and is especially suitable for dynamic topology scenarios such as non-terrestrial networks (NTN). Background Technology

[0002] The integration of sensing, communication, and computing is a core technology direction of 6G, aiming to achieve data transmission and environmental perception simultaneously using wireless signals. In a sensing, communication, and computing system, multiple network nodes (such as base stations and dedicated sensing nodes) need to cooperate to detect targets. Typically, a known pilot signal is transmitted as the detection waveform. The receiving node uses a copy of this pilot signal for matched filtering to extract information such as the time, angle, and Doppler of the target's reflected echo.

[0003] However, existing inductive computing schemes face the following core challenges in pilot generation and distribution: 1. High pilot signaling overhead: In traditional solutions, the pilot sequence used by the transmitting node needs to be communicated to all cooperating receiving nodes in advance via signaling. In dynamic networks or large-scale cooperative scenarios, pilot configuration signaling consumes a large amount of air interface resources, increasing latency and energy consumption.

[0004] 2. Difficulty in multi-node synchronization: Coherent sensing requires the receiving node to know precisely the transmission time and pilot content of the transmitting node. Existing solutions rely on high-precision external synchronization protocols such as GPS or IEEE 1588, which are difficult to guarantee in indoor, underground, or denial-of-service environments.

[0005] 3. Security and privacy risks: If the pilot sequence is fixed or predictable, it is easily forged or interfered with by malicious nodes, resulting in unreliable perception results.

[0006] 4. Poor adaptability to network topology changes: In dynamic scenarios such as non-terrestrial networks (NTN), satellite movement and power supply link switching lead to changes in service nodes. Traditional architectures require frequent reconfiguration of pilots, affecting the continuity of perception.

[0007] 5. Physical layer uncertainties affect sensing accuracy: Physical layer factors such as Doppler frequency offset, channel time variation, and clock drift can disrupt the coherence of matched filtering, leading to a decrease in sensing performance. Traditional solutions require nodes to estimate and compensate themselves, increasing complexity and power consumption, and making it difficult to guarantee real-time performance.

[0008] 6. Inconsistent physical layer parameters between nodes lead to cooperation failure: When the local clocks of cooperating nodes deviate or use different physical layer configurations, even if the pilot replicas are consistent, misalignment of transmission and reception times can cause sensing failure. Existing solutions lack an effective management mechanism for such deviations.

[0009] To address the aforementioned issues, this invention proposes a sensing pilot generation method based on cryptographic primitives. This method enables cooperating nodes to independently generate identical pilot copies locally, completely eliminating pilot signaling overhead. Simultaneously, cryptographic mechanisms ensure the uniqueness, randomness, and security of the pilots. Building upon this foundation, the invention further provides multiple implementation paths adaptable to different network architectures and introduces a two-layer decoupled architecture to make physical layer uncertainties transparent to sensing nodes. Specifically, a centralized clock skew management mechanism is designed to address the issue of inconsistent physical parameters between nodes, providing end-to-end deterministic guarantees for the integration of sensing, computing, and communication technologies, from logical resources to physical channels. Summary of the Invention

[0010] Core concept definition To facilitate understanding of this invention, the following core concepts are first defined. Some of these concepts are derived from the applicant's previous series of patents, while others are newly added to this invention.

[0011] Concept 1: Dynamic Security Foundation (DSF) A shared cryptographically secure triple `(K_sec, Init_Anchor, Rule_ID)` is used for generating all deterministic parameters and synchronizing states. `K_sec` is the shared key, `Init_Anchor` defines the starting point and initial state of the logical timeline, and `Rule_ID` identifies the evolution rule of the protocol's secure state machine. The DSF is the security foundation and synchronization basis of this invention. The possible values ​​for `Rule_ID` include, but are not limited to: - Rule A: Broadcast clock driven - Rule B: Hash Chain Driven - Rule C: Logical Epoch Driven - Rule D: Physical anchor-driven (state `S(t) = T_anchor`) (New in this invention) Concept 2: Protocol Security State Machine (PSSM) A deterministic state machine based on the DSF evolves independently between the communicating parties, outputting a time-varying logic state `S(t)`. The evolution of the PSSM is driven by the logical tick defined by `Rule_ID`, independent of the physical clock. As long as the communicating parties share the same DSF, their PSSMs remain synchronized on the logical timeline.

[0012] Concept 3: Logical Decision Instant Discrete moments on the logical timeline defined by `Rule_ID`. At each logical decision moment, the PSSM outputs the current state `S(t)` and triggers physical layer operations. The logical decision moments of different user devices are independent and asynchronous, but are all mapped to a unified physical timeline for arbitration.

[0013] Concept 4: Unified Anchor Time (T_anchor) At the logical decision moment, the node instantaneously reads the value obtained from its local physical clock, denoted as `T_anchor`. This operation uniquely and instantaneously binds the abstract, physical time-independent logical moment to a specific physical time point, providing a unified and computable physical time reference for all subsequent physical layer operations based on this logical moment (such as signal transmission and reception). Although `T_anchor` is based on the local physical clock and may have single-shot errors, these errors only affect the absolute time of this transmission and will not accumulate or propagate. The network side can absorb these single-shot errors by adjusting the timing advance or receive window in the dynamic parameter set.

[0014] Explanation regarding `T_anchor` as a time-varying synchronization parameter: It should be noted that the unified anchor time `T_anchor` can not only be used to determine the transmission and reception times, but it can also be used as a form of the time-varying synchronization parameter `Sync(t)` to directly participate in the generation of sensing pilots: ``` Pilot_sensing(t) = F( K_sec, T_anchor, Context ) ``` This design has unique value: - Physical randomness injection: `T_anchor` contains the instantaneous reading of the local physical clock. Even if the logic state `S(t)` of different nodes is the same, their `T_anchor` will differ due to the small deviation of the physical clock, thus injecting additional randomness into the pilot and further enhancing security and collision resistance. - Timing and content binding: Binding "when to send" with "what to send" through cryptographic functions makes the pilot sequence strongly correlated with the specific sending time, and attackers cannot forge the sensing signal by replaying old pilots; - Simplified design: In some complexity-sensitive scenarios (such as low-cost IoT devices), the logical state S(t) output by PSSM can be directly replaced by T_anchor, reducing the maintenance overhead of the state machine.

[0015] Therefore, within the framework of this invention, `T_anchor` serves as both a bridge connecting logical time and physical time, and can also be used as an independent time-varying synchronization parameter, providing richer design dimensions for the integrated sensing and computing system.

[0016] Concept 5: Protocol Security State Machine Rule D – Physical Anchor Point Driven Rule This rule directly uses the unified anchor time `T_anchor` as the output state `S(t)` of the protocol's security state machine, representing a novel state machine evolution paradigm. - State Definition and Evolution: At each logical decision moment, the node instantaneously samples the local physical clock and uses the sampled value `T_anchor` as the current state `S(t)`. The "evolution" of the state is driven by the periodic arrival of the logical decision moment, but the state value itself is determined by the instantaneous reading of the physical clock, which has inherent randomness and unpredictability.

[0017] - Consistency Guarantee: Both communicating parties define the starting reference for the first logical decision moment (either absolute or relative anchoring) through a shared `Init_Anchor`. Subsequently, both parties independently arrive at their respective logical decision moments based on the same logical clock tick (defined by `Rule_ID`) and independently sample their local physical clocks to obtain `T_anchor`. Although there may be slight differences between the two parties' `T_anchor`s (caused by physical clock deviations), these differences are encapsulated in a single state value and do not accumulate or propagate. For scenarios requiring strict consistency (such as resource conflict detection), the network side can absorb these differences through centralized calibration or receive window adjustment.

[0018] - Rule_ID characteristics: The state update tick of rule D is defined by the logical tick T_logical (e.g., 0.125ms, 1ms, 10ms, etc.), decoupled from the instantaneous value of the physical clock. Its Init_Anchor can be defined as an absolute physical time or a relative offset. The state value S(t) = T_anchor can be directly used as a time-varying synchronization parameter to participate in the generation of various parameters, following the unified parameter generation formula of the DSF paradigm: text Param = F(K_sec, S(t), Context) Param can be any communication or sensing parameter that requires zero-signaling coordination, including but not limited to: Sensing pilot sequences and communication pilot sequences; Logical resource index and physical resource location; HARQ process number, beam identifier, slot format indicator, and other higher-level protocol parameters; Timing parameters such as transmit / receive time offsets Δ_tx and Δ_rx; Physical layer parameters such as scrambling sequence and interleaving pattern; Control channel parameters such as uplink scheduling request resources and downlink listening resources.

[0019] This characteristic makes rule D not only applicable to perception scenarios, but also widely applicable to the zero signaling generation of various communication parameters, providing terminals and network nodes with a very simple deterministic parameter acquisition method.

[0020] The unique value of rule D lies in: 1. Physical randomness injection: The state value contains microscopic uncertainties of the local physical clock, injecting an additional source of random entropy into the generated parameters, further enhancing security and collision resistance; 2. Extremely simplified: No need to maintain complex hash chains or logical counters, suitable for complexity-sensitive IoT devices; 3. Timing and Content Binding: By physically sampling, "when to make a decision" and "decision output" are naturally bound together, making it impossible for attackers to forge legitimate behavior by replaying old state values; 4. Complementary to existing rules: Rule D fills the gap of "based on physical instantaneous values ​​but with logical decoupling", and together with Rule A (based on broadcast timing), Rule B (based on secret chain), and Rule C (based on public logic), it constitutes a complete protocol security state machine system.

[0021] Concept 6: Fixed Offset A fixed amount of time used to calculate the specific physical operation moment from `T_anchor`. This invention distinguishes between two types of fixed offsets: - Fixed communication offsets: `Δ_dl` (downlink listening) and `Δ_ul` (uplink sending), used for regular uplink and downlink communication links.

