Deterministic wireless communication method, system and apparatus based on active resource planning

Abstract

The application discloses a kind of certainty wireless communication method, system and device based on active resource planning, belong to 6G mobile communication field.The application utilizes the global predictability of network side running all user equipment protocol safety state machine (PSSM) under dynamic security foundation (DSF) framework, accurately obtains the logical decision time sequence of each service flow in future planning window, directly allocates deterministic time-frequency resource for each logical decision time by resource allocation algorithm, generates resource use time table and issues to user equipment.User equipment transmits and receives according to table in planning window, without real-time signaling interaction.The application also provides three kinds of implementation modes of explicit planning, implicit planning (derived by planning key) and temporary planning (switching dedicated), and works cooperatively through resource pool division and mapping table version control and two-step arbitration.Resource use time table allows multiple user equipment to share the same physical resource at different logical times through time domain multiplexing, fundamentally eliminates resource collision, realizes zero signaling, zero queuing, sub-millisecond switching interruption deterministic communication, provides complete deterministic scheduling system from "security bottom" to "extreme performance" for 6G network.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, specifically relating to resource scheduling and mobility management in sixth-generation (6G) mobile communication systems. In particular, it relates to a method, system, and apparatus based on the Dynamic Security Foundation (DSF) framework, which enables zero-signaling and collision-free communication by pre-planning resources within a future time window on the network side. Background Technology

[0002] The Dynamic Security Foundation (DSF) and Protocol Security State Machine (PSSM) framework previously proposed by the applicant (see patent applications 2026100015014, 2026100463580, etc.) achieves zero signaling parameter generation through cryptographic synchronization, and transforms the resource collision caused by the "modulo funnel" effect into a bounded queuing process through downlink implicit authorization and two-step deterministic arbitration, thus realizing a mathematically provable upper bound on delay for the first time in scheduling-free multiple access.

[0003] However, two-step arbitration is a real-time response mechanism, where the network passively executes arbitration at each logical decision point. For periodic, predictable services (such as URLLC control signaling, eMBB video streaming, and periodic sensor reporting), as well as predictable mobility scenarios based on ephemeris tables, the network is fully capable of anticipating resource demands at multiple future logical points. Therefore, a proactive planning method is urgently needed to upgrade resource allocation from "passive response" to "proactive planning."

[0004] Existing technologies also include resource reservation schemes based on known trajectories (such as high-speed rail lines, autonomous driving routes, and satellite ephemeris), as well as semi-persistent scheduling (SPS) in 5G. These schemes typically reserve resources in advance by notifying the target base station via core network signaling before the UE moves, or configure semi-static resources at a relatively coarse period. However, this type of reservation is merely a coarse-grained reservation at the "resource pool" level—the target base station reserves a resource range or an access opportunity for the UE, but after the UE enters the target cell, it still needs to initiate a random access or scheduling request (such as RACH or SR) to compete for specific transmission resources within the reserved resource pool. This cannot guarantee transmission times accurate to the symbol level or micro-timeslot level, nor can it eliminate real-time signaling interactions during handover. Furthermore, these schemes lack the concept of "logical decision moment" in the DSF framework, and cannot achieve deterministic resource allocation for each UE and each service flow at each logical moment. Summary of the Invention

[0005] I. Definition of Key Concepts To make the technical solution of this invention clear and easy to understand, the following key concepts are first defined. Some concepts are inherited from the applicant's previous patent applications, and are hereby clarified to maintain the integrity of the system.

[0006] Dynamic Security Foundation (DSF): A cryptographically secure triple `(K_sec, Init_Anchor, Rule_ID)` shared by both communicating parties, used for the generation of all deterministic parameters and state synchronization. Here, `K_sec` is the shared security key, `Init_Anchor` defines the starting point of the logical timeline, and `Rule_ID` identifies the evolution rule of the protocol's security state machine. For detailed definitions, see previous patent application 2026100015014.

[0007] Protocol Security State Machine (PSSM): A deterministic state machine based on DSF (Digital State Function), evolving independently between communicating parties and outputting a time-varying logic state `S(t)`. Its evolution is decoupled from the physical clock and driven only by the logical ticks defined by `Rule_ID`. Logical decision moments refer to the discrete moments of the PSSM's output state, defined by `Rule_ID`, at which the two parties achieve behavioral synchronization. For detailed definitions, see the preceding patent application 2026100015014.

[0008] Micro-moment: This is a core concept of Rule C (Logical Era-Driven Rule) defined by the applicant in its previous patent application (2026100463580). Rule C divides each long period (epoch) into multiple micro-moments, each corresponding to a time-domain resource unit, which, together with the frequency-domain resource unit, forms a two-dimensional resource grid for low-power, high-concurrency access of massive IoT devices. In this invention, the micro-moment, as one of the basic granularities of time-domain resources, can be specifically defined by the N_symbol parameter in RULE_ID (for example, N_symbol=1 corresponds to one symbol, N_symbol=7 corresponds to half of a time slot, etc.).

