CN120825384A - Dynamic slicing architecture and control method of self-organizing networks based on software-defined radio - Google Patents
Dynamic slicing architecture and control method of self-organizing networks based on software-defined radioInfo
- Publication number
- CN120825384A CN120825384A CN202511059444.7A CN202511059444A CN120825384A CN 120825384 A CN120825384 A CN 120825384A CN 202511059444 A CN202511059444 A CN 202511059444A CN 120825384 A CN120825384 A CN 120825384A
- Authority
- CN
- China
- Prior art keywords
- node
- nodes
- cluster
- task
- self
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Mobile Radio Communication Systems (AREA)
Abstract
The invention relates to the field of wireless communication and intelligent networks, in particular to an ad hoc network dynamic slicing architecture and a management and control method based on software defined radio, which comprise a basic communication layer, an intelligent control layer and a service execution layer. The basic communication layer adopts a multi-distributed multi-center clustering architecture, the intelligent control layer is responsible for intelligent management and resource scheduling of the self-organizing network, the intelligent control layer comprises a plurality of finger control nodes, the service execution layer adopts a micro-service-based architecture, and different network functions and service logics are packaged into independent micro-service modules. The invention solves the technical problems of strong plane coupling, stiff resource scheduling, insufficient survivability and high hardware dependence in the prior art.
Description
Technical Field
The invention relates to the field of wireless communication and intelligent networks, in particular to an ad hoc network dynamic slicing architecture based on software defined radio and a management and control method.
Background
In many fields such as modern communication, intelligent transportation, industrial Internet of things, etc., the self-organizing network and the control method thereof have important uses and values. The self-organizing network and the control method thereof are not only used for autonomous construction, dynamic adjustment and optimization of the network, but also used for changeable connection requirements under different scenes, greatly improve the flexibility of the network and the adaptability, and have important value and significance for improving the overall performance and the operation efficiency of the system.
The common construction mode of the traditional self-organizing network technology is mainly based on a fixed topological structure and a traditional routing algorithm, a distributed control mechanism is adopted, the connectivity of the network is maintained through negotiation and cooperation among nodes, and the control method focuses on static allocation of static network resources of network resources and simple flow control so as to ensure basic communication function realization.
However, existing ad hoc network technologies expose many drawbacks when dealing with complex scenarios such as multi-tasking, heterogeneous traffic mix transmission, and complex electromagnetic environments. The traditional self-organizing network service data, routing information and management and control instructions share transmission resources, so that delay jitter is aggravated under high load, control signaling blocking even occurs, differentiated service capability aiming at different tasks is lacked, real-time changing service requirements (such as low-delay communication of unmanned aerial vehicle clusters and mass data return of ground sensors) are difficult to dynamically adapt, in addition, when nodes fail or environment is interfered, a path reconstruction of the traditional self-organizing network depends on a static algorithm, optimal topology is difficult to quickly generate to ensure normal operation of the network, and a communication module with a fixed function cannot flexibly support novel protocols and waveforms, so that expandability and technical iteration speed of the network are limited. In summary, the traditional self-organizing network has the defects of strong planar coupling, stiff resource scheduling, insufficient survivability and high hardware dependence, and is difficult to meet the strict requirements on network elasticity and service quality under the development of the fusion of 6G communication, edge computing and artificial intelligence technology.
Disclosure of Invention
The application aims to provide an ad hoc network dynamic slice architecture and a management and control method based on software defined radio, which solve the technical problems of strong plane coupling, stiff resource scheduling, insufficient survivability and high hardware dependence in the prior art.