[0022] - Sensing transmit / receive offsets: `Δ_tx` (transmit offset) and `Δ_rx` (receive window start offset), specifically designed for sensing tasks. The offsets can be statically configured by the system or dynamically generated from the PSSM state via cryptographic functions.

[0023] Concept 7: Logic Processing Center (LPC) and Physical Access Node (PAN) The LPC is centrally deployed at ground stations, GEO satellites, or the core network side, responsible for centralized processing functions such as logical resource management, mapping table maintenance, global arbitration, location calculation, and clock deviation management. The PAN is deployed at base stations, satellites, UAVs, or dedicated sensing nodes, responsible for physical layer transmission and reception, pre-compensation, measurement reporting, and local signal processing. The LPC and PAN are connected through a high-precision synchronous fronthaul network, together forming a decoupled architecture that separates logic and physical layers.

[0024] Concept 8: Two-Step Deterministic Arbitration When multiple users' resource requests conflict due to the "modulo funnel" effect, the network side performs two steps: "conflict arbitration" and "resource availability check," determining the winner and notifying it via downlink implicit authorization. The order of these two arbitration steps can be: - Order 1: First, arbitrate the conflict between users, and then check the resource availability from the winner; - Sequence 2: First, perform a resource availability pre-check (temporary reservation) on all conflicting users, then arbitrate to select the winner and convert the temporary reservation into a formal use.

[0025] Both sequences are within the scope of protection of this invention and can be flexibly selected according to the system design. This mechanism transforms random collisions into a deterministic queuing process.

[0026] Concept 9: Sensing Pilot The pilot sequence, specifically designed for integrated sensing and computation, is generated by the transmitting node using a cryptographic deterministic function based on the DSF, current logical state, and context information. Cooperative receiving nodes independently generate identical pilot copies based on the same input for matched filtering and target detection. Sensing pilots can exist independently or be multiplexed with communication pilots. The copy used for matched filtering by the receiving node must have the exact same context as the transmitted pilot (i.e., the same label and node identifier) ​​to ensure coherent detection.

[0027] Concept 10: Two-Layer Decoupling Architecture - First layer of decoupling - Resource location decoupling: Mapping logical resource indexes to physical resource locations `(UE_ID, logical index) → (PAN_ID, beam ID, time-frequency coordinates)`, so that user equipment is unaware of changes in physical network topology.

[0028] - Second-layer decoupling – parameter decoupling: The physical layer transmission parameters are divided into a reference parameter set and a dynamic parameter set. This ensures that the user equipment is only aware of the reference parameters and is completely unaware of the dynamic parameters used in actual transmission (such as subcarrier spacing, frequency offset compensation value, CP length, etc.). The dynamic parameter set is dynamically adjusted by the network side based on real-time channel conditions, location information, ephemeris, etc., and through mechanisms such as pre-compensation, resampling, and receive compensation, the real physical dynamic channel is converted into a virtual ideal channel transparent to the user equipment.

[0029] Concept 11: Baseline Physical Layer Parameters The user equipment (UE) baseband processing is based on parameters configured during connection establishment and can remain unchanged for a long period. These parameters include, but are not limited to: reference subcarrier spacing, reference CP length, reference sampling rate, reference resource grid definition, operating frequency, and logical resource pool size. All UE baseband processing (including sensing pilot generation) is based on these reference parameters.

[0030] Concept 12: Dynamic Physical Layer Parameter Set The network side uses parameters for each transmission, which can be dynamically adjusted based on real-time information and are completely transparent to the user equipment. These parameters include, but are not limited to: actual physical resource locations, actual subcarrier spacing, actual CP length, downlink frequency offset compensation value `f_comp_dl`, uplink frequency offset compensation value `f_comp_ul`, timing advance, actual pilot pattern, and antenna port set. The dynamic parameter set can be maintained by the LPC and guide the PAN in transmission and reception processing.

[0031] Concept 13: Pre-Conversion One of the core technologies for achieving parameter decoupling is as follows: When transmitting downlink signals, the network side first generates a baseband signal based on the user equipment's reference parameters. Then, through resampling, frequency domain mapping, and other methods, the signal is converted into a form that conforms to the actual dynamic parameters before transmission. When the user equipment receives the signal, it is pre-converted to conform to the reference parameters and can be directly demodulated by the reference receiver. This technology is completely transparent to the user equipment and is the preferred method for achieving parameter decoupling.

[0032] I. Technical Problem to be Solved by the Invention The applicant's prior patent series (including the DSF basic patent, LPC+PAN patent, deterministic arbitration patent, and unified anchor point timing patent) has solved problems such as zero signaling generation of communication parameters, resource location decoupling, multi-user conflict arbitration, and timing determinism. However, in the integrated communication and computing scenario, the following unresolved technical problems still exist: 1. Collaborative Generation and Distribution of Sensing Pilots: In existing sensing algorithms, multi-node collaborative sensing requires exchanging pilot information via signaling, which is costly and time-consuming. How can collaborative nodes obtain the pilot copies they need from each other without signaling interaction? 2. Impact of physical layer uncertainties on sensing: Factors such as Doppler frequency offset, channel time variation, and clock drift can disrupt the coherence of matched filtering, reducing sensing accuracy. Traditional solutions require nodes to estimate and compensate themselves, increasing complexity and power consumption, and making it difficult to guarantee real-time performance.

[0033] 3. Inconsistent physical layer parameters between nodes lead to cooperation failure: When the local clocks of cooperating nodes deviate or use different physical layer configurations, even if the pilot replicas are consistent, misalignment of transmission and reception times can cause sensing failure. Existing solutions lack an effective management mechanism for such deviations.

[0034] 4. Compatibility of multi-path architecture: In the evolution from existing 5G networks to future 6G NTN, a unified solution that can adapt to different network architectures is needed.

[0035] The purpose of this invention is to solve the above problems and provide a deterministic sensing method that integrates sensing and computation based on cryptographic primitives to generate sensing pilots. This method achieves zero signaling cooperation, physical layer transparency, and supports the smooth evolution of various network architectures. In particular, it provides a centralized management mechanism to address the problem of inconsistent physical parameters between nodes.

[0036] II. Core Inventive Concept The core of this invention lies in: utilizing pre-shared cryptographic materials (such as session keys) and time-varying synchronization parameters between the communicating parties, a sensing pilot sequence for coherent sensing computation is independently generated locally through a cryptographic deterministic function. The transmitting node uses the generated pilot as a probe signal, and the cooperating receiving node independently calculates the same pilot copy using the same cryptographic materials and synchronization parameters, thereby achieving coherent sensing without any real-time signaling interaction.

[0037] This core idea can be uniformly expressed as: ``` Pilot_sensing(t) = F( K_sec, Sync(t), Context ) ``` in: - `K_sec`: Shared session key, pre-established through secure negotiation; - `Sync(t)`: Time-varying synchronization parameter, which can be in various forms such as physical layer broadcast timing (e.g., system frame number SFN, corresponding to rule A), logical state `S(t)` output by the protocol security state machine (corresponding to rule B / C), or unified anchor point time `T_anchor` (corresponding to rule D, where `S(t) = T_anchor`). - `Context`: Context information used to distinguish different nodes, different services, or different parameter types. Crucially, the copy used for receiving matched filtering must use the exact same Context (i.e., the same tag and node identifier) ​​as the transmit pilot to ensure consistency. The Context is an optional parameter. - `F`: Cryptographic deterministic functions, such as HMAC, KDF, hash functions, etc., which guarantee the randomness and uniqueness of the output.

[0038] Building upon this, the present invention further introduces a two-layer decoupled architecture to make physical layer uncertainties transparent to the sensing nodes: - By decoupling resource locations, the sensing nodes become unaware of changes in network topology; - By decoupling parameters, the sensing nodes are unaware of the actual transmission parameters (subcarrier spacing, frequency offset compensation, etc.), and all dynamic adaptations are centrally processed by the network side.

[0039] To address the issue of inconsistent physical parameters between nodes, this invention introduces a centralized clock skew management mechanism in path C: LPC establishes a clock skew table for each PAN through periodic synchronization beacons or measurement reports, unifying the time axis in the calculation of transmission and reception times and location resolution, thereby eliminating the impact of local clock skew.

[0040] Based on the above core principles, this invention provides three implementation paths, each corresponding to a different network architecture and evolution stage.

[0041] III. Explanation of PSSM Synchronization and Physical Clock Deviation In this invention, the evolution of the Protocol Secure State Machine (PSSM) is a deterministic logical process, entirely determined by the shared DSF triples (specifically `Init_Anchor` and `Rule_ID`). Therefore, as long as both communicating parties possess the same DSF, their independently operating PSSMs will evolve along the exact same logical timeline, remaining logically synchronized. This logical synchronization is independent of the precision of the physical clock, as state transitions are driven by logical ticks and are unrelated to physical time.

[0042] However, when it is necessary to map logical decision moments to the physical timeline for actual signal transmission and reception, a unified anchor moment `T_anchor` needs to be obtained using a local physical clock. In this case, the accuracy and synchronization error of the physical clock will affect the absolute value of `T_anchor`. The design of this invention addresses this issue in the following way: - Definition of `Init_Anchor`: `Init_Anchor` can have two forms: - Absolute Anchoring: Specifies a specific future physical moment (such as system frame number SFN, timeslot number), requiring both communicating parties to have achieved physical layer synchronization through external means (such as GPS, IEEE 1588). Applicable to some scenarios of rules A and B / C.

[0043] - Relative Anchoring: Specifies an offset relative to a common event (such as the time when this DSF configuration message is received), or specifies a mutually observable synchronization beacon. In this mode, both communicating parties independently calculate the start time based on their respective local clocks, and the start deviation is isolated and absorbed by the subsequent `T_anchor` mechanism. Applicable to rules B, C, and D.