[0009] PSSM Mirroring: A core mobility management mechanism defined by the applicant in its prior patent application (2026102049886). It refers to a state machine copy that the target base station instantiates locally, maintaining complete logical synchronization with the terminal's PSSM, based on the complete context information (DSF and current state snapshot `S(t0)`) of a terminal's PSSM obtained from the serving base station. This mirror is used to simulate the terminal's behavior after access within the target base station's localized parameter set environment and participates in the target base station's local deterministic arbitration, thereby achieving zero-signaling handover. In this invention, the PSSM mirroring mechanism can be used in conjunction with proactive planning: for scenarios where the handover time can be accurately predicted, proactive planning (mode C) is used to directly issue resource usage timetables; for scenarios where handover cannot be accurately predicted, PSSM mirroring handover is used as a fallback.

[0010] Logical resource index and physical resource index: - Logical Resource Index: An index within the logical resource space independently calculated by the UE or service flow based on its DSF state, with a value range of `0 ~ L-1`. This index does not directly correspond to physical resources and requires mapping table conversion.

[0011] - Physical Resource Index: A unique identifier that directly identifies physical layer time-frequency resources, determined by both the frequency domain location (starting PRB index, number of PRBs N_PRB) and the time domain location (starting symbol index, number of symbols N_symbol). The values ​​of N_PRB and N_symbol are determined by the RULE_ID configuration of the service flow. For specific determination methods, please refer to the applicant's prior patent application, "A Method and System for QoS-Aware Fine-Grained Resource Mapping and Adaptive Configuration Based on Dynamic Security Foundation".

[0012] - Mapping table: A data structure maintained on the network side, used to map logical resource indexes to physical resource indexes. It can associate UE_ID and LCID to achieve service flow granularity.

[0013] Resource Usage Schedule: The network side allocates a set of time-frequency resource units to each service flow of each UE within each logical decision time within the planning window. This is represented as `Schedule[UE_ID][LCID][t]`, where `t` is the logical time number. This schedule is sent to the user equipment via signaling, and the user equipment sends and receives the schedule according to the schedule within the planning window.

[0014] Planning window: A continuous logical time interval containing `N` logical decision moments. Its length `N` can be dynamically configured based on factors such as service cycle, channel coherence time, and UE mobility speed. After the window ends, the UE can either revert to the two-step arbitration mode or start a new round of planning.

[0015] Planning key (K_{plan}}): A random seed generated by the network side for implicit planning, used to drive the pseudo-random function, enabling the UE to independently calculate the resource usage schedule. The planning key is only used for resource allocation and does not involve user data encryption; its security requirements are lower than `K_sec`.

[0016] Logic Processing Center (LPC) and Physical Access Node (PAN): This invention applies to LPC+PAN separation architectures, where the LPC is centrally deployed at ground stations or the core network, responsible for logic processing and planning; the PAN is deployed on satellites, base stations, or UAVs, responsible for physical layer transmission and reception. This invention also applies to integrated base station architectures where the LPC and PAN are combined, in which case the LPC function is implemented by the baseband processing unit (BBU) or distributed unit (DU) of the base station.

[0017] II. Purpose of the Invention The purpose of this invention is to provide a deterministic wireless communication method based on proactive resource planning, which inherits the advantages of DSF's zero signaling and determinism. This method enables the network side to allocate future resources in advance based on global information, fundamentally avoiding collisions, reducing real-time processing pressure, and working in conjunction with the original two-step arbitration to form a complete deterministic scheduling system.

[0018] III. Technical Solution To address the aforementioned technical problems, this invention provides a deterministic wireless communication method based on active resource planning (DRF). This invention leverages the global predictability of the DRF framework—the network side operates a copy of the PSSM for all UEs, enabling precise knowledge of the existence of each service flow for each UE at every logical decision point in the future—to generate a resource usage schedule. This schedule specifies which specific time-frequency resource will be used for each service flow of each UE at a precise logical point, much like a "subway timetable." UEs transmit and receive according to the schedule within the planning window, without requiring any real-time signaling or random access.

[0019] To achieve the above objectives, the present invention adopts the following technical solution.

[0020] 3.1 Core Steps of Proactive Resource Planning Step 1: Obtain the time sequence of logical decision-making The network side uses copies of the PSSMs of all UEs it maintains to pre-run the PSSMs of each service flow for each participating UE, accurately determining the logical decision-making time sequence of each service flow within the future planning window (i.e., at which logical moments each service flow needs to send and receive). These moments are determined by the `Rule_ID` in the DSF triplet of each service flow, and can be fully predicted by the network side.

