Search

CN-122002558-A - Unmanned cluster hybrid access method and system based on fusion of statistical priority and time division scheduling

CN122002558ACN 122002558 ACN122002558 ACN 122002558ACN-122002558-A

Abstract

A method for accessing unmanned cluster based on fusion of statistical priority and time-division scheduling includes such steps as building three-stage macro-super frame of beacon segment-competition segment-scheduling segment in time domain, introducing cyclic micro-frame mechanism in scheduling segment TDMA, dispersing micro-frame into two-dimensional orthogonal time-frequency resource matrix containing multiple frequency bands and time slots, obtaining three resource attributes, quickly obtaining dedicated resource coordinates, providing global view to synchronous resource, building global three-dimensional load view, implementing interference of micro time slot level to local node with extreme unbalance, reallocating macro-time-space resource by regulating boundary, and outputting optimal access threshold for whole network by cross-layer calculation and dynamic optimizing. The invention meets the high dynamic concurrency demands of multi-level tidal services.

Inventors

  • ZHANG SHUN
  • CHEN PING
  • MA JIANPENG
  • HE XIANDENG

Assignees

  • 西安电子科技大学

Dates

Publication Date
20260508
Application Date
20260316

Claims (10)

  1. 1. The unmanned cluster hybrid access method based on the fusion of the statistical priority and the time division scheduling is characterized by comprising the following steps of; Step 1, constructing a three-stage macro superframe frame of a beacon section, a competition section and a scheduling section on a time domain, introducing a cyclic micro-frame mechanism into a scheduling section TDMA, discretizing the interior of the micro-frame into a two-dimensional orthogonal time-frequency resource matrix containing a plurality of frequency bands and time slots, and acquiring three resource attributes; Step2, based on the macro super frame constructed in the step 1 and three resource attributes, the unmanned cluster node rapidly acquires the exclusive resource coordinates according to the logic identification allocated by the cluster head when accessing the network, and provides a global view for the use state of the synchronous resources of the whole network; Step 3, based on the global view, forcing the cluster periodic cooperative service to anchor to the 0 th micro-frame MF0 transmission of the scheduling segment TDMA, implicitly embedding a traffic intensity vector TIV in an MAC header extension field of the MF0 time period by an online node, wherein the TIV comprises a burst traffic intensity subdomain subjected to high-precision logarithmic mapping and a periodic service overflow pre-metering subdomain; step 4, based on the global three-dimensional load view, implementing micro time slot level intervention on local nodes with extremely unbalanced flow, and starting bidirectional circulation when the TDMA resources are stiff due to local flow surge; Step 5, after solving the local microcosmic congestion based on the step 4, dynamically adjusting the proportion of the competition section and the scheduling section to adjust the boundary of the competition section and the scheduling section to realize the redistribution of macroscopic time-space resources and realize the abnormal self-healing under high dynamic state; And 6, finishing macroscopic physical time domain boundary adjustment of the competition section and the scheduling section based on the step 5, performing microscopic precise control on conflict caused by high concurrency of differentiated QoS service in the competition section in an established framework, and outputting an optimal access threshold value for the whole network by means of cross-layer calculation and dynamic optimization according to the global high-fidelity load view obtained in the step 3.
  2. 2. The method for hybrid access of unmanned cluster based on the combination of statistical priority and time division scheduling according to claim 1, wherein in step1, the whole network time axis is divided into consecutive superframe periods The super-frame conforms to the three-stage architecture of beacon segment-contention segment-schedule segment in time domain, and adapts to the multi-stage QoS difference and tidal effect of unmanned cluster service: the system comprises a beacon section, a contention section SPMA, a scheduling section TDMA, a control section and a control section, wherein the beacon section is used for whole network synchronization and parameter issuing; A cyclic micro-frame mechanism is introduced into the TDMA section which is equally divided into With a period of And the micro-frame period constitutes a physical upper bound of the traffic delay constraint, Inside the micro frame, discretize into containing Each time slot Two-dimensional orthogonal time-frequency resource matrix of each frequency band, on-line node By its unique logical identification Mapping to obtain absolute orthogonal dedicated resource coordinates : 。