[0044] - Error encapsulation of `T_anchor`: The local clock is sampled instantaneously at the logic decision moment. The error only affects the current `T_anchor` value and does not accumulate or propagate. Subsequent `Δ_tx` and `Δ_rx` are fixed offsets relative to this `T_anchor`, so the relative relationship remains unchanged.

[0045] - Centralized calibration: In path C, LPC establishes clock deviation tables for each PAN via synchronization beacons or measurement reports, and performs unified calibration when necessary.

[0046] Therefore, this invention does not require strict synchronization of the physical clocks of all nodes. Instead, it achieves the containment and isolation of physical clock errors through flexible `Init_Anchor` definition, `T_anchor` mechanism, and centralized calibration. In particular, under rule D, `T_anchor` itself serves as a state value, and the small deviations it contains become the source of state randomness rather than errors that need to be eliminated. This characteristic makes rule D particularly suitable for scenarios that require the injection of physical entropy sources.

[0047] IV. Overview of the Three Implementation Paths | Path | Network Architecture | Synchronization Parameter Source | Clock Skew Management | Applicable Scenarios | |------|----------|--------------|--------------|----------| | Path A | Traditional 5G architecture (base station + core network) | Physical layer broadcast timing (SFN, timeslot number) | Relies on GPS / 1588, no active calibration | Rapid deployment, compatible with existing terminals, suitable for short-term or coarse-grained sensing | | Path B | Traditional 5G architecture + DSF + PSSM | Logical state `S(t)` output by PSSM (rules A / B / C / D) + `T_anchor` | Relies on initial synchronization, error isolation but no centralized calibration | High performance requirements, pursuing deterministic latency and forward security | | Path C | LPC+PAN Decoupled Architecture | PSSM runs on PAN, `S(t)` is generated by rules A / B / C / D, LPC provides clock skew calibration | LPC provides centralized management, proactive calibration, and strong tolerance | Ultimate goal: complete decoupling of logic and physical components, especially suitable for dynamic topologies such as NTN | The specific implementation of the three paths is described in detail below. In the description of each path, rules A / B / C / D will be flexibly selected according to the needs of the scenario. Among them, rule D (physical anchor point driven) is a newly added rule, and its state `S(t) = T_anchor` can be directly used for sensing pilot generation.

[0048] V. Path A: 5G traditional network architecture + cryptographic primitive generation sensing pilot 5.1 System Composition - Base station (gNB): A 5G base station with sensing computing capabilities, capable of transmitting sensing pilots and receiving echoes.

[0049] - Collaborating Node: This can be another base station, a dedicated sensing node, or a UE, which shares a key with the transmitting base station.

[0050] - Core network: Responsible for key negotiation and distribution, such as establishing shared keys between base stations through the 5G AKA process.

[0051] 5.2 Key Establishment Base stations establish a shared key `K_sec` through the core network. For example, the AMF can configure temporary keys for cooperative sensing for neighboring base stations, which are securely transmitted through the N2 interface.

[0052] 5.3 Sensing Pilot Generation The base station generates sensing pilots based on the current system frame number (SFN) and slot number. Let the SFN corresponding to the current time be `SFN_t`, and the slot index be `Slot_t`, then the synchronization parameters are defined as follows: ``` Sync(t) = SFN_t * N_slot + Slot_t ``` Where `N_slot` represents the number of time slots per frame. The transmitting base station (denoted as gNB_A) generates the transmission pilot: ``` P_tx_A = HMAC-SHA256( K_sec, Sync(t) || "TX" || gNB_A_ID ) mod 2^L ``` The first L bits are truncated as the pilot sequence. The cooperating receiving base station (denoted as gNB_B) independently calculates the same `Sync(t)` and generates a replica of the received matched filter: ``` P_rx_for_A = HMAC-SHA256( K_sec, Sync(t) || "TX" || gNB_A_ID ) mod 2^L / / Note: Same context as the transmit pilot ``` Note that `Sync(t)` must be strictly aligned, which requires that the base stations have already achieved frame-level synchronization via air interface synchronization or GPS. Path A assumes that the base stations participating in the cooperation have already achieved high-precision time synchronization through traditional methods.

[0053] 5.4 Collaborative Perception Process 1. Configuration phase: The core network distributes the shared key `K_sec` and cooperative task information (such as sensing period and target area) to the base stations participating in the cooperation.

[0054] 2. Periodic perception: - In each sensing cycle, gNB_A transmits `P_tx_A` on a predetermined time-frequency resource (configured by the upper layer).

[0055] - gNB_B opens a receiving window on the same time-frequency resources, uses locally generated `P_rx_for_A` for matched filtering, and detects target echoes.

[0056] - At the same time, gNB_B can also transmit its own pilot signal, which gNB_A receives, to achieve two-way sensing.

[0057] 3. Measurement Reporting: gNB_B reports the detected echo arrival time (TOA), angle of arrival (AOA), and other measurements to the positioning server via the X2 interface or the core network (which can be integrated into the core network).

[0058] 4. Location calculation: The positioning server uses measurements from multiple base stations to perform TDOA / AOA fusion positioning.

[0059] 5.5 Advantages and Limitations - Advantages: No need to add new signaling pilots, utilizes existing broadcast timing, is compatible with 5G networks, and is easy to deploy.

[0060] - Limitations: Relies on GPS / 1588 synchronization, and physical clock errors directly affect sensing accuracy; SFN cycle is limited, and the loopback problem needs to be solved; key updates require additional signaling; there is no clock deviation management, and sensing fails when parameters between nodes are inconsistent.

[0061] 6. Path B: Traditional 5G architecture + DSF + PSSM 6.1 Introduction of DSF The Dynamic Security Foundation (DSF) is defined by a triple: `DSF = (K_sec, Init_Anchor, Rule_ID)`. The core network generates the DSF for the awareness task and distributes it to participating nodes via RRC signaling or the core network interface. `Rule_ID` can be selected from a set of rules A / B / C / D.

[0062] 6.2 System Composition - Base station: integrates DSF function and runs PSSM (using one of the rules A / B / C / D).

[0063] - Core Network: Responsible for the generation and distribution of DSF.

[0064] 6.3 PSSM Synchronization and Init_Anchor The logical synchronization of PSSM depends on the shared DSF. `Init_Anchor` can take two forms: - Absolute Anchoring: Specifies a future physical time (e.g., `(SFN=1000, Slot=0)`), requiring frame-level synchronization to be achieved between base stations. Applicable to certain scenarios of Rule A and Rule B / C.

[0065] - Relative Anchor: Specifies the offset relative to the received DSF configuration message (e.g., "Starting in the 500th time slot after receiving this signaling"). Each base station counts based on its local clock, and minor deviations in startup time are isolated by subsequent `T_anchor` mechanisms. Applicable to rules B, C, and D.

[0066] 6.4 Sensing Pilot Generation The base station and cooperating nodes share the same DSF, but operate the PSSM independently. At each logical decision time, the PSSM outputs the logical state `S(t)`. The sensing pilot generation formula is: ``` P_tx_i = F( K_sec, S(t), "TX" || Node_ID_i ) / / Transmit pilot P_rx_for_j = F( K_sec, S(t), "TX" || Node_ID_j ) / / Receive node detects the pilot signal from j ``` The value of `S(t)` depends on the selected rule: - Rule A: `S(t) = SFN||Slot` (broadcast timing) - Rule B: `S(t) = T_State[n]` (hash chain value) - Rule C: `S(t) = (Epoch_ID, Counter)` (Logical epoch state) - Rule D: `S(t) = T_anchor` (physical anchor time) 6.5 Enhanced Timing Determinism – Unifying Anchor Point Timing and Fixed Offset A unified anchor time `T_anchor` and fixed offsets `Δ_tx` and `Δ_rx` are introduced. At the logical decision time, the node instantaneously samples its local physical clock to obtain `T_anchor`, and then calculates: - Launch time: `T_tx = T_anchor + Δ_tx` - Receive window start time: `T_rx_start = T_anchor + Δ_rx` Here, `Δ_tx` and `Δ_rx` are fixed values ​​predefined by the system, or generated by the PSSM state `S(t)` through a cryptographic function (e.g., `Δ_tx = G(K_sec, S(t), "TX_OFFSET" || Node_ID)`). They must satisfy: `Δ_rx > Δ_tx + T_process + T_prop_max`, where `T_process` is the node's handover time from transmission to reception, and `T_prop_max` is the maximum possible round-trip propagation delay to ensure that the reception window can cover the echo arrival time.

[0067] The network side (base stations) operates the same PSSM, thus enabling the prediction of each node's `T_anchor` and transmit / receive times, and performing two-step arbitration based on `T_anchor`. While there may be slight deviations in the physical clocks of each base station, these deviations only affect the absolute value of `T_anchor`, while the relative relationship between `T_tx` and `T_rx` (determined by a fixed offset) remains unchanged, allowing for accurate resource conflict detection. For more precise synchronization, the network can periodically calibrate its local clock using synchronization beacons.

[0068] Explanation of the relationship between communication offset and sensing offset: It should be specifically noted that this invention distinguishes between two types of fixed offsets—communication fixed offsets (`Δ_dl`, `Δ_ul`) and sensing transmit / receive offsets (`Δ_tx`, `Δ_rx`), which serve different purposes and are independent of each other: - The communication fixed offset is used to ensure bidirectional interaction between the UE and the network. It must be strictly aligned with the radio frame structure to ensure that the network can accurately predict the UE's transmit and receive times. Its value is usually specified by the network static configuration or RRC signaling.