[0021] Step 2: Generate a resource usage schedule On the network side, using `(UE_ID, LCID, logical time t)` as an index, and combining global resource status (including interference map, UE location, movement trajectory, ephemeris, beam coverage, MIMO port status, transmit power limit, service priority, service quality requirements, etc.), and considering factors such as interference map, geographical isolation constraints, and timing advance differences, a deterministic time-frequency resource is allocated to each index using a resource allocation algorithm (e.g., graph coloring algorithm based on interference map, maximum independent set search, heuristic scheduling, or artificial intelligence algorithm), generating a resource usage schedule `Schedule[UE_ID][LCID][t]`. The granularity of the time-frequency resource is determined by the N_PRB (number of physical resource blocks) and N_symbol (number of OFDM symbols) parameters in the RULE_ID of the service flow. Specifically, N_PRB consecutive PRBs are occupied in the frequency domain, and N_symbol consecutive symbols are occupied in the time domain (for rule C scenario, N_symbol can correspond to one or more micro-times). This timetable specifies the precise resource units for each logical decision time (e.g., "At logical time 100, the traffic flow of UE_A with LCID=1 uses 2 PRBs starting from PRB 5 and 14 symbols starting from symbol 0").

[0022] Step 3: Issue the timetable The network side sends the resource usage schedule to the UE via signaling. There are two methods of sending this schedule: explicit planning and implicit planning (see Section 3.2 below). Regardless of the method used, the signaling can be encrypted using a security key \(K_{\text{sec}}\) to ensure the confidentiality of the resource allocation information.

[0023] Step 4: UE sends and receives data according to the table Within the planning window, at each logical decision moment, the UE obtains the time-frequency resources by looking up a table or calculating based on its own UE_ID, LCID and the current logical time offset `t`, and directly sends and receives data without any real-time signaling interaction.

[0024] 3.2 Construction method of planning key (key to implicit programming) To achieve implicit planning (i.e., instead of issuing a complete timetable, issuing a key for the UE to derive it itself), the network side needs to construct a planning key \(K_{\text{plan}}\) such that the resource allocation results generated by the predetermined pseudo-random function \(F\) are basically consistent with the target timetable. The specific construction method is as follows: - The network side randomly generates a root key \(K_{\text{plan\_root}}\) and sets a counter \(c\) (the value range is 0~C-1, where C is the preset upper limit of the number of searches, such as 256).

[0025] - The network side defines the resource mapping function (G) as follows: \[\text{Resource}(t, \text{UE\_ID}, \text{LCID}) = \text{Map}\left( \text{Hash}(K_{\text{plan\_root}} \| c \| t \| \text{UE\_ID} \| \text{LCID}) \bmod R \right)\] Where \(R\) is the total number of available physical resources, Hash is the cryptographic hash function, and Map is the function that maps hash values ​​to specific time-frequency resources.

[0026] - The value of the network-side traversal counter \(c\) is used to calculate the corresponding resource allocation result and compare it with the target timetable. The \(c\) with an acceptable conflict probability is selected as the final parameter (due to the uniform distribution characteristics of the pseudo-random function, when the counter value space is large enough, the conflict probability can be calculated by mathematical expectation and is lower than the preset threshold, such as 0.1%).

[0027] The network side sends the planning parameters (K_{plan_root}}, the selected counter (c), the logical start time of the planning window (T_{start}}), and the window length (N) to the UE via signaling. To enhance security, the planning parameters (K_{plan_root}}) and (c) can be encrypted using a security key (K_{sec}}).

[0028] - After receiving the data, the UE decrypts it locally using \(K_{\text{sec}}\) to obtain \(K_{\text{plan\_root}}\) and \(c\). At each logical moment, it calculates \(G(t, \text{UE\_ID}, \text{LCID})\) to obtain the time-frequency resources, and then directly transmits and receives data.

[0029] 3.3 Three Implementation Modes This invention provides three active planning implementation modes that can be used independently or in combination: Mode A: Explicit planning (directly issuing resource usage schedules) 1. Generate a resource usage schedule on the network side.

[0030] 2. The timetable is sent to the UE via a single downlink signaling message (such as RRC reconfiguration, MAC CE, downlink pilot control frame, etc.). For security reasons, this signaling message can be encrypted with \(K_{\text{sec}}\).

[0031] 3. The UE can directly look up tables and send / receive data within the window.

[0032] 4. After the window ends, the UE reverts to the two-step arbitration mode.

[0033] Mode B: Implicit Programming (Programming Key Derivation) 1. Generate a resource usage schedule on the network side.

[0034] 2. On the network side, construct the planning parameters (root key (K_{plan_root}}) and counter (c) as described in Section 3.2.

[0035] 3. Send the planning parameters, \(T_{\text{start}}\) and window length \(N\) through a single signaling.

[0036] 4. The UE decrypts the planning parameters and calculates \(G(t, \text{UE\_ID}, \text{LCID})\) at each logical moment to obtain the resources, and then directly sends and receives data.

[0037] 5. After the window ends, the UE reverts to the two-step arbitration mode.