  3. 3. The method for hybrid access of unmanned cluster based on the fusion of statistical priority and time-division scheduling according to claim 2, wherein in step 1, the robustness of dynamic networking is improved and resource idling is eliminated by the following state maintenance and management mechanism; The cluster head node maintains and records the online state of the active condition of all member nodes of the whole network in real time locally, and dynamically divides the resource blocks in the two-dimensional orthogonal time-frequency resource matrix into three resource attributes, namely public resources, private resources and soft reservation resources; The cluster head divides dynamic attributes of resource blocks in the micro-frame two-dimensional matrix into three states, when a node is offline, corresponding resources are public resources and are classified into a public resource pool, when the node is online and active, the node is activated to be a private resource, the node does not monitor to directly send primary data, when the node is online but temporarily has no sending requirement, the resources are degraded to be a soft reserved resource, and other nodes are allowed to borrow transmission after monitoring LBT confirmation channels are idle before executing sending.
  4. 4. The method for hybrid access of unmanned cluster based on fusion of statistical priority and time-division scheduling according to claim 3, wherein in the step 2, the member node listens to the beacon frame to extract synchronization and threshold parameters after power-on, evaluates channel load in a competition section and sends an association request; If the response is lost due to channel interference, the node performs implicit matching and state self-healing by monitoring an association notification LAA field carrying a MAC address hash-logic identifier mapping relation in a beacon frame extension field of the next period; After network access is completed, the cluster head locally maintains a global resource state bitmap GRB based on 2-bit coding, maps and provides a global view support for subsequent cross-domain borrowing to public resources, private resources and soft reserved resources tri-state attributes of each time slot in the whole network synchronous two-dimensional resource matrix in real time.
  5. 5. The method for hybrid access of unmanned cluster based on the fusion of statistical priority and time division scheduling according to claim 4, wherein in the step 3, the traffic sensing mechanism based on collaborative frame piggybacking is specifically as follows: the cluster head has the accurate sensing capability on the QoS service load of the whole network, realizes the cross-domain resource scheduling with high timeliness and full dimensionality, and designs a flow state feedback mechanism with zero additional signaling overhead in a high dynamic scene; Utilizing strict periodic characteristics of cluster cooperative service, forcing the cluster cooperative service to be anchored in a No. 0 micro-frame MF0 of a TDMA scheduling section for exclusive transmission; During the MF0 period of the current superframe, all online nodes must embed a local logical traffic intensity vector TIV in the MAC header extension field; if the node currently has no cooperative task load, the protocol forces the node to send an idle heartbeat frame only containing physical header, MAC header and TIV information.
  6. 6. The method for hybrid access of an unmanned cluster based on statistical priority and time division scheduling fusion of claim 5, The burst flow intensity subdomain is used for representing the burst residence queue intensity of each level of business in the SPMA competition section, and adopts a high-precision logarithmic mapping algorithm to nonlinearly map the bit backlog rate of the bottom layer into a quantization index so as to give consideration to both low-load sensitivity and heavy-load wide range; the periodic service overflow prediction operator field is used for representing expected overflow quantity of each level of periodic service due to limited time slot capacity And after the cluster head is used for intensively analyzing the TIV vector of the whole network, a three-dimensional high-fidelity global load view of coverage space node-service type-flow intensity can be constructed locally.
  7. 7. The method for hybrid access of unmanned cluster based on the fusion of statistical priority and time division scheduling according to claim 6, wherein the bidirectional flow in step 4 is specifically: (1) The forward reservation borrowing comprises the steps that a high-load node applies for borrowing idle resources to a cluster head, if the soft reservation resources are obtained, the high-load node monitors LBT before sending to avoid the original master node absolutely, and if the soft reservation resources are not obtained, the high-load node does not monitor extremely low-delay direct sending; (2) Reverse overflow isolation, namely, depriving original road rights of the residual periodic long wake flow of the TDMA section, and mapping logic degradation into the lowest virtual priority, wherein when the residual periodic long wake flow of the undigested section floods into an SPMA competitive domain, a phase isolation algorithm based on random back-off is forcedly executed, so that the living space of key control signaling is prevented from being polluted: when a node bursts a large amount of high definition video traffic due to task demands, such a node becomes a high load node in the network.