[0069] - The sensing transmit / receive offset is used for cooperative sensing between PANs. It only needs to satisfy the physical layer feasibility constraint `Δ_rx > Δ_tx + T_process + T_prop_max`, and has no necessary binding relationship with the wireless frame structure. In path B, the sensing offset is limited by the alignment of `T_anchor` with the frame structure, but can still be freely selected within the frame; in path C, the sensing offset can be completely independent of the frame structure, providing maximum flexibility for sensing tasks.

[0070] Both types of offsets share the same `T_anchor` as a time reference, but are otherwise uncoupled. This design allows communication and sensing to operate independently in their respective optimal ways, achieving efficient resource reuse. For example, in path C, the PAN can be dedicated to sensing during periods when the UE is inactive, achieving time-division multiplexing of communication and sensing; sensing signals can also be inserted during communication gaps, achieving frequency-division or code-division multiplexing. The independence of the two types of offsets is one of the core design concepts of this invention and is also key to achieving optimal performance of the integrated communication and sensing system.

[0071] 6.6 Collaborative Awareness Process 1. Initialization: The core network generates a DSF for the cooperative sensing task and distributes it to participating base stations via security signaling. Simultaneously, each base station calibrates its local clock offset using high-precision time synchronization (e.g., 1588) to ensure a unified reference for `T_anchor`, or employs relative anchoring for startup.

[0072] 2. Autonomous operation: - Base stations A and B independently operate PSSM (using one of the rules A / B / C / D), generating `S(t)` at logical time t respectively, and then calculating `P_tx_A`, `P_tx_B`, `Δ_tx_A`, `Δ_rx_B`, etc.

[0073] Base station A transmits `P_tx_A` at `T_anchor_A + Δ_tx_A`; base station B opens a receive window at `T_anchor_B + Δ_rx_B` (the window length is sufficient to cover the maximum possible propagation delay) and uses the locally generated `P_rx_for_A` for matched filtering.

[0074] 3. Measurement and Reporting: The base station reports the measured values ​​(TOA, AOA) and `T_anchor` value to the positioning server. The reporting process can reuse existing communication mechanisms or reserved reporting resources.

[0075] 4. Location Calculation: The positioning server uses the `T_anchor` reported by each base station and the measured values, combined with the known base station locations, to calculate the target location. Since `T_anchor` is bound to logical time, and all base stations use the same logical timeline, even if there is a deviation in the physical clock, the error can be eliminated by the difference in `T_anchor` (provided that the deviation is within a reasonable range).

[0076] 6.7 Advantages and Limitations - Advantages: Decoupling of logic and physical components, preventing the accumulation of physical clock errors; PSSM provides forward safety and determinism; supports finer-grained slot-level awareness; partially tolerates initial synchronization errors through relative anchoring; Rule D provides an extremely simplified solution for scenarios requiring physical randomness.

[0077] - Limitations: There is no centralized clock skew management, and long-term operation still relies on coarse synchronization; inconsistencies in physical parameters between nodes (such as excessive crystal oscillator drift) may still lead to misalignment of transmit and receive windows; and the adaptability to large-scale dynamic networks is limited.

[0078] VII. Path C: LPC+PAN decoupling architecture 7.1 System Architecture - Logical Processing Center (LPC): Deployed at ground stations, GEO satellites, or the core network, it serves as the global logical anchor point and is responsible for: - Generate perceptual DSF triples and distribute them to PAN, where `Rule_ID` can be selected from the set of rules A / B / C / D; - Maintain a global mapping table to record the physical location, coverage area, and resource status of each PAN; - Coordinate sensing tasks and perform location calculations; - Maintain a dynamic parameter set to achieve parameter decoupling; - Establish and maintain clock offset tables for each PAN, and calibrate the local clock using periodic synchronization beacons or measurement reports.

[0079] - Physical Access Node (PAN): Deployed at base stations, satellites (usually LEO satellites), drones, etc., responsible for: - Run PSSM independently based on shared DSF (using one of the rules A / B / C / D); - Generate sensing pilots, transmit / receive time offsets, and time-frequency resources at each logical decision-making moment; - Instantaneously sample the local physical clock to calculate the actual transmission and reception times; - Perform signal transmission and reception and local detection, and report the measured quantities (with local `T_anchor`). - Perform physical layer processing such as frequency offset compensation and waveform adaptation based on the dynamic parameter set; - Receive and respond to the LPC synchronization beacon, and report clock skew information.

[0080] 7.2 PSSM Synchronization and Clock Deviation Management All PANs share the same perceptual DSF, so their independently operating PSSMs are fully synchronized on the logical timeline. Logical synchronization is guaranteed by the DSF, but the sampled value of `T_anchor` depends on the local physical clock. LPC manages clock skew through the following mechanisms: 1. Periodic Synchronization Beacon: LPC broadcasts a high-precision synchronization beacon via a feed link or inter-satellite link. The beacon encodes the global reference time `T_ref` (i.e., the precise time of beacon transmission).

[0081] 2. Deviation Measurement and Reporting: Upon receiving a beacon, each PAN records its local time `T_local` and calculates the deviation `Δ_calib = T_ref - T_local`. The PAN reports `Δ_calib` to the LPC, or directly includes `T_anchor` in subsequent measurements for unified processing by the LPC.

[0082] 3. Deviation Table Maintenance: LPC establishes and maintains the clock deviation table `Δ_calib_i(t)` for each PAN, and can update it smoothly over time (e.g., using Kalman filtering).

[0083] 4. Unified Time Axis: During position calculation, LPC converts all local `T_anchor` reported by the PANs into global time: `T_global = T_anchor_i + Δ_calib_i(t)`, thereby eliminating the influence of local clock deviations. Furthermore, LPC can proactively send fine-tuning commands to the PANs based on deviation information, guiding them to adjust the counting frequency of their local clocks, thus achieving proactive closed-loop calibration.

[0084] Even if the PAN uses a low-cost crystal oscillator (e.g., ±20ppm), the residual error can be controlled at the nanosecond level through periodic calibration (e.g., once every 1 second).

[0085] Regarding the coordination between Rule D and LPC clock skew management: It is important to note that when the PAN adopts rule D (physical anchor driven), its state `S(t) = T_anchor` is directly based on the instantaneous sampling of the local physical clock. While this design brings advantages such as extremely simple implementation complexity and natural physical randomness injection (see Concept 5 for details), it also introduces the problem of `T_anchor` inconsistency caused by the local clock deviation of each PAN—the `T_anchor` obtained by different PANs at the same logical decision moment may have slight differences. If directly used for measurement reporting and position calculation, it will affect the accuracy of collaborative sensing.

[0086] This invention perfectly resolves this contradiction through the LPC centralized clock skew management mechanism. As mentioned earlier in this section, LPC establishes and maintains the clock skew table `Δ_calib_i(t)` for each PAN using periodic synchronization beacons. When a PAN adopts rule D, it includes a local `T_anchor` in its measurement report. LPC uses the skew table to unify all `T_anchor`s to the global time axis: `T_global = T_anchor_i + Δ_calib_i(t)`. Subsequently, all calculations of measurements based on the unified global time (such as time difference of arrival and distance calculation) can obtain results equivalent to high-precision synchronization.

[0087] This collaborative mechanism forms a complete closed-loop solution: Rule D is responsible for minimal state generation and physical randomness injection; - LPC centralized calibration is responsible for eliminating the resulting clock skew and ensuring the accuracy of multi-node collaboration; - Combining the two retains the simplification and security advantages of rule D while achieving high-precision perception, making it the preferred implementation method in path C.

[0088] 7.3 Sensing Pilot Generation and Transmit / Receive Offset PAN_i generates the following parameters at logical time t: - Transmit pilot: `P_tx_i = F( K_sec, S(t), "TX" || PAN_ID_i )` - Receive a copy (for detecting signals from PAN_j): `P_rx_for_j = F( K_sec, S(t), "TX" || PAN_ID_j )` - Time-frequency resources: `Resource_i = F( K_sec, S(t), "RES" || PAN_ID_i ) mod N_res` - Launch offset: `Δ_tx_i = G( K_sec, S(t), "TX_OFFSET" || PAN_ID_i ) modT_slot_max` - Receive offset: `Δ_rx_i = G( K_sec, S(t), "RX_OFFSET" || PAN_ID_i ) modT_slot_max` Where `G` is another cryptographic deterministic function (or the same as `F` but using a different context), and `T_slot_max` is the maximum slot length (e.g., 1ms) to ensure the offset is within the slot range. The value of `S(t)` depends on the selected rule (A / B / C / D), under rule D `S(t) = T_anchor`.

[0089] The system must ensure that `Δ_rx_i > Δ_tx_i + T_process + T_prop_max`, where `T_process` is the PAN's switching time from transmission to reception, and `T_prop_max` is the maximum possible round-trip propagation delay (e.g., hundreds of milliseconds between LEO satellites). When configuring sensing tasks, the network sets the output range of the `G` function and defines the reception window length `W_rx` (e.g., equal to twice `T_prop_max`), ensuring that the echo always falls within the reception window regardless of when it arrives. If a generated `Δ_rx_i` does not meet the above constraints, the PAN can automatically adjust to the effective range according to preset rules (e.g., adding a base offset `Δ_base`). The impact of this adjustment on sensing performance can be absorbed by the system parameter design.

[0090] Key point: The replica used by the receiving node for matched filtering must use the exact same context as the transmitting pilot ("TX" || Transmitter Node ID"), while the "RX" context is only used to generate the receiving node's own control parameters (such as the receive window offset), and the two do not interfere with each other. This design ensures the consistency of coherent detection.