In order to solve the technical problems, the application adopts the following scheme:
The self-networking dynamic slicing architecture based on the software defined radio is characterized by comprising a layered three-dimensional network plane system;
the layered three-dimensional network plane system comprises a basic communication layer, an intelligent control layer and a service execution layer;
the base communication layer adopts a multi-distribution multi-center clustering architecture, and cluster nodes of the base communication layer comprise cluster head nodes and common nodes outside the cluster head nodes; generating a cluster structure based on the comprehensive performance of the cluster nodes, wherein the cluster head nodes elect by adopting a multi-round voting mechanism based on the comprehensive performance of the cluster nodes, and the comprehensive performance comprises signal strength, computing capacity, storage resources and residual energy;
The intelligent control layer comprises a plurality of finger control nodes, and each finger control node is provided with a phased array antenna and an AI decision engine; when the basic communication layer detects that the self-organizing network structure or configuration information is changed, the cluster head nodes send the change information to the command nodes of the intelligent control layer, and each command node of the intelligent control layer sends a control command to each cluster head node of the basic communication layer according to the change information to reprogram the self-organizing network structure, so that the intelligent control layer manages the basic communication layer;
The intelligent control layer sends out control instructions, the service execution layer receives and executes the control instructions, and the intelligent control layer monitors the execution process of the service execution layer and receives feedback signals of the service execution layer;
the micro-service module consists of sub-network nodes, the sub-network nodes cooperate to form a micro-service module for executing specific task functions and services, and the micro-service module comprises a reconnaissance task module, a relay task module and an attack task module.
Preferably, each cluster node of the basic communication layer is internally provided with a GPS receiving module, and receives satellite signals in real time through the built-in GPS receiving module to obtain the position information of the cluster node;
When information interaction is performed, the information package sent by the cluster node comprises time and position information of the cluster node, and after the rest cluster nodes receive the information package, the time is synchronized through the self-adaptive filtering algorithm and the multipath compensation algorithm, so that the time references of the cluster nodes are the same, and the communication and cooperative work of the cluster nodes under the same time reference are ensured.
Preferably, the communication protocol of the self-organizing network adopts a mixed routing protocol, and the mixed routing protocol comprises an optimized link state routing protocol and an on-demand distance vector routing protocol-enhanced type;
In the running process of the self-organizing network, one of the two communication protocols is dynamically selected as a dominant route protocol, wherein the dynamic selection rule is that when the self-organizing network topology is stable, an optimized link state route protocol is selected as a dominant route to establish a global optimal route, and the on-demand distance vector route protocol-enhanced auxiliary main communication protocol is used as the dominant communication protocol.
Preferably, the communication signal transmission waveforms of the self-organizing network comprise OFDM and SC-FDE, the two signal transmission waveforms are switched according to real-time channel state information, OFDM is selected when the RMS time delay spread is greater than 100ns and the available spectrum resource is smaller than 30%, the carrier frequency offset is greater than 10% subcarrier spacing, the SC-FDE is switched, and the real-time channel state information comprises multipath intensity, signal to noise ratio and frequency offset.
Preferably, the service execution layer adopts a micro-service-based architecture, self-organizing network sub-network nodes which are responsible for different network functions and service logics are packaged into independent micro-service modules at the service execution layer, the micro-service modules which are responsible for the same service can form a task sub-network, the task sub-network loads and unloads corresponding micro-service modules according to the creation and destruction requirements of new tasks, the micro-service modules comprise a reconnaissance task module, a relay task module and an attack task module, and the reconnaissance task module, the relay task module and the attack task module are respectively responsible for image acquisition and transmission, data relay and target attack.
Preferably, the control method comprises the following steps:
step S1, building the architecture of a basic communication layer, an intelligent control layer and a service execution layer;
Step S2, according to the requirements of different planes of the self-organizing network, determining the frequency spectrum ranges and access modes of the different planes according to plane-level resource allocation rules;
And step S3, monitoring the node state and the link quality, and performing survivability enhancement treatment after the node fails or the link is terminated, wherein the self-organizing network is specifically used for continuously monitoring the node state and the link quality, and when detecting that a certain node fails or the link is interrupted, immediately triggering a route repair algorithm, and recalculating the link path based on the route repair algorithm. After the new link path calculation is completed, the routing table of the affected node is updated, and the update information is broadcast to the relevant nodes. Meanwhile, the affected node caches data, and after a new path is established, the cached data is retransmitted to the target node.