[0038] Mode C: Temporary Planning (Dedicated to Switching) Suitable for predictable handover scenarios (such as satellite handover based on ephemeris tables, and base station handover along high-speed rail lines): 1. When the network sends a handover command to the UE, it plans a resource usage schedule for the UE for several logical moments after the handover, based on the resource status of the target cell and the predicted trajectory of the UE.

[0039] 2. The switching command carries this timetable (or planning parameter) and can be encrypted with \(K_{\text{sec}}\).

[0040] 3. After the UE switches to the target cell, it can directly send and receive data according to the table within the planning window without initiating random access or participating in two-step arbitration.

[0041] 4. After the planning window ends, the UE will fall back to the standard two-step arbitration mode of the target cell.

[0042] 3.4 Coordination of Resource Pool Allocation and Two-Step Arbitration Within the planning window where proactive planning takes effect, the network side divides the physical resource pool into a planning resource pool and an arbitration resource pool. The planning resource pool is strictly allocated to planning UEs according to the resource usage schedule, while the arbitration resource pool is reserved for UEs that need to arbitrate in two steps. The two resource pools are orthogonal or quasi-orthogonal in the time, frequency, or spatial domains (e.g., spatial isolation is achieved through beamforming), fundamentally avoiding resource collisions between the two types of UEs.

[0043] When the network side performs proactive resource planning, it updates the resource mapping table used for two-step deterministic arbitration based on the planning results. Specifically, this includes: - The network side maintains a mapping table for two-step arbitration UEs, which maps to physical time and frequency resources with `(UE_ID, LCID, logical resource index)` as the key.

[0044] - When a portion of physical resources are actively occupied, the network side updates the mapping table, remapping the logical index corresponding to the occupied physical resources to the free resources.

[0045] - The network side periodically sends out updated mapping tables via broadcast or multicast signaling, carrying the logical time range in which the mapping table takes effect (e.g., \([T_{\text{start}}, T_{\text{end}}]\)). When the two-step arbitration UE performs resource intention calculation at logical time \(t\), it needs to select the corresponding version of the mapping table based on the current time.

[0046] - If the UE does not receive the new mapping table, the old mapping table will be used by default. However, the network side will automatically avoid the resources occupied by the planning in the two-step arbitration UE's resource mapping within the planning window (this is achieved by setting the planning resource mapping table to the "invalid / reserved" entry in the two-step arbitration mapping table).

[0047] - In the two-step arbitration, the UE still uses the original logical resource pool size to calculate the logical resource index within the planning window, but automatically avoids the occupied resources after the table lookup, thereby reducing the probability of arbitration failure.

[0048] For UEs not included in the long-term plan, the network side can still provide temporary proactive planning at critical moments such as handover to eliminate real-time arbitration during the handover process and further reduce interruption latency.

[0049] 3.5 Relationship with the existing DSF system and applicable scenarios This invention does not negate or replace the applicant's existing two-step arbitration and PSSM mirroring switching technology, but rather constitutes an enhancement layer built upon it. The specific relationships are as follows: - Two-step arbitration provides a deterministic fallback mechanism applicable to all scenarios, with zero real-time signaling and bounded latency, forming the foundation of the DSF system.

[0050] - PSSM mirroring (see previous patent application 2026102049886) solves the zero signaling mobility problem in inter-frequency / asynchronous scenarios, but still requires two-step arbitration after entering the target cell.

[0051] - Seamless switching of dynamic resource pools (see previous patent applications 202610313177X, etc.) solves the problem of zero interruption under continuous movement at the same frequency, and the UE is unaware of network topology changes.

[0052] This invention (proactive planning), building upon the previous three, further eliminates arbitration waiting and reduces interruption latency to the physical limit through temporary proactive planning for handover scenarios that can be accurately predicted. For handover scenarios that cannot be accurately predicted, PSSM mirror handover still provides a reliable fallback mechanism. The two complement each other, together forming a complete mobility management solution.

[0053] This invention, together with the two-step arbitration, constitutes a complete deterministic scheduling system, each applicable to different scenarios: - For predictable, high-load services (such as periodic URLLC control flow and continuous eMBB video stream), proactive planning is adopted to achieve zero queuing and extreme performance through resource usage schedules; - For unpredictable, sparse services (such as web browsing and random access), a two-step arbitration is adopted, and bounded queuing is used to achieve zero real-time signaling and flexible scheduling. - For handover scenarios, a temporary proactive planning (mode C) is adopted to eliminate handover interruptions while maintaining daily communication through a two-step arbitration process.

[0054] These four elements form a complete spectrum from "safety fallback" to "ultimate performance." The network side can flexibly select and combine them based on service characteristics, security requirements, channel conditions, and resource adequacy. In particular, for services with extremely high security requirements, two-step arbitration can be used first; for services with ultimate performance requirements, proactive planning can be used; the two can work together through mapping table updates without conflict.