  8. 8. The method for hybrid access of unmanned cluster based on the fusion of statistical priority and time-division scheduling according to claim 7, wherein in the step 5, the scheduling trend potential energy and the competition trend potential energy in the unmanned cluster network are calculated, and the differential potential energy under the current unmanned cluster network is obtained by calculating the difference between the scheduling trend potential energy and the competition trend potential energy; when the differential potential energy breaks down to stabilize the capacity, the physical boundary between SPMA and TDMA in the super frame is driven to dynamically move leftwards or rightwards, the boundary reset adopts a three-stage atomization view switching process of 'forecast-reciprocal-submit', and the out-of-step node triggers an active silence protection locking transmitter until an effective beacon is captured again, so that abnormal self-healing under high dynamic state is realized; In the step 6, a cluster head builds a cross-layer data mapping model for connecting the instantaneous load of the MAC layer and the multi-packet receiving MPR demodulation limit of the PHY layer, the access threshold of low-priority service is preferentially lowered by a reverse greedy peak clipping strategy to reduce the noise of the whole network, if the system falls into high-priority service self-congestion, the system is triggered to actively degrade and test, and finally the optimal service access threshold of the whole network is output to ensure the rigid QoS constraint of the high-priority service.
  9. 9. A system for implementing the unmanned cluster hybrid access method based on the fusion of statistical priority and time-division scheduling according to any one of claims 1-8, comprising a physical resource layer, a mechanism and control layer, and an algorithm decision layer; the physical resource layer comprises a multi-dimensional hybrid access superframe structure module, and strict space-time isolation of multi-dimensional heterogeneous service flows is realized by constructing a two-dimensional time-frequency resource matrix, so that a bottom physical bearing is provided for deterministic transmission of upper-layer high-quality services and dynamic tri-state circulation of resources; The mechanism and control layer comprises a high-robustness dynamic networking and global state maintenance module, a flow sensing and bidirectional cross-domain circulation module and a boundary reconstruction and view self-healing module driven by differential potential energy to form a local closed loop of sensing-decision-execution; The algorithm decision layer comprises a QoS dynamic threshold optimization module, and outputs an optimal service access threshold of the whole network in real time based on cross-layer load perception.
  10. 10. The system of claim 9, wherein the system architecture is tightly coordinated to form a global closed loop by the following information reporting and instruction issuing channels: The physical resource layer supports the idle/busy state of the channel at the bottom layer and the physical time slot allocation situation upwards to the mechanism and the control layer, the mechanism and the control layer extract the queue residence state and overflow difference value of the nodes of the whole network through implicit signaling convergence, construct a high-fidelity global three-dimensional load view, and then accurately report the quantized load and differential potential energy state under the cluster network to the algorithm decision layer; the algorithm decision layer receives the network load state, and then solves the anti-interference optimal admittance configuration by combining with the preset rigid QoS requirement, and transmits the calculated optimal threshold value instruction to the mechanism and control layer; After receiving the optimal threshold value, the mechanism and control layer translates the superframe boundary reconstruction decision calculated based on differential potential energy into a specific resource scheduling instruction and transmits the specific resource scheduling instruction to a physical resource layer; and the physical resource layer executes the bottom-layer radio frequency actions strictly according to the instructions, and finally completes the elastic expansion and deterministic access of the space-time physical resources of the whole unmanned cluster.