Preferably, the step S1 is implemented by establishing a basic communication layer, an intelligent control layer, and a service execution layer according to the following steps:
step S1.1, initializing a node of a basic communication layer and constructing topology;
Each node of the self-organizing network periodically broadcasts a broadcast frame containing self identification, position and energy state, and adjacent nodes receive the broadcast frame, extract and record information of the transmitting node; meanwhile, each node transmits an active detection frame to the adjacent nodes, the adjacent nodes respond to the active detection frame, the nodes mutually transmit and receive node information, and an initial topological graph of the self-organizing network is constructed based on the collected adjacent node information;
s1.2, dynamic architecture and optimization of an intelligent control layer;
The control node identifies the cluster head node of each cluster in the self-organizing network based on the initial control link, the control node configures the area controller acted by the cluster head node according to the scale of each cluster, the distribution and service requirements of the node, the area controller collects the state data of the nodes in the cluster in real time, performs preprocessing on the collected data, extracts key information and compresses the data, can continuously monitor the state of the self-organizing network by a control layer, adjusts a control strategy by the control node after the state of the self-organizing network changes, generates a new control command, sends the control command to the area controller through the phased array antenna, converts the control command into a specific configuration command, distributes the specific configuration command to the common node in the cluster, and performs the operation of the command after the common node receives the configuration command, feeds back the execution result to the area controller, gathers the feedback information and feeds back the feedback information to the control node to form closed loop control;
S1.3, service execution layer dynamic task deployment and management;
The command node receives an external command, analyzes the command to obtain task parameters, wherein the task parameters comprise task type, execution time, target area, time delay and bandwidth requirement, and adds the task into a scheduling queue according to task priority. And screening network nodes meeting the conditions according to the resource requirements of the tasks, and finally distributing resources such as communication frequency bands, time slots, bandwidths and the like to the network nodes according to the task types and the network node distribution planning subnet topological structure. After the configuration is completed, the network node enters an activated state, and the sub-network is put into use to start executing tasks;
Preferably, the node initialization and topology construction of the base communication layer in the step S1.1 is implemented according to the following steps:
s1.1.1, constructing a self-discovery and neighbor node list of a self-organizing network node;
Periodically broadcasting a broadcasting frame containing self identification, position and energy state by nodes in the self-organizing network, extracting and recording information of a transmitting node after receiving the broadcasting frame by adjacent nodes, wherein the information of the nodes comprises the identification, the position and the energy state, and constructing a preliminary neighbor node list;
Step S1.1.2, initializing a cluster structure and electing cluster head nodes;
after the preliminary cluster structure is formed, each cluster node in the cluster performs voting according to the comprehensive performance index of the cluster head candidate cluster node, and a plurality of cluster nodes with the highest vote count commonly serve as cluster heads to form a multi-center architecture with a plurality of cluster head nodes in the cluster.
Preferably, the dynamic architecture and optimization of the intelligent control layer in the step S1.2 is implemented according to the following steps:
S1.2.1, initializing and scanning a command node;
And activating the command node in the construction process of the intelligent control layer, performing full-network scanning by the command node by using a phased array antenna, collecting topology information of the self-organizing network, and rapidly acquiring state information such as spatial position information, node residual energy and the like of all nodes in the self-organizing network by using the phased array antenna.