[0091] 7.4 Dynamic Parameter Sets and Parameter Decoupling LPC maintains a dynamic parameter set for each PAN, including the actual subcarrier spacing, frequency offset compensation value, timing advance, etc. The PAN performs physical layer processing based on the dynamic parameter set during transmission and reception, but the baseband signal is always generated based on reference parameters. For example: - Before transmission, the PAN performs frequency offset pre-compensation on the signal based on `f_comp_dl`; - After receiving the signal, the PAN performs frequency offset compensation based on `f_comp_ul`; - If the actual subcarrier spacing differs from the reference, the PAN uses a pre-conversion technique (such as frequency domain mapping) to convert the reference signal into the actual waveform.

[0092] The dynamic parameter set can be updated in real time by LPC based on ephemeris, target location, channel conditions, etc., and notified to PAN via downlink signaling (or implicitly).

[0093] 7.5 Collaborative Awareness Process 1. Task Configuration: LPC generates DSF_sensing for the sensing area and selects participating PANs (e.g., PAN_A, PAN_B). The DSF (including the selected Rule_ID), the peer PAN_ID, baseline parameters, and sensing transmit / receive offset configuration are distributed to each PAN via a secure link. Simultaneously, LPC reports the logical start time (which can use relative anchoring, such as "starting on the 100th logical tick after receiving this configuration").

[0094] 2. Autonomous operation: - PAN_A and PAN_B run PSSM independently (using one of the rules A / B / C / D), obtain `S(t)` at logical time t, and generate their respective parameters.

[0095] - PAN_A transmits `P_tx_A` to `Resource_A` at `T_anchor_A + Δ_tx_A`, and performs frequency offset pre-compensation based on `f_comp_dl` in the dynamic parameter set before transmission.

[0096] - PAN_B opens a receiving window (window length `W_rx`) at `T_anchor_B + Δ_rx_B`, performs matched filtering on `Resource_B` using `P_rx_for_A`, and performs frequency offset compensation based on `f_comp_ul` in the dynamic parameter set after reception. If the target reflected echo arrives, a peak appears in the matched filter output.

[0097] 3. Measurement and Reporting: After PAN_B detects the target echo, it records the arrival time `TOA` and the local `T_anchor_B`. The reporting process can reuse the communication mechanism in this invention: LPC configures a low-priority DSF for measurement reporting, and PAN_B reports the measurement to LPC according to the communication process (calculating resources, listening for downlink authorization, and sending data) at subsequent logical decision moments.

[0098] 4. Clock Deviation Calibration: LPC periodically broadcasts synchronization beacons, and each PAN reports its local time or includes `T_anchor`. LPC then updates the clock deviation table.

[0099] 5. Position Calculation: LPC collects measurements from multiple PANs, uses the `T_anchor` reported by each PAN and the known clock offset to unify all times onto the global time axis, and combines the known PAN positions (obtained from the mapping table) to calculate the target position.

[0100] 7.6 Advantages of NTN Scenarios - No handover awareness: When a PAN (such as a LEO satellite) moves, the LPC only updates the mapping table, and the PAN's sensing mission is not affected.

[0101] - Power supply interruption tolerance: The PAN locally caches DSF and mapping tables, and can autonomously continue to detect during power supply link interruptions and report them after recovery.

[0102] - Doppler transparency: LPC accurately calculates the frequency offset compensation value based on ephemeris, and PAN transmits and receives according to the reference parameters, so it is not affected by Doppler.

[0103] - Clock skew tolerance: Through centralized calibration, long-term sensing accuracy can be guaranteed even if the PAN uses a low-cost crystal oscillator.

[0104] - Physical layer parameter inconsistency adaptation: LPC can allocate different dynamic parameter sets (such as subcarrier spacing) to different PANs according to their capabilities and channel conditions, and ensure signal compatibility through pre-conversion.

[0105] 7.7 Advantages and Limitations - advantage: - Completely eliminate signaling overhead; all parameters are generated locally. - Naturally synchronized, logic and physical decoupling, low external synchronization dependency; - Centralized clock skew management, tolerating inconsistencies in physical parameters between nodes; - Dynamic parameter sets enable physical layer transparency and reduce the complexity of perception nodes; - Supports ultra-large-scale dynamic networks, enabling seamless mobility; - Strong tolerance to power supply interruptions and high robustness; Rule D provides an extremely simplified access solution for low-cost, low-complexity nodes.

[0106] - Limitations: It requires the introduction of new network element LPC, which involves significant changes to the existing network architecture, but it provides the optimal solution for 6G NTN.

[0107] 7.8 On the application of two-step deterministic arbitration between PANs It is worth noting that, unlike previous patents which primarily addressed uplink resource conflicts between multiple user equipment (UEs), the two-step deterministic arbitration mechanism of this invention is creatively applied here to multiple network-side physical access nodes (PANs) to coordinate their spontaneously generated sensing signal transmission resources. This signifies that this invention successfully extends the concept of deterministic conflict management from the "terminal side" to the "network infrastructure side," laying the foundation for large-scale, self-organizing collaborative sensing networks.

[0108] The specific implementation of arbitration is similar to that of UE-to-UE arbitration: LPC runs a PSSM copy synchronized with each PAN to accurately predict the resource intention `Resource_i` that each PAN will generate at each logical decision moment in the future. When LPC finds that at the same `T_anchor` moment, multiple PANs' `Resource_i` points to the same physical resource, it determines that a resource conflict has occurred, and then performs a two-step arbitration (two orders are optional) and notifies the winner via downlink.

[0109] Explanation of the licensing method between LPC and PAN: It should be further pointed out that the downlink authorization between LPC and PAN differs from the downlink implicit authorization between LPC and UE in its implementation, but both follow the same core idea—to transmit authorization information through downlink signals and avoid allocating independent authorization signaling to each node.

[0110] Specifically: - Between LPC and UE: Due to limitations in air interface resources and UE power consumption, a strict "implicit licensing" approach is adopted. The UE listens on a pre-determined downlink resource `R_d` and determines whether it has been granted licensing based on the presence or absence of the signal. The signal itself does not contain any UE identifier; the UE confirms whether the signal is for itself through the binding relationship between `R_d` and its own DSF. This method compresses the licensing information to 1 bit (present / absent) and requires no signaling overhead.

[0111] - Between LPC and PAN: Thanks to the high bandwidth and reliability of the feeder link, a more direct granting method can be adopted. The LPC sends explicit grant messages (such as `Grant(PAN_ID, Resource, T_tx)`) to the PAN via the feeder link, or broadcasts a grant list containing all winning PANs. Although this method is "explicit" in form, it is essentially an extension of the "downlink implicit granting" concept—centralizing the arbitration result to all PANs, rather than allocating a separate granting channel to each PAN.

[0112] The core of both approaches lies in: 1. Authorization information is uniformly generated by LPC, based on the results of a two-step arbitration process, ensuring global consistency; 2. The transmission of authorization information does not require a request from the PAN; it is a proactive decision made by the network side. 3. Authorization information is bound to DSF status to ensure security and uniqueness; 4. The receiving party (UE or PAN) independently determines whether it is authorized, and when, where, and with which resources to perform the operation based on the local DSF and the received signal.

[0113] Therefore, despite their different physical implementations, they together constitute the complete meaning of the "downlink implicit authorization" concept of this invention—unified decision-making on the network side and autonomous action on the execution side based on shared state and downlink signals. This design concept runs through all nodes from the terminal to the infrastructure, reflecting the unity and scalability of the DSF paradigm.

[0114] Explanation of PSSM's runtime location and prediction mechanism: From a broader perspective, this is precisely the core advantage of the DSF paradigm—achieving distributed consensus through shared cryptographic state, enabling the network to "predict" rather than "perceive" the behavior of nodes: | Dimensions | Communication Services (UE Side) | Sensing Services (PAN Side) | |------|-----------------|-------------------| | PSSM Operating Location | UE and LPC | PAN and LPC | | Resource Intent Generation | Independent Calculation by Both Parties | Independent Calculation by Both Parties | How LPC learns intent | By running the same PSSM forecast | By running the same PSSM forecast | | Is reporting required? | No | No | | Arbitration Subject | LPC | LPC | | Authorization Methods | Implicit Authorization over the Air (Signal Presence / Absence) | Explicit Message / Broadcast on the Feeder Link | | Execution Unit | UE | PAN | This symmetrical design ensures that: - Zero signaling: All collaboration is based on a shared cryptographic state, requiring no real-time signaling interaction; - Predictability: The network side has a global view and can accurately predict future resource needs; - Arbitrability: Conflicts can be detected and resolved before they occur, avoiding actual interference; - Unified architecture: Both terminals and network nodes follow the same DSF paradigm.

[0115] Therefore, although resource intentions in the perception service are generated by the PAN and arbitrated by the LPC, prediction is achieved by running the same PSSM through the LPC, forming a natural closed loop between the two, and there is no problem of "separation of decision-making and execution". This is clear evidence of the successful extension of the DSF paradigm from the "terminal side" to the "network infrastructure side", and it is also the theoretical foundation for the end-to-end deterministic perception of this invention.

[0116] VIII. Beneficial Effects Compared with the prior art, the present invention has the following significant advantages: | Dimension | Existing Solutions | Path A | Path B | Path C | |------|----------|-------|-------|-------| | Pilot signaling overhead | High | Zero | Zero | Zero | | Synchronization Dependency | GPS / 1588 | GPS / 1588 | Initial Synchronization + Error Isolation | Logical Decoupling + Centralized Calibration, No External Synchronization Required | | Clock Deviation Management | None | None | None | Centralized calibration, strong tolerance | | Deterministic time delay | Statistical | Statistical | Mathematically provable | Mathematically provable | | Mobility Support | Signaling Required | Signaling Required | Signaling Required | Smooth Mapping Table Migration, Zero Interruption | | NTN adaptability | Poor | Poor | Average | Excellent | | Physical Layer Uncertainty Handling | Node Self-Handling | Node Self-Handling | Node Self-Handling | Network Transparency Handling | | Tolerance for parameter inconsistencies between nodes | None | None | Weak | Strong | | Network Changes | None | Software Upgrade | Added DSF Module | New Network Element LPC | Standardization Difficulty | Easy | Easy | Medium | Difficult (but with good prospects) Specifically: 1. Zero-signaling-aware pilot generation: Collaborating nodes independently generate pilots based on shared cryptographic materials, completely eliminating pilot configuration signaling overhead.