[0055] IV. Beneficial Effects Compared with the prior art, the present invention has the following beneficial effects: 1. Eliminate collisions at the root: By planning in advance, the allocation of resources within the planning window is completely orthogonal, thus completely avoiding random collisions caused by the "modulus funnel".

[0056] 2. Significantly reduces real-time processing pressure: The network side does not need to perform real-time arbitration at every logical moment, but only needs to perform a calculation once at the beginning of the planning period.

[0057] 3. Zero signaling communication: In implicit planning mode, the UE only needs to receive the key once to independently deduce resources within the window; in explicit planning mode, the signaling frequency is extremely low.

[0058] 4. Support differentiated business: predictable business is included in the planning, and unexpected business retains a two-step arbitration process to form a unified and flexible service system.

[0059] 5. Perfectly compatible with the existing DSF framework: As an enhanced extension of DSF, it seamlessly collaborates with the original two-step arbitration, PSSM image switching, and dynamic resource pool.

[0060] 6. Timetable-level determinism: Specifying precise resources for each logical decision time, which is fundamentally different from existing trajectory-based resource reservation (which only reserves resource pools).

[0061] 7. Enhanced Zero-Interruption Handover: Through proactive temporary planning, the UE can directly transmit and receive according to the table after handover, with an interruption time of only radio frequency retuning (<0.25ms), which is better than PSSM mirror handover.

[0062] 8. Achieve precise planning by leveraging the global predictability of DSF: The network side learns in advance the future resource intentions of each UE and performs global optimal allocation, which is something that traditional communication systems cannot achieve.

[0063] 9. Efficient resource reuse: Through deterministic time-domain reuse, the resource usage schedule allows multiple UEs to share the same physical resource at different logical times under globally optimal arrangement, avoiding resource idleness caused by traditional exclusive resource reservation. Attached Figure Description

[0064] Figure 1 This is a schematic diagram of the overall process of proactive resource planning in this invention.

[0065] Figure 2 This is a signaling flowchart for the explicit planning (issuing resource usage schedules) of this invention.

[0066] Figure 3 This is a schematic diagram illustrating the principle of implicit programming (programming key derivation) of the present invention.

[0067] Figure 4 This is a schematic diagram illustrating the application of the graph coloring algorithm based on interference graphs of the present invention to resource allocation.

[0068] Figure 5 This is a diagram of the hybrid scheduling architecture of the present invention and two-step arbitration collaboration.

[0069] Figure 6 This is a schematic diagram of the switching based on ephemeris table active planning in the NTN scenario of the present invention.

[0070] Figure 7This is a functional block diagram of the network device (LPC) of the present invention.

[0071] Figure 8 This is a functional block diagram of the user equipment (UE) of the present invention. Detailed Implementation

[0072] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and embodiments. The described embodiments are some, but not all, of the embodiments of this invention.

[0073] Example 1: Explicit planning applied to LEO satellite periodic services (Rule C and micro-moments) Scenario: In a low-Earth orbit satellite constellation, 100 UEs need to report sensor data every second, with each report occupying one time-frequency resource unit (TFLU). Adopting rule C defined in the applicant's previous patent application (2026100463580), the logical period `T_logical=1s` is used, dividing each period into `N_s=100` micro-times (each micro-time is 10ms long). Each UE only needs to transmit within one micro-time in one period. A single satellite beam has 50 available PRBs in the frequency domain. Each service flow's RULE_ID is configured with N_PRB=1 and N_symbol=1 (i.e., occupying 1 PRB and 1 micro-time).

[0074] step: 1. Planning Generation: The network side utilizes copies of the PSSM of all UEs to accurately determine the logical decision time (i.e., the micro-time position of each UE) within each cycle. Since multiple UEs may be assigned to the same micro-time, the network side runs a graph coloring algorithm based on the interference graph to allocate conflicting UEs to different micro-times or different PRBs, generating a resource usage schedule `Schedule[UE_ID][LCID][Cycle]`, where each entry contains (micro-time, starting PRB, N_PRB=1, starting symbol, N_symbol=1).

[0075] 2. Explicit Distribution: Before the start of a cycle, the network side broadcasts the timetable for the next 100 cycles to all UEs via RRC signaling (each UE only receives entries relevant to itself). For security reasons, this signaling is encrypted using `K_sec`.

[0076] 3. Intra-window communication: Within each cycle, the UE wakes up at its own micro-time and directly transmits data using the time-frequency resources specified in the timetable, without any arbitration. The network side receives the data on the same resource.

[0077] 4. Window ends: After 100 cycles, the network side re-plans and sends the data again based on the new load and channel status.

[0078] Results: Achieves zero-collision, zero-signaling communication, UE duty cycle as low as 0.1%, and extremely low power consumption.