Step S1.2.2, generating an initial control link according to the self-organizing network topology information of step S1.2.1;
After obtaining the topology information of the whole network of the self-organizing network, generating an initial control node-area controller-common node control link;
S1.2.3, identifying cluster head nodes of each cluster in the self-organizing network by the command node, electing a regional controller and configuring;
after an initial control link is generated, the command node identifies cluster head nodes of all clusters in the self-organizing network, the cluster head nodes double as area controllers, the command node configures the area controllers according to the scale of each cluster, the distribution of the nodes and the service demand, and the configuration process comprises the steps of distributing control tasks, setting communication parameters and defining management strategies;
S1.2.4, collecting state data of nodes in the cluster by the regional controller and preprocessing;
The regional controller collects state data of nodes in the cluster in real time, pre-processes the collected data, extracts key information and compresses the data, and reduces the data quantity transmitted to the command node;
step S1.2.5, continuously monitoring the state of the self-organizing network;
The intelligent control layer continuously monitors the state of the self-organizing network, and after the state of the self-organizing network changes, the control layer instructs the nodes to adjust the control strategy and generate new control instructions, wherein the adjustment of the control strategy comprises the steps of reconfiguring the task of the regional controller, adjusting the link bandwidth and optimizing the data transmission path;
Step S1.2.6, distributing the control instruction in step S1.2.5 to common nodes in the cluster and executing;
The command node generates a new control command according to the adjusted control strategy, the control command is sent to the area controller through the phased array antenna, the area controller converts the control command into a specific configuration command and distributes the specific configuration command to the common nodes in the cluster, the common nodes execute corresponding operations after receiving the configuration command and feed back execution results to the area controller, and the area controller gathers feedback information and reports the feedback information to the command node to form closed loop control.
Preferably, the step S1.3 performs dynamic task deployment and management on the service execution layer according to the following steps:
step S1.3.1, external instruction receiving and analyzing;
the command node receives an instruction initiated by an entity outside the service execution layer of the self-organizing network and is used for guiding the self-organizing network to execute a specific task or perform a specific operation, and meanwhile, the command node analyzes the task instruction and extracts key task parameters;
step S1.3.2, matching task scheduling with resources;
Adding the tasks into a scheduling queue according to the priority and the emergency degree of the tasks, and then analyzing the demands of the tasks on network resources according to the task parameters;
Step S1.3.3, creating and configuring a task subnet;
According to the task demand in the step S1.3.2 and the screened subnet node distribution, creating a task subnet, and distributing independent space-time-frequency resource slices for the task subnet, and simultaneously, sending a configuration instruction to the task subnet by a command node;
step S1.3.4, executing tasks by the task sub-network;
After configuration is completed, the task sub-network executes tasks, and meanwhile, the command node monitors the task execution state in real time through the regional controller, wherein the task execution state comprises data transmission progress, node energy consumption and task completion conditions;
step S1.3.5, task completion and resource recovery;
when the task target is achieved or the task execution time is finished, judging that the task is finished, after the task is finished, issuing a resource recovery instruction by the command node, releasing the allocated resources including frequency spectrum, time slot, calculation resources and the like by the subnet node in the task subnet, and returning the subnet node to an idle state after the resources are recovered, so that the subnet node can be reused by other tasks.
The technical scheme of the application has at least the following advantages and beneficial effects:
1. in the invention, the self-organizing network is divided into three independent network plane systems of a basic communication layer, an intelligent control layer and a service execution layer. The base communication layer adopts a distributed multi-center clustering architecture, a cluster structure is generated based on an improved K-means algorithm, each cluster node in a cluster performs voting according to the comprehensive performance index of cluster head candidate cluster nodes, a plurality of cluster nodes with the highest number of votes jointly serve as cluster heads, and a multi-center architecture with a plurality of cluster head nodes in the cluster is formed.
The phased array antenna is configured by the finger control nodes of the intelligent control layer, and a hierarchical control chain of finger control nodes, regional controllers and common nodes is constructed. The architecture of the hierarchical control chain can realize efficient management and flexible configuration of the self-organizing network resources.