[0117] 2. Multi-node natural synchronization: By decoupling the logical timeline from the physical clock (path B / C) and centralized calibration (path C), the dependence on external synchronization is reduced, and clock deviation is tolerated.

[0118] 3. High-precision sensing: Deterministic pilot signals ensure coherent processing gain, multi-node collaboration improves positioning accuracy and detection probability; centralized clock calibration eliminates synchronization errors.

[0119] 4. Intrinsic security: The pilot is generated based on cryptography, making it unpredictable and resistant to forgery and interference.

[0120] 5. Flexible architecture evolution: Three paths cover the smooth transition from 5G to 6G NTN.

[0121] 6. Manageable resource conflicts: A deterministic arbitration mechanism is introduced to transform collisions into orderly queuing, and arbitration can be applied between PANs.

[0122] 7. Integrated communication and sensing: Multiplexing communication signals enables sensing, doubling spectral efficiency.

[0123] 8. Supports non-cooperative target detection: Passive reflection mode is suitable for security, anti-drone and other scenarios.

[0124] 9. Physical layer transparency: Through two-layer decoupling, the sensing node is completely unaware of Doppler, waveform changes, clock drift, etc., which greatly reduces the complexity.

[0125] 10. Strong tolerance to inconsistencies in physical parameters between nodes: Path C centrally manages clock skew and dynamic parameters through LPC, enabling precise collaborative sensing even if the PAN uses a low-cost crystal oscillator and different physical layer configurations. This is the core advantage of this invention that distinguishes it from existing solutions.

[0126] 11. The unique value of rule D: The newly added physical anchor-driven rule (rule D) provides a brand-new solution for scenarios that require physical randomness injection or extreme simplification. It directly uses the unified anchor time `T_anchor` as the state value, and together with rules A / B / C, it forms a complete protocol security state machine system. Attached Figure Description

[0127] Figure 1 : Schematic diagram of the core invention point - cryptographic pilot generation and zero signaling cooperation.

[0128] Figure 2 Flowchart of collaborative awareness for path A (5G+password).

[0129] Figure 3 Collaborative awareness flowchart for path B (DSF+PSSM).

[0130] Figure 4 System architecture diagram for path C (LPC+PAN).

[0131] Figure 5: Flowchart of the internal functional modules of PAN and the operation of PSSM in path C.

[0132] Figure 6 Flowchart of resource conflicts and two-step arbitration for multi-sensor nodes (including two arbitration sequences).

[0133] Figure 7 : A schematic diagram of sensing timing synchronization based on a unified anchor point.

[0134] Figure 8 : Schematic diagram of a two-layer decoupled architecture.

[0135] Figure 9 A comparative diagram of the four rules (A / B / C / D) of the protocol security state machine. Detailed Implementation

[0136] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be described in detail below with reference to the accompanying drawings and embodiments. The described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0137] Example 1: Dual-base station cooperative target detection based on path A (relying on GPS synchronization) This embodiment demonstrates how, in an existing 5G network, two base stations use a shared key and system frame number to generate sensing pilot signals, enabling cooperative detection of drone targets. It is assumed that the base stations have already achieved high-precision synchronization via GPS.

[0138] System Configuration: - Both base station A and base station B are 5G gNBs, supporting the extension of communication and computing functions. The two establish a shared session key `K_sec` through the core network.

[0139] - The core network configures collaborative sensing tasks for the two base stations, with a sensing period of 10ms.

[0140] - The two base stations have achieved high-precision time synchronization via GPS (accuracy better than 100ns).

[0141] Sensing pilot generation: Base stations A and B generate sensing pilots based on the current system frame number `SFN_t` and time slot number `Slot_t`. A synchronization parameter `Sync(t) = SFN_t * N_slot + Slot_t` is defined, where `N_slot = 20`. Base station A generates a transmit pilot `P_A = HMAC - SHA256(K_sec, Sync(t) || "TX" || CellID_A)`, and base station B generates a receive copy `P_recv_A = HMAC - SHA256(K_sec, Sync(t) || "TX" || CellID_A)`.

[0142] Collaborative detection process: 1. On the reserved time and frequency resources, base station A transmits P_A in time slot `Slot_t`, and base station B opens a receiving window in the same time slot and uses P_recv_A for matched filtering.

[0143] 2. If the target appears in the coverage overlap area, the signal of base station A will be received by base station B after being reflected by the target, and a peak will appear in the matched filter output.

[0144] 3. Base station B records the echo arrival time `TOA_B` and reports it to the positioning server via the X2 interface.

[0145] 4. The positioning server calculates the distance using `TOA_B` and the transmission time (known) of base station A, and combines the base station location to solve for the target.

[0146] Effect: Achieves zero-signaling pilot collaboration, but relies on GPS synchronization.

[0147] Example 2: High-precision positioning based on multi-base station cooperation along path B (using rule D) This embodiment demonstrates how, in a 5G network incorporating DSF, multiple base stations utilize rule D (physical anchor driven) to generate time-varying sensing pilots, achieving high-precision positioning of mobile terminals. Relative anchoring is employed, leveraging the physical randomness of `T_anchor` to enhance security.

[0148] System Configuration: - The core network generates a DSF for the positioning service: `DSF_loc = (K_sec, Init_Anchor, Rule_ID=D)`, where `Rule_ID` is rule D, the logical tick `T_logical=0.125ms`, and `Init_Anchor` uses relative anchoring: "started in the 500th time slot after receiving this signaling".

[0149] - All participating base stations (gNB1, gNB2, gNB3) and terminal UEs obtain this DSF. After receiving the signaling, each base station starts PSSM based on 500 time slots counted by its local clock.

[0150] Sensing pilot generation: At each logical decision time t, the UE and all base stations independently sample their local physical clocks to obtain `T_anchor`, which serves as the current state `S(t) = T_anchor`. The UE generates an uplink positioning pilot `P_UL(t) = HMAC-SHA256(K_sec,T_anchor, "TX" || UE_ID)`. The base station generates the same `P_UL(t)` as a receive copy.

[0151] Timing determinism: The UE samples its local clock at the logical decision point to obtain `T_anchor`, and transmits `P_UL(t)` at `T_UL = T_anchor + Δ_ul`. Each base station records the arrival time `t_rx_i`, calculates the absolute propagation time `τ_i = t_rx_i - T_UL` (assuming the base station knows `T_UL` and can predict it through PSSM), and reports it to the positioning server.

[0152] Position calculation: The positioning server uses multiple `τ_i` to calculate the UE's location through polygonal positioning. While there is a slight deviation at startup, the deviation of `T_UL` is isolated in this measurement and can still be calculated through measurements from multiple base stations. Furthermore, since the pilot signals are generated based on `T_anchor`, the pilot signals at different times possess physical randomness, enhancing resistance to replay attacks.

[0153] Effect: No GPS required; randomness is injected by utilizing the microscopic uncertainty of the physical clock, thus improving security.

[0154] Example 3: Wide-area sensing of LEO satellite constellation based on path C (centralized clock calibration, using rule D) This embodiment demonstrates how to achieve global UAV target detection using an LPC+PAN architecture in NTN, and shows how the LPC calibrates the PAN clock via a synchronization beacon. The PAN adopts a rule-D simplified design.

[0155] System Configuration: - LPC is deployed at ground stations, with multiple LEO satellites serving as PANs.

[0156] - LPC generates DSF_sensing for the sensing task, with `Rule_ID` as rule D, `T_logical=1s`, `N_s=1000`, and `M=50`. `Init_Anchor` uses relative anchoring: "starts on the 10th logical cycle after receiving this DSF".

[0157] - LPC distributes DSF to each PAN via the feed link and configures reference parameters (15kHz subcarrier spacing). The values ​​of the transmit / receive offsets `Δ_tx` and `Δ_rx` range from 0 to 0.5ms, and the receive window length `W_rx=20ms` covers the maximum round-trip time.

[0158] Clock skew calibration: The LPC broadcasts a high-precision synchronization beacon every 10 seconds, encoding the transmission time `T_ref`. Each PAN receives the beacon, records its local time `T_local`, calculates the deviation `Δ_calib = T_ref - T_local`, and reports it to the LPC. The LPC establishes a deviation table and updates it smoothly over time. Simultaneously, the LPC can issue fine-tuning commands to the PANs based on the deviation trend, adjusting their local clock frequencies.

[0159] PAN autonomous perception: Each PAN independently samples its local physical clock at the logic decision time to obtain `T_anchor` as `S(t)`, generating a transmit pilot `P_tx_i = F(K_sec, T_anchor, "TX" || PAN_ID_i)`, a receive replica `P_rx_j`, a resource `Resource_i`, an offset `Δ_tx_i`, and `Δ_rx_i`. The PAN transmits the pilot at `T_anchor_i + Δ_tx_i` and opens the receive window (continuously `W_rx`) at `T_anchor_i + Δ_rx_i`. All baseband signals are generated based on a 15kHz reference.

[0160] Target detection and reporting: After PAN_B detects the echo from PAN_A, it records TOA and the local `T_anchor_B`. In subsequent logical moments, PAN_B uses the low-priority DSF configured for reporting to report the measurements to LPC according to the communication procedure.

[0161] Position calculation: LPC uses the offset table to convert the `T_anchor` of each PAN into global time, and combines the known PAN position and launch time to calculate the target position.