Description

Unmanned cluster hybrid access method and system based on fusion of statistical priority and time division scheduling Technical Field The invention belongs to the technical field of wireless communication and ad hoc networks, and particularly relates to an unmanned cluster hybrid access method and system based on integration of statistical priority and time division scheduling. Background When the unmanned cluster network performs complex cooperative tasks, the carried service has obvious multi-level quality of service (QoS) requirements and non-stable high dynamic characteristics. In order to meet the cooperative requirements of different application scenarios, the clustered network generally needs to process service flows of multiple dimensions in parallel. Depending on the quality of service (QoS) requirements, these traffic flows are strictly divided into five Priorities (PL), PL1 to PL5 in order from highest to lowest. The system comprises a safety control type service (PL 1) with extremely low time delay, a command signaling type service (PL 2), a periodical cluster cooperation type service (PL 3), a task load type service (PL 4) with large bandwidth and a background maintenance type service (PL 5). In addition to the static multi-level priority attribute, clustered network traffic may exhibit severe global hops, i.e., a "tidal effect," during different phases of takeoff aggregation, cruise transition, and task execution. In addressing the complex scenario described above, conventional MAC protocols expose significant adaptation limitations: (1) Traditional statistical priority competitive protocols (such as SPMA) are characterized in that when the network faces extremely heavy concurrency, the collision avoidance mechanism which completely depends on statistics and probability can cause the channel collision probability to rise exponentially, and the backoff avalanche is extremely easy to be caused. This results in the underlying critical life-saving services being at extremely high risk of packet loss, with thoroughly uncontrolled delay. In addition, existing protocols lack cross-layer awareness of the underlying physical layer (PHY) real channel state, non-differentiated access thresholds are very prone to cause high priority traffic to fall into "self-congestion" while low priority traffic faces "full starvation". (2) Conventional time division scheduling class protocols (e.g., TDMA) suffer from a lack of resource flexibility in the static rigid allocation pattern. In the face of non-stable burst service, large-area spectrum resources idle in idle due to light load, and nodes face serious buffer queuing backlog due to heavy load, so that agile response capability to burst microsecond signaling is thoroughly lost. (3) The limitation of the traditional fixed boundary hybrid access architecture is that the traditional thinking tries to compromise by statically dividing the competition section and the scheduling section, but the stiff space-time physical boundary cannot adapt to the flow mutation caused by macroscopic task switching. In particular, in the stage of burst signaling density, the solidified boundary often causes the SPMA competition section space to be limited, so that the SPMA competition section space cannot process massive signaling data in time, and serious local queuing backlog and congestion are caused. Due to deterministic resource allocation in the TDMA segment, transient local congestion cannot be resolved by borrowing idle resources, and overflow of large packet load can even reverse pollute the living environment of the contention domain. Disclosure of Invention In order to overcome the defects of the prior art, the invention aims to provide an unmanned cluster hybrid access method and system based on the fusion of statistical priority and time division scheduling, aiming at the problems of resource misallocation and cross-domain pollution caused by the traditional static boundary, according to the invention, a multidimensional hybrid superframe structure of a deep fusion statistical priority competition (SPMA) and time division scheduling (TDMA) mechanism is constructed, and a bottom lightweight state maintenance mechanism is matched, so that a space-time physical island is thoroughly broken, two-dimensional elastic on-demand circulation of time-frequency resources of a physical layer is realized, and the high dynamic concurrency requirement of a multistage tidal service is met. In order to achieve the above purpose, the technical scheme adopted by the invention is as follows: An unmanned cluster hybrid access method based on the fusion of statistical priority and time division scheduling comprises the following steps of; step 1, constructing a multi-dimensional hybrid access superframe and a two-dimensional time-frequency resource matrix: Constructing a three-stage macro superframe frame of 'beacon section-competition section-scheduling section' on a time domain, introducing