The service execution layer adopts a micro-service-based architecture, different network functions and service logics are packaged into independent micro-service modules, the micro-service modules can communicate with each other, and in addition, the micro-service modules can quickly establish and destroy task sub-networks according to service requirements. Meanwhile, the micro-service module supports dynamic loading and unloading, and can dynamically load required modules or unload modules which are not required any more according to the change of task demands, thereby improving the utilization rate and flexibility of self-organizing network system resources.
In general, the basic communication layer, the intelligent control layer and the service execution layer operate independently, cross-layer interference is avoided through resource isolation, and resources of the self-organizing network can be flexibly allocated.
2. In the invention, the self-organizing network continuously monitors the node state and the link quality, when a certain node failure or link interruption is detected, the link path can be recalculated to generate a new link path, meanwhile, the node which fails caches data, and after the new path is established, the cached data is retransmitted to the target node, so that the self-organizing network has the survivability, and the operation of the self-organizing network is not influenced when part of the nodes fail or the link is interrupted.
Drawings
FIG. 1 is a block diagram of a dynamic slice architecture of an ad hoc network in accordance with the present invention;
FIG. 2 is a flow chart of a control method of the present invention;
FIG. 3 is a flow chart of the architecture of the basic communication layer, the intelligent control layer and the service execution layer of the present invention;
FIG. 4 is a dynamic architecture and optimization flow chart of the intelligent control layer of the present invention;
FIG. 5 is a flow chart of the dynamic task deployment and management of the service execution layer according to the present invention.
Detailed Description
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the present invention will be briefly described below with reference to the accompanying drawings and the description of the embodiments or the prior art, and it is obvious that the following description of the structure of the drawings is only some embodiments of the present invention, and other drawings can be obtained according to these drawings without inventive effort to a person skilled in the art. It should be noted that the description of these examples is for aiding in understanding the present invention, but is not intended to limit the present invention.
It should be understood that although the terms first, second, etc. may be used herein to describe various modules, these modules should not be limited by these terms. These terms are only used to distinguish one module from another. For example, a first module may be referred to as a second module, and similarly a second module may be referred to as a first module, without departing from the scope of example embodiments of the invention.
It should be understood that for the term "and/or" that may appear herein, it is merely one kind of association relation describing the associated object, it may be indicated that there may be three kinds of relations, for example, a and/or B, it may be indicated that there are a alone, B alone, and there are three cases of a and B together, for the term "and" that may appear herein, it is another kind of association relation describing the other kind of association object, it may be indicated that there may be two kinds of relations, for example, a/and B, it may be indicated that there are a alone, there are two cases of a and B alone, and in addition, for the character "/" that may appear herein, it is generally indicated that the associated object is an "or" relation.
Examples:
Referring to fig. 1-5, an aspect of the present invention provides a software defined radio based ad hoc network dynamic slice architecture, including a hierarchical stereo network plane architecture.
Further, the layered three-dimensional network plane system comprises a basic communication layer, an intelligent control layer and a service execution layer.
Specifically, the basic communication layer adopts a multi-distribution multi-center clustering architecture, a cluster structure is dynamically generated based on an improved K-means algorithm, and cluster head nodes elect by adopting a multi-round voting mechanism according to the comprehensive performance of the cluster nodes. The basic communication layer is responsible for basic communication functions of the network, and ensures stable connection and data transmission between network nodes.
Specifically, the cluster nodes of the basic communication layer include cluster head nodes and common cluster nodes except the cluster head nodes.
More specifically, the cluster nodes are basic constituent units in the self-organizing network, specifically, the cluster nodes are various devices and terminals in the network, and the cluster nodes realize communication and data transmission tasks of the network through mutual cooperation. In the scheme, the multi-dimensional information of the cluster nodes is used as a feature vector, and an improved K-means algorithm is input to determine an optimal cluster structure. The multi-dimensional information of the cluster nodes comprises the geographic position, the residual energy and the communication capability of the coarse cluster nodes.