[0162] Results: Even with a PAN crystal oscillator drift of ±20ppm, the residual error is <200ns after calibration every 10 seconds, meeting the requirements for high-precision sensing. The PAN uses regular D, eliminating the need to maintain a complex state machine and achieving extreme simplification.

[0163] Example 4: Cooperative sensing based on Doppler transparency compensation for path C (a mixture of rule B and rule D) This embodiment demonstrates in detail how to achieve Doppler transparency compensation using the dynamic parameter set provided by LPC under path C, while simultaneously using rules B and D in combination.

[0164] Scenario: Multiple LEO satellites (PANs) collaborate to detect high-altitude drones. The high-speed motion of the satellites causes Doppler frequency offsets as high as ±50kHz. PAN_A uses rule B (hash chain driven) to generate high-security pilot signals, while PAN_B uses rule D (physical anchor driven) to simplify the design.

[0165] process: 1. Based on the ephemeris and the preset target area, LPC calculates the Doppler frequency offset `f_doppler_dl` (PAN→target) and `f_doppler_ul` (target→PAN) for each PAN in the current period, and sets `f_comp_dl = -f_doppler_dl` and `f_comp_ul = +f_doppler_ul` in the dynamic parameter set.

[0166] 2. PAN_A generates `S_A(t) = T_State[n]` according to rule B, and PAN_B samples `T_anchor` according to rule D to obtain `S_B(t) = T_anchor`.

[0167] 3. The two PANs generate pilot signals based on their respective states: `P_A = F(K_sec, S_A(t), "TX" || PAN_A_ID)`, `P_B = F(K_sec, S_B(t), "TX" || PAN_B_ID)`.

[0168] 4. Before transmission, the two PANs perform frequency offset pre-compensation on the signal according to `f_comp_dl`, so that the signal has no frequency offset when it reaches the target area.

[0169] 5. The target reflected signal is received by the other party. The receiving PAN performs frequency offset compensation on the received signal according to `f_comp_ul`, and then uses the locally generated copy of the other party's pilot signal for matched filtering.

[0170] 6. Report the detected echo.

[0171] Results: PANs do not require real-time frequency offset estimation, achieving compensation accuracy at the Hz level and ensuring sensitivity in weak echo detection. PANs of different rules can work seamlessly together, demonstrating the uniformity of the DSF paradigm.

[0172] Example 5: Sensing for Dynamic Adaptation of Waveform Parameters via Pre-conversion Method This embodiment demonstrates how, in path C, the actual subcarrier spacing is dynamically adjusted through pre-conversion technology to optimize sensing performance.

[0173] Scenario: Ground-based PAN needs to detect long-range hypersonic targets, requiring high Doppler resolution, and LPC decision-making uses an actual subcarrier spacing of 60kHz.

[0174] process: 1. PAN generates the sensing pilot frequency domain symbol `X[k]` at a reference of 15kHz.

[0175] 2. Pre-conversion processing: Map each 15kHz subcarrier to four consecutive 60kHz subcarriers, repeating the following pattern: `Y[4k]=X[k], Y[4k+1]=X[k], Y[4k+2]=X[k], Y[4k+3]=X[k]`.

[0176] 3. Perform a 4-point IFFT on `Y` to generate a time-domain signal with a 60kHz sampling rate, add an extended CP, and then transmit.

[0177] 4. The PAN is received and demodulated using a reference 15kHz FFT. Since the spectrum is composed of repeating subcarriers, the demodulation result is naturally equivalent to combining four subcarriers to recover `X[k]`.

[0178] 5. Perform matched filtering using locally generated pilot copies.

[0179] Effect: The PAN baseband processing is always based on 15kHz, and the actual transmitted 60kHz signal has better anti-Doppler performance. Waveform adjustment is transparent to the PAN.

[0180] Example 6: Resource Conflicts and Two-Step Deterministic Arbitration in Multi-Sensing Nodes This embodiment demonstrates how LPC resolves conflicts in transmit resources among multiple PANs through a two-step arbitration process in path C.

[0181] Scenario: PAN_A and PAN_B generate the same resource intent `Resource` at the same `T_anchor` (e.g., both pointing to PRB#10). PAN_A uses rule B, and PAN_B uses rule D.

[0182] Conflict Detection: LPC accurately predicts the resource intention `Resource_i` that each PAN will generate at each logical decision moment by running a PSSM replica synchronized with each PAN (running rule B for PAN_A and rule D for PAN_B). When LPC detects that multiple PANs' `Resource_i` points to the same physical resource at the same `T_anchor` moment, it determines that a resource conflict has occurred.

[0183] Arbitration process (taking sequence one as an example): 1. First step: Arbitration of user conflicts. The winner, PAN_A, is selected based on preset rules (such as priority, assuming PAN_A has higher priority).

[0184] 2. Second step: Resource availability check. LPC queries the global resource calendar to confirm that the winner's `Resource` is available at the corresponding `T_tx` time.

[0185] 3. Downlink Implicit Granting: LPC sends granting signals to the winner and silent instructions to the loser via the downlink.

[0186] 4. The winner, PAN_A, transmits a pilot signal at `T_tx`, while the loser, PAN_B, remains silent and waits for the next cycle to retry.

[0187] If the second order is adopted: LPC first temporarily reserves the conflicting resources, and then arbitrates the winner from the successfully reserved users to convert the temporary reservation into a formal occupation. Both orders are within the scope of protection of this invention.

[0188] Example 7: Multi-node cooperative sensing for pilot contamination elimination This embodiment demonstrates how, in path C, orthogonal pilot physical resources are allocated to different PANs using a dynamic parameter set to eliminate pilot pollution.

[0189] Scenario: Multiple PANs simultaneously transmit sensing pilot signals to detect the same area, using a combination of rules B and D.

[0190] process: 1. LPC allocates orthogonal resources to each PAN in a dynamic parameter set based on the number of PANs and the size of the resource pool: - Time-domain offset: PAN_A is emitted at symbol #0, and PAN_B is emitted at symbol #1; - Or comb tooth offset: PAN_A uses comb tooth 0, PAN_B uses comb tooth 2; - Or circular shift: PAN_A uses a circular shift of 0, PAN_B uses a circular shift of 6.

[0191] 2. PANs generate pilots according to their respective rules (rule B or rule D), but in actual transmission, they only place pilots on the allocated time domain / frequency domain / code domain resources.

[0192] 3. The receiving PAN is matched and filtered on the corresponding resources, which can perfectly separate the echoes of each transmitting PAN.

[0193] Example 8: Non-cooperative target detection and identification This embodiment demonstrates the use of sensing pilots to detect non-cooperative targets (such as unauthorized drones flying without any communication modules).

[0194] process: 1. PAN_A transmits pilot `P_A` at `T_tx` (a simplified design using rule D can be adopted).

[0195] 2. The target reflected signal is received by PAN_B. PAN_B uses a locally generated `P_rx_for_A` matched filter to detect the echo.

[0196] 3. PAN_B not only measures TOA and AOA, but also extracts the micro-Doppler features of the echoes for identification of drone types.

[0197] 4. Multiple PANs illuminate the target from different angles to obtain multi-view features, which are then reported to LPC.

[0198] 5. LPC uses machine learning models for target identification and threat assessment.

[0199] Example 9: Sensing and Communication Multiplexing – The Same Pilot Frequency is Used Simultaneously for Data Transmission and Target Detection This embodiment demonstrates that in path B, the communication pilot is also used as a sensing signal.

[0200] Configuration: - The UE establishes a DSF connection with the base station for data transmission. The UE uses rule B to generate communication pilots.

[0201] - The base station simultaneously activates its sensing function, utilizing the UE's uplink communication pilots to detect the environment.

[0202] process: 1. The UE sends an uplink data frame in `T_UL`, which includes a security pilot `P_UL` generated based on rule B.

[0203] 2. The base station not only receives the direct signal from the UE for data demodulation, but also opens an additional receiving window to listen for possible target reflected echoes.

[0204] 3. The base station uses the same `P_UL` replica to perform matched filtering on the echo. If a peak is detected, it is determined that a target exists.

[0205] 4. The base station can combine the uplink signals of multiple UEs to achieve multi-angle detection of the area.

[0206] Example 10: Generation and physical constraints of receive offset under path C (example of rule D) This embodiment details the generation method and physical feasibility of the receive offset `Δ_rx_i` under rule D.

[0207] Generation method: ``` Δ_rx_i = G(K_sec, T_anchor, "RX_OFFSET" || PAN_ID_i) mod T_slot_max ``` Where `T_slot_max` is the maximum slot length (e.g., 1ms), ensuring the offset is within the slot range. `G` is the cryptographic hash function.

[0208] Physical constraints: The system must ensure that `Δ_rx_i > Δ_tx_i + T_process + T_prop_max`, where `T_process` is the PAN's switching time from transmission to reception (including RF switching, baseband processing, etc.), and `T_prop_max` is the maximum possible round-trip propagation delay. When configuring sensing tasks, the network sets the output range of the `G` function and defines the receive window length `W_rx` (e.g., twice `T_prop_max`), ensuring that the echo always falls within the receive window regardless of when it arrives. If a generated `Δ_rx_i` does not meet the above constraints, the PAN can automatically adjust to the effective range according to preset rules (e.g., adding a base offset `Δ_base`). The impact of this adjustment on sensing performance can be absorbed by the system parameter design.

[0209] Effects: The receiving windows of different PANs are randomly and uniformly distributed along the time axis, reducing the probability of interference from simultaneous reception by multiple nodes while ensuring timing feasibility. Since `Δ_rx_i` is generated based on `T_anchor`, and `T_anchor` itself contains physical randomness, the position of the receiving window also has natural randomness, further enhancing the robustness of the system.