[0079] Example 2: Implicit Programming Applied to URLLC Control Signaling (Rule B) Scenario: An industrial robot needs to receive motion commands at a period of 0.25ms, with a latency of <0.5ms and zero jitter. Rule B is adopted, `T_update=0.25ms`. The robot's motion stream RULE_ID is configured with N_PRB=1 and N_symbol=14 to obtain the maximum channel estimation gain. Two-step arbitration is used when the robot simultaneously receives low-priority data (such as status monitoring).

[0080] step: 1. Planning Generation: The network side utilizes a copy of the robot's PSSM to accurately obtain the decision time sequence for the next 400 logical moments (corresponding to 100ms). Then, using `(UE_ID, LCID, t)` as the index, it runs a graph coloring algorithm based on the interference graph to directly allocate unique time-frequency resources (starting PRB, N_PRB=1, starting symbol, N_symbol=14) for each logical moment, generating a resource usage time table `Schedule[UE_ID][LCID][t]`.

[0081] 2. Planning Parameter Construction: The network side randomly generates a root key \(K_{\text{plan\_root}}\) and iterates through counters \(c\) (0~255) to calculate the corresponding resource allocation results, selecting \(c\) with the lowest conflict rate with the target timetable. \(K_{\text{plan\_root}}\) and \(c\) are then encrypted using `K_sec`.

[0082] 3. Signaling transmission: The network side transmits the encrypted planning parameters, \(T_{\text{start}}\) (current logical time) and window length `N=400` to the UE through a single RRC signaling.

[0083] 4. Intra-window communication: The UE decrypts to obtain \(K_{\text{plan\_root}}\) and \(c\). At each logical decision point within the planning window, it calculates \(G(t, \text{UE\_ID}, \text{LCID})\) to obtain the time-frequency resources and directly sends or receives commands. The network side calculates simultaneously and receives on the same resources. The entire process involves no arbitration and no signaling.

[0084] 5. Collaboration with two-step arbitration: The robot's low-priority operations still use the regular PSSM and two-step arbitration, without interfering with the motion flow.

[0085] 6. Window End: After 400 logical moments, the robot automatically reverts to the two-step arbitration. If the network side needs to continue planning, it can issue new planning parameters (or reuse old parameters to continue).

[0086] Results: Motion streams achieve zero-jitter hard deterministic transmission with a constant latency equal to the transmission time; low-priority services maintain operation through two-step arbitration.

[0087] Example 3: Active planning and zero-signaling handover based on ephemeris tables in NTN scenarios (LPC+PAN architecture) Scenario Description: The Low Earth Orbit (LEO) satellite constellation adopts an LPC+PAN separation architecture. The LPC is deployed at ground stations, and the PAN is deployed on satellites. The UE is currently served by PAN_A (source satellite), and the service flow includes: URLLC control flow (LCID=1, period 0.25ms, N_PRB=1, N_symbol=14) and eMBB video stream (LCID=2, period 1ms, N_PRB=2, N_symbol=7). According to the ephemeris, PAN_A will fly out of the UE's coverage area after 100ms, and PAN_B (target satellite) will enter. A cross-satellite handover needs to be completed.

[0088] step: 1. Service selection and planning decision: Based on the high periodicity and high determinism of UE services, as well as the accurate handover time prediction provided by the ephemeris, LPC decides to include both service flows in proactive planning.

[0089] 2. Planning Window Determination: LPC sets the planning window length to 50ms (i.e., 200 logical cycles, `T_dsf=0.25ms`), covering 25ms before and after the switchover. The window starts at the current logical time \(T_{\text{start}}\) and ends at \(T_{\text{start}}+200\).

[0090] 3. Logical Decision Timing and Resource Allocation: LPC utilizes its maintained PSSM copies of each UE service flow to accurately determine the existence of each service flow at each logical decision timing within the planning window. Then, using `(UE_ID, LCID, logical timing t)` as an index, it runs a graph coloring algorithm based on the interference graph. Considering handover and interference constraints, it directly allocates time-frequency resources to each index, generating a conflict-free resource usage timetable. Simultaneously, it uses the ephemeris table to calculate the precise logical timing `t_switch` (e.g., the 100th logical timing) where the handover occurs. The specific allocation is as follows: - For the first 100 logical moments (`t=0` to `t_switch-1`), resources are allocated to specific beams and PRBs in PAN_A.

[0091] - From the 100th logical moment onwards (`t_switch` to `199`), resources are allocated to specific beams of PAN_B and PRB.

[0092] - The algorithm simultaneously optimizes the beam direction of the target satellite PAN_B, calculates the azimuth and elevation angles of the UE relative to PAN_B after the handover time based on the UE's predicted trajectory, and encodes the beam direction parameters into the physical resource configuration.

[0093] 4. Distribution of Planning Results: LPC adopts an implicit planning mode, constructs planning parameters as described in Section 3.2, and distributes the encrypted parameters, \(T_{\text{start}}\) and window length `N=200` to the UE through a single RRC signaling.