It should be noted that, the traditional K-means algorithm is a clustering algorithm based on partitioning, and the similarity of data points in cluster nodes is maximized and the similarity between cluster nodes is minimized by iteratively optimizing cluster centers and cluster members. The conventional K-means algorithm is known in the art, and those skilled in the art will understand the algorithm and therefore will not be described in detail herein.
More specifically, in the invention, the improvement of the traditional K-means algorithm is in three aspects of energy sensing, dynamic adjustment and multi-center election.
More specifically, the improvement on the energy sensing aspect refers to introducing a cluster node residual energy weight factor when calculating the similarity among cluster nodes, so that the cluster head node is prevented from being rapidly disabled due to excessive energy consumption, and the stable operation period of the self-organizing network is prolonged. The remaining energy weight factor is the ratio of the cluster node remaining energy E p to the cluster node initial energy E 0, namely:
After the residual energy weight factors are introduced, the algorithm considers not only the geographical positions of the cluster nodes but also the residual energy states of the cluster nodes when calculating the similarity among the cluster nodes. Specifically, when the similarity between two cluster nodes is calculated, the residual energy weight factors of the two cluster nodes are compared, and the cluster nodes with more residual energy are given larger weight, so that the cluster nodes with more residual energy have relatively larger influence in the cluster structure forming and cluster first-choice lifting processes, and the cluster nodes with more sufficient energy are guided to be selected as cluster head nodes by an algorithm, and the cluster head nodes are prevented from being rapidly failed due to excessive energy consumption.
More specifically, the improvement on the aspect of dynamic adjustment refers to monitoring the state change of cluster nodes in real time in the running process of the self-organizing network, triggering a K-means algorithm to re-cluster when the degree of the change of the cluster nodes reaches a certain threshold value, and dynamically adjusting the cluster structure to ensure that the self-organizing network is always in a good running state.
The threshold includes a cluster node energy change threshold and a cluster node position change threshold. When the rest energy of the cluster nodes is lower than 50% of the initial energy, the energy state of the cluster nodes is judged to be changed obviously, and then the re-clustering operation is triggered. For a mobile cluster node, when its location changes beyond a certain distance (in a wireless communication environment, the moving distance exceeds 20% of the communication radius) or the moving speed exceeds a set value (5 m/s), the quality of the communication link between the mobile cluster node and other nodes in the cluster is reduced, and the re-clustering is triggered to optimize the cluster structure.
More specifically, when the re-clustering operation is performed, an improved K-means algorithm is operated, and the latest state information of all cluster nodes is updated, wherein the latest state information comprises the cluster node residual energy and the space position of the cluster nodes. After the data updating of the improved K-means algorithm is completed, the improved K-means algorithm is reinitialized, the residual energy and the space position coordinates of the cluster nodes are respectively normalized, then the weighted average of each cluster node is calculated according to the weight coefficients of the residual energy and the space position coordinates, the cluster nodes are ordered according to the weighted average, the first K cluster nodes with the highest score are selected as initial cluster centers, then the improved K-means algorithm is operated, the distance between the cluster nodes and the cluster centers is calculated according to the residual energy and the space position coordinates of the cluster nodes, and the cluster attribution of the cluster nodes is updated according to the distance measurement formula:
Where D i,j is the distance from node i to cluster center j, E r,i is the remaining energy of node i, E 0,i is the initial energy of node i, D i,j is the position distance from node i to cluster center j, and α and β are weight coefficients representing the weights of the remaining energy and the spatial position coordinates of the cluster nodes, respectively. K is the number of cluster head nodes required, and can be selected according to the cluster head nodes required in the actual application process. After the algorithm iteration converges, updating the cluster structure according to the attribution relation of the new cluster center and the nodes, and reallocating network resources according to the new cluster structure so as to adapt to the dynamic change of the network.