Claims

1. A deterministic sensing method integrating sensing and computation based on cryptographic primitives to generate sensing pilots, characterized in that, include: - The communicating parties share cryptographic materials and time-varying synchronization parameters in advance; - The transmitting node generates and transmits a sensing pilot sequence based on the cryptographic materials and synchronization parameters using a cryptographic deterministic function; - Cooperative receiving nodes independently generate identical sensing pilot copies based on the same cryptographic materials and synchronization parameters, which are used for matched filtering of the received signals to extract target reflection information; - The entire process does not require the transmission of pilot sequences through signaling interaction.

2. The method according to claim 1, characterized in that, The cryptographic material includes a session key K_sec, which is pre-established through an authentication and key negotiation process; the time-varying synchronization parameters include at least one of the following: physical layer broadcast timing, logical state output by the protocol security state machine, or unified anchor time T_anchor.

3. The method according to claim 1, characterized in that, The generation of the sensing pilot copy must use the exact same context information as the transmission pilot, which includes the transmission node identifier and direction label.

4. The method according to claim 1, characterized in that, It also includes the construction and operation of the protocol security state machine: - Both communicating parties share a dynamic security foundation DSF, which includes K_sec, an initial anchor point Init_Anchor, and a state machine rule Rule_ID; - Both parties evolve synchronously on the logical time axis based on the DSF Independent Operation Protocol Security State Machine PSSM, outputting a time-varying state S(t); - The sensing pilot is generated based on S(t).

5. The method according to claim 4, characterized in that, The state machine rule Rule_ID includes at least one of rule A, rule B, rule C, and rule D: - Rule A: Broadcast clock driven, state S(t) is generated based on physical layer broadcast timing; - Rule B: Hash chain driven, state S(t) is the hash chain value, which evolves through iterative hash function; - Rule C: Logic epoch driven, state S(t) is a public logic state, consisting of a logic counter and an epoch identifier; - Rule D: Physical anchor point driven, state S(t) is the instantaneous sampled value of T_anchor at the unified anchor point time.

6. The method according to claim 5, characterized in that, In rule D: - The unified anchor point time T_anchor is obtained by instantaneously sampling the local physical clock at the logical decision time and is used as the current state S(t); - The T_anchor contains microscopic uncertainties of the local physical clock, injecting physical randomness into the generated parameters; - The state S(t) = T_anchor can be directly used as a time-varying synchronization parameter to generate various communication parameters and sensing parameters, including but not limited to: * Sensing pilot sequences and communication pilot sequences; * Logical resource index and physical resource location; * High-level protocol parameters such as HARQ process number, beam identifier, and slot format indicator; * Timing parameters such as transmit / receive time offsets Δ_tx and Δ_rx; * Physical layer parameters such as scrambling sequence and interleaving pattern; - The generation of the parameters follows a unified formula `Param = F(K_sec, S(t), Context)`, where Context is used to distinguish parameter types and node identities.

7. The method according to claim 5, characterized in that, Multiple nodes participating in collaborative sensing can adopt different state machine rules, including any combination of rules A, B, C, and D. Synchronous collaboration can still be achieved based on the shared DSF. In particular, nodes using rule D can directly use their local T_anchor as their state value and work collaboratively with nodes using other rules without additional state transitions or signaling interactions.

8. The method according to claim 1, characterized in that, It also includes the determination and use of unified anchor point times: - The sensing node instantaneously samples the local physical clock at the moment of logical decision-making to obtain the unified anchor point time T_anchor; - Calculate the actual transmission time T_tx = T_anchor + Δ_tx and the start time of the reception window T_rx_start = T_anchor + Δ_rx based on the fixed offsets Δ_tx and Δ_rx; - The network side uses T_anchor as the benchmark for resource arbitration and availability checks.

9. The method according to claim 8, characterized in that, The fixed offsets Δ_tx and Δ_rx are generated by a cryptographic deterministic function based on K_sec, S(t) and the context, and the system ensures that Δ_rx > Δ_tx + T_process + T_prop_max, where T_process is the node processing time and T_prop_max is the maximum round-trip propagation delay.

10. The method according to claim 1, characterized in that, Three implementation paths are provided: - Path A: Based on the traditional 5G network architecture, the time-varying synchronization parameters are the system frame number (SFN) and the time slot number, and the shared key between base stations is distributed by the core network; - Path B: Based on the traditional 5G architecture, introduce the dynamic security foundation DSF, and the base station operation protocol security state machine PSSM outputs the logical state S(t); - Path C: Based on the decoupled architecture of the logical processing center LPC and the physical access node PAN, the PSSM runs on the PAN, the LPC is responsible for DSF configuration and global coordination, the PAN independently generates sensing pilots and transmit / receive times, and the LPC maintains a mapping table to absorb changes in physical topology.

11. The method according to claim 10, characterized in that, In path C, the LPC also maintains a dynamic physical layer parameter set, which includes at least one of the actual subcarrier spacing, frequency offset compensation value, and timing advance. The PAN performs physical layer processing based on the dynamic parameter set, but its baseband signal is always generated based on the reference physical layer parameters, thus achieving parameter decoupling.

12. The method according to claim 11, characterized in that, Before transmitting the sensing pilot, the PAN performs frequency offset pre-compensation on the signal based on the downlink frequency offset compensation value f_comp_dl in the dynamic parameter set; after receiving the echo, it performs frequency offset compensation on the received signal based on the uplink frequency offset compensation value f_comp_ul, making the Doppler frequency offset transparent to the sensing node.

13. The method according to claim 11, characterized in that, When the PAN transmits sensing pilot signals, if the actual subcarrier spacing is different from the reference subcarrier spacing, it adopts a pre-conversion method: first, a baseband signal is generated based on the reference subcarrier spacing, and then it is converted into a signal that conforms to the actual subcarrier spacing through resampling or frequency domain mapping before transmission.

14. The method according to claim 10, characterized in that, In path C, the LPC establishes and maintains a clock offset table for each PAN through periodic synchronization beacons or measurement reports. In the position calculation, the local time reported by each PAN is unified to the global time axis to eliminate the influence of local clock offset.

15. The method according to claim 14, characterized in that, When the PAN reports measurements, it includes the local unified anchor time T_anchor. The LPC uses a clock offset table to convert the T_anchor of each PAN into global time, thereby achieving accurate time difference of arrival calculation.

16. The method according to claim 10, characterized in that, In path C: - The PAN uses rule D to generate state S(t) = T_anchor, where T_anchor is the local physical clock value sampled instantaneously at the time of logical decision; - The LPC establishes and maintains clock offset tables for each PAN through periodic synchronization beacons or measurement reports; - In the location calculation, the LPC uses the clock offset table to unify the T_anchor reported by each PAN to the global time axis, eliminating the influence of local clock offset; - The PAN performs measurement calculations and reporting based on a unified global time axis.

17. The method according to claim 4 or 10, characterized in that, The initial anchor point Init_Anchor is defined using either absolute anchoring or relative anchoring. Absolute anchoring specifies a specific physical time, while relative anchoring specifies the offset relative to a common event.

18. The method according to claim 1, characterized in that, It also includes deterministic arbitration and authorization for resource conflicts: - The network side predicts the resource intentions of multiple sensing nodes. If a conflict occurs, a two-step deterministic arbitration is performed. - The order of the two-step arbitration can be conflict arbitration first and resource check second, or resource pre-check first and conflict arbitration third; - The network side generates authorization information based on the arbitration result and notifies relevant nodes via the downlink; - The receiving node determines whether it has been authorized and the authorized resource information based on the local DSF and the received downlink signals.

19. The method according to claim 18, characterized in that, The notification methods for the downlink authorization information include: - For air interface links, an implicit authorization method is adopted, and nodes determine whether to obtain authorization on the reserved downlink resources by the presence or absence of signals; - For power supply links, an explicit authorization method is adopted, in which the network side sends an explicit authorization message or broadcasts an authorization list containing node identifiers and authorization parameters.

20. The method according to claim 1, characterized in that, Communication services and sensing services at the same node can operate independently using different state machine rules and different fixed offsets, without interfering with each other.

21. The method according to claim 1, characterized in that, The sensing pilot signal is also used for communication data transmission, realizing the integration of communication and sensing.

22. A deterministic sensing system integrating sensing and computation based on cryptographic primitives to generate sensing pilots, characterized in that, include: - Key management unit, used to establish shared cryptographic materials among communication nodes; - Synchronization parameter acquisition unit, used to acquire time-varying synchronization parameters; - Pilot generation unit, based on cryptographic materials and synchronization parameters, generates sensing pilots through cryptographic deterministic functions; - Transceiver unit, used to transmit generated pilot signals or receive signals and perform matched filtering using locally generated pilot copies; - Measurement reporting unit, used to extract target information and report it; - Position calculation unit, used to calculate the target position by integrating measurements from multiple nodes.

23. The system according to claim 22, characterized in that, The system can be deployed in one of three architectures: a traditional 5G base station architecture, a base station architecture with integrated DSF, or an LPC+PAN decoupled architecture.

24. The system according to claim 23, characterized in that, In the LPC+PAN decoupled architecture, LPC also includes a clock skew management module, which is used to maintain the clock skew tables of each PAN and to unify the time during position calculation.

25. The system according to claim 23, characterized in that, In the LPC+PAN decoupled architecture, the PAN also includes a pre-conversion processing module, which is used to convert the baseband signal generated by the reference parameters into a form that conforms to the actual dynamic parameters before transmission.

26. A computer-readable storage medium having a computer program stored thereon, the program being executed by a processor to implement the method as claimed in any one of claims 1 to 21.

27. A communication device comprising a processor, a memory, and a transceiver, wherein the processor, when executing a program in the memory, implements the method as claimed in any one of claims 1 to 21.

Images