[0094] 5. Communication within the planning window: Within the window, at each logical time `t`, the UE calculates \(G(t, \text{UE\_ID}, \text{LCID})\) to directly obtain the time-frequency resources (PAN identifier, beam, start PRB, start symbol) and performs transmission and reception. The UE does not need to know the beam direction and only transmits and receives according to the specified resources.

[0095] - The LPC side performs synchronous calculations and receives signals on the corresponding resources. Based on the joint optimization results, the network side automatically adjusts the beam direction of PAN_B before the handover time to ensure optimal signal quality for the UE after handover.

[0096] - The handover occurs automatically at the `t_switch` time. The UE does not need any handover signaling, only radio frequency retuning (or even no tuning at the same frequency). The interruption time is only the retuning time (<0.25ms).

[0097] 6. Collaboration with Two-Step Arbitration: Before the planning window begins, LPC updates the mapping table of the two-step arbitration UE based on the planning results, and sets the effective time range of the mapping table to `[T_{\text{start}}, T_{\text{start}}+200]`, remapping the logical indexes corresponding to the physical resources to be planned and occupied to idle resources. When the two-step arbitration UE calculates its resource intention within the window, it automatically avoids already occupied resources, thereby preventing invalid arbitration.

[0098] 7. Window End: After the window ends, the UE automatically reverts to the two-step arbitration mode. LPC can then be replanned based on subsequent ephemeris data.

[0099] Results: Handover is entirely planned, with zero signaling and only radio frequency retuning time (<0.25ms). Within the planning window, the UE enjoys conflict-free, zero-signaling deterministic communication.

[0100] Example 4: Comparison of Temporary Planned Switching (Mode C) and PSSM Mirror Switching Scenario: Within the same LEO satellite environment, the UE's service is an unpredictable, bursty activity (such as web browsing). While the network can still plan resources long-term, considering resource utilization efficiency (the UE may not send data during most planning periods), long-term planning would result in resource waste. Therefore, the network decides not to perform long-term proactive planning, but only to use temporary proactive planning (Mode C) during handover to eliminate handover interruptions, while maintaining daily communication through a two-step arbitration process. The UE's RULE_ID is configured with N_PRB=1 and N_symbol=14 (assuming small data packets).

[0101] step: 1. LPC decided not to conduct long-term proactive planning, and the UE will still use two-step arbitration before handover.

[0102] 2. When handover is triggered, LPC uses Mode C in the handover command: It plans a resource usage schedule for the UE for 10 logical time intervals after handover (e.g., on the target satellite PAN_B, it allocates 10 consecutive logical time-frequency resource units for each service flow of the UE, each unit being specified by the starting PRB, starting symbol, and N_PRB and N_symbol in RULE_ID). This schedule is encrypted with `K_sec` and sent with the handover command.

[0103] 3. After receiving the handover command, the UE decrypts the timetable, performs radio frequency retuning, and directly transmits and receives according to the timetable after entering the target cell, without any random access or two-step arbitration.

[0104] 4. After 10 logical time intervals, the timetable expires, and the UE automatically reverts to the two-step arbitration mode of the target cell.

[0105] In contrast, if the PSSM mirror handover method in the applicant's prior patent application (2026102049886) is used, the UE still needs to undergo two-step arbitration with the PSSM mirror of the target cell after entering the target cell, and the first transmission may wait for 1 to 2 logical moments before obtaining authorization. However, the temporary active planning of this invention completely eliminates this waiting time, and the interruption time is only radio frequency retuning.

[0106] Results: Temporary proactive planning reduces interruption latency to the physical limit in predictable switching scenarios, outperforming PSSM mirror switching.

Claims

1. A deterministic wireless communication method based on active resource planning, characterized in that, include: The network side determines a planning window, which includes at least one logical decision moment, and the logical decision moment is determined by the user equipment's protocol security state machine (PSSM) based on the dynamic security foundation (DSF). Before the planning window begins, the network side obtains the resource requirements of at least one service flow of at least one user device within the future planning window. Based on the resource requirements and global resource status, the network side uses a resource allocation algorithm to determine a resource usage schedule for each service flow of each user device at each logical decision moment within the planning window. The resource usage schedule specifies the time-frequency resources that the service flow should use at each logical decision moment. The network side sends the resource usage schedule to the user equipment via signaling; Within the planning window, the user equipment uses the specified time-frequency resources for communication at the corresponding logical decision time according to the resource usage schedule, without needing to perform real-time signaling interaction with the network side; The resource usage schedule allows multiple user devices to use the same physical resource at different logical times through time-domain multiplexing.

2. The method according to claim 1, characterized in that, The network side sends the resource usage schedule to the user equipment in the following ways: - Directly issue a mapping table containing the resource usage schedule; or - Issue planning parameters so that the user equipment can independently calculate the resource usage schedule based on the planning parameters.