More specifically, the improvement in multi-center election is to elect cluster head nodes using a multi-round voting mechanism. I.e. on the basis of the initial clustering, each cluster gets a temporary cluster head node. When the cluster head nodes are selected by the multi-round voting mechanism, each cluster node in the cluster performs voting according to the comprehensive performance index of the cluster head candidate cluster node, and a plurality of cluster nodes with the highest number of votes jointly serve as the cluster head, so that a multi-center architecture with a plurality of cluster head nodes in the cluster is formed, the fault tolerance and the load balancing of the self-organizing network are further enhanced, and the situation that a certain cluster node is selected as a unique cluster head only because of the fact that one performance index is highlighted is avoided. The comprehensive performance of the cluster head candidate cluster node comprises signal strength, computing power, storage resources and residual energy.
More specifically, the base communication layer incorporates a space-time reference synchronization technique. Specifically, a GPS receiving module is built in the cluster nodes of the basic communication layer, satellite signals can be received in real time, geographical position information of the cluster nodes can be obtained, meanwhile, a communication link based on Ethernet is established between the cluster nodes, time synchronization between the cluster nodes is realized by sending and receiving accurate time protocol messages, and a time synchronization process of a time protocol is optimized by adopting a self-adaptive filtering algorithm and a multipath compensation algorithm, so that time synchronization of a whole network nanosecond level is realized.
It should be noted that, the adaptive filtering algorithm and the multipath compensation algorithm are related art, and those skilled in the art will understand that the description is omitted here.
More specifically, the communication protocol of the ad hoc network adopts a hybrid routing protocol, and the hybrid routing protocol includes an optimized link state routing protocol (OLSR) and an on-demand distance vector routing protocol-enhanced (AODV-UU), and dynamically selects the optimized link state routing protocol (OLSR) or the on-demand distance vector routing protocol-enhanced (AODV-UU) as a dominant routing protocol during the operation of the network. Specifically, when the self-organizing network topology is relatively stable, a global optimal route is established based on an optimized link state routing protocol (OLSR), and when the self-organizing network local topology rapid change or specific traffic burst is detected, an on-demand distance vector routing protocol-enhanced (AODV-UU) rapid response is established, a temporary on-demand route is established, and a routing path of the OLSR is supplemented and optimized.
More specifically, the communication signal transmission waveform of the self-organizing network adopts Orthogonal Frequency Division Multiplexing (OFDM) and single carrier frequency domain equalization (SC-FDE) self-adaptive switching, and can dynamically select OFDM or SC-FDE as the current transmission waveform according to the real-time channel state information. Specifically, OFDM is selected when the channel multipath effect is significant and the spectrum resources are intense, and SC-FDE is selected when the channel frequency offset is large. The real-time channel state information includes multipath strength, signal to noise ratio, frequency offset. Specifically, OFDM is selected when the RMS delay spread is greater than 100ns and the available spectral resources are less than 30%, and is switched to SC-FDE when the carrier frequency offset is greater than 10% subcarrier spacing.
Specifically, the basic communication layer is used for establishing and maintaining physical connection of a network on one hand, ensuring stable transmission of data between nodes, providing a basis for network connection and data transmission for the intelligent control layer on the other hand, and managing and optimizing the basic communication layer by the intelligent control layer.
Specifically, the intelligent control layer is responsible for intelligent management and resource scheduling of the self-organizing network, and comprises a plurality of finger control nodes, wherein the finger control nodes are configured with phased array antennas and AI decision engines, and the intelligent control layer has strong communication capacity and intelligent decision capacity.
More specifically, when the basic communication layer detects that the network structure or configuration changes, the cluster head node sends the change information to the intelligent control layer. The intelligent control layer re-plans the network structure according to the change and sends the updated configuration information back to the basic communication layer. The command node refers to equipment with the capabilities of command, control, communication, calculation and the like in the communication network, is a core control point of the self-organizing network and is responsible for managing and coordinating other nodes in the network. The intelligent control layer sends control instructions to cluster head nodes of the basic communication layer through the command control nodes, wherein the control instructions comprise communication parameter adjustment instructions and routing optimization instructions. The network structure or configuration includes the addition, subtraction, movement, establishment of links, disconnection of links, link quality change, device failure, link failure of cluster nodes.