3. The method according to claim 2, characterized in that, The planning parameters include at least a root key, a counter, a planning window start logical time, and a window length. The user equipment independently calculates the resource usage timetable based on the root key, the counter, the offset between the current logical time and the start logical time, and the user equipment's identifier and service flow identifier using a predetermined pseudo-random function.

4. The method according to claim 1, characterized in that, The network side decides whether to include the user equipment's service flow in the proactive planning based on at least one of the following factors: the user equipment's service attributes, the predictability of its movement trajectory, and the quality of service requirements.

5. The method according to claim 1, characterized in that, The resource usage timetable assigns a unique time-frequency resource unit to each logical decision time. The time-frequency resource unit is jointly determined by the frequency domain position (starting PRB index, number of PRBs) and the time domain position (starting symbol index, number of symbols), wherein the number of PRBs and the number of symbols are determined by the RULE_ID configuration of the service flow.

6. The method according to claim 1, characterized in that, The resource allocation algorithm adopts a joint optimization approach, which uses the resource usage schedule and the physical resource configuration on the network side as unified optimization variables for joint solution. The physical resource configuration includes at least one of beam direction, MIMO port configuration, and transmit power.

7. The method according to claim 1, characterized in that, When the network side performs proactive resource planning, it updates the resource mapping table used for two-step deterministic arbitration based on the planning results. The updated mapping table carries the effective logical time range, so that the user equipment participating in the two-step arbitration can avoid the time-frequency resources that have been planned and occupied when calculating the resource intentions at the corresponding time.

8. The method according to claim 1, characterized in that, Within the planning window, the network side divides the physical resource pool into a planning resource pool and an arbitration resource pool. The planning resource pool is allocated to user equipment included in the planning according to the resource usage schedule, while the arbitration resource pool is reserved for user equipment participating in the two-step arbitration. The two parts of resources are orthogonal or quasi-orthogonal in the time domain, frequency domain, or spatial domain.

9. The method according to claim 1, characterized in that, The network architecture includes a logical processing center (LPC) and physical access node (PAN) architecture. The LPC is responsible for determining the planning window, executing the resource allocation algorithm, and issuing the resource usage schedule. The PAN is responsible for physical layer transmission and reception within the planning window according to the resource usage schedule. Alternatively, the network architecture is an integrated base station architecture, in which case the LPC function is implemented by the baseband processing unit of the base station.

10. The method according to claim 1, characterized in that, The length of the planning window is dynamically configured according to the application scenario: a long planning window is configured for periodic business, and a short planning window is configured for switching scenarios.

11. The method according to claim 10, characterized in that, The resource usage schedule for the short planning window is sent to the user equipment in the handover command. After handover to the target cell, the user equipment can directly send and receive according to the schedule without participating in the two-step arbitration.

12. The method according to claim 11, characterized in that, The handover scenarios include satellite handover based on satellite ephemeris, ground base station handover based on trajectory prediction, or handover triggered by measurement reports.

13. The method according to any one of claims 1 to 12, characterized in that, The resource usage timetable is represented in the form of a three-dimensional table, and the index of the three-dimensional table is the user equipment identifier, the service flow identifier, and the logical time number.

14. A deterministic wireless communication system based on active resource planning, characterized in that, include: A network-side logical processing function entity is configured to perform the steps executed by the network side in the method of any one of claims 1 to 13; At least one physical access node is configured to perform the steps of the method described in any one of claims 1 to 13, which are executed by the physical access node. At least one user equipment is configured to receive the resource usage schedule and communicate within the planning window according to the resource usage schedule.

15. The system according to claim 14, characterized in that, The network-side logical processing function entity includes: Transceivers are used to communicate with the core network and physical access nodes; The processor is configured to determine the planning window, obtain resource requirements, execute resource allocation algorithms, and generate resource usage schedules. The memory is used to store user device context, global resource calendar, and mapping tables.

16. The system according to claim 14, characterized in that, The physical access nodes include: Transceivers are used for air interface communication with user equipment and for communication with network-side logical processing functional entities. The processor is configured to store a copy of the protocol-secure state machine of the storage service user equipment and transmit downlink signals on the specified time-frequency resources according to the received authorization instructions; The memory is used to store the protocol security state machine context.

17. A network device, characterized in that, It includes a processor, a memory, and a transceiver, wherein the processor executes a computer program stored in the memory to implement the steps performed by the network side in the method of any one of claims 1 to 13.

18. A user equipment, characterized in that, It includes a processor, a memory, and a transceiver, wherein the processor executes a computer program stored in the memory to implement the steps performed by the user equipment in the method of any one of claims 1 to 13.

19. A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method of any one of claims 1 to 13.

20. A communication device, characterized in that, The apparatus includes a processor and a memory, the memory storing instructions that, when executed by the processor, cause the apparatus to perform the method of any one of claims 1 to 13.

Images