More specifically, the phased array antenna has high gain, wide beam coverage and fast beam switching capability, the working frequency range covers a main wireless communication frequency band, the phased array antenna can adapt to communication requirements under different service scenes, the gain of the phased array antenna can reach more than 30dBi, and the beam width can be adjusted between 10 degrees and 60 degrees. The AI decision engine has high concurrency processing capability and low delay response characteristics, can process state information of thousands of finger control nodes in real time, makes optimal resource allocation decisions, and ensures efficient operation of the network.
More specifically, the topology structure of the intelligent control layer is to construct an SDN hierarchical control chain of 'finger control node-regional controller-common node' based on a Software Defined Network (SDN) concept, wherein the regional controller is doubled by cluster head nodes of the basic communication layer.
More specifically, the common node represents a device with a basic communication function in the network, can perform data transmission with other nodes, is controlled by the area controller, and is basically the same device or terminal as the common cluster node of the basic communication layer, but the intelligent control layer emphasizes that the common node is controlled by the area controller, and the basic communication layer emphasizes that the common cluster node belongs to a certain cluster instead of the identity of the cluster head node, so that the common node is called differently.
More specifically, the command node is a core control node of the self-organizing network and is responsible for issuing control instructions to the regional controller so as to realize control of the common node, and in addition, the command node is configured with a phased array antenna and an AI decision engine and is responsible for resource management and instruction distribution of the global network. The control node communicates with the regional controllers through the phased array antenna, acquires topology information of the self-organizing network and the states of the nodes in real time, and coordinates the work of the regional controllers.
More specifically, the area controller is doubled by the cluster head node of the basic communication layer and is responsible for managing the communication and resource allocation of the nodes in the cluster. The regional controller communicates with the command node through the phased array antenna, receives the control command of the command node, converts the control command into a configuration command and sends the configuration command to the common node in the cluster. The regional controller has a certain local decision-making capability, and can be quickly recovered and adjusted under the condition of partial network abnormality.
More specifically, the common node performs communication and data transmission according to the configuration instruction of the regional controller, has computing capability and storage capability, and can perform simple data processing and buffering.
More specifically, the SDN controller operates on an independent server in the ad hoc network and is configured to cooperate with the regional controller to implement intelligent management and control of the ad hoc network. The SDN controller communicates with the regional controller by adopting an OpenFlow protocol, so that flexible control of network traffic and dynamic allocation of resources are realized. Specifically, the SDN controller adjusts link bandwidths and priorities among the regional controllers in real time according to changes of network traffic, so as to ensure efficient operation of the network.
More specifically, the control instruction transmission mechanism of the intelligent control layer is Time Division Duplex (TDD) matched with reserved access.
More specifically, the control instruction transmission adopts a Time Division Duplex (TDD) mode, a wireless channel is divided into an uplink time slot and a downlink time slot, the bidirectional transmission of data is realized through time slot allocation, and the proportion of the uplink time slot and the downlink time slot is dynamically adjusted according to the network service requirement and the communication environment. In the instruction transmission process, a reserved access mechanism is adopted to reserve special time slots and spectrum resources for key instructions. The reserved access mechanism ensures reliable transmission of key instructions by pre-allocating specific time slots and spectrum resources, and avoids instruction loss or delay caused by network congestion or other interference.
Specifically, the intelligent control layer is responsible for generating and issuing control instructions, supervising and managing the execution process of the service execution layer, receiving information fed back by the service execution layer and adjusting task instructions according to the fed back information, and receiving and executing the control instructions of the intelligent control layer and feeding back the execution state and results of the instruction tasks to the intelligent control layer.
Specifically, the service execution layer adopts a micro-service-based architecture, and encapsulates different network functions and service logic into independent micro-service modules. Specifically, the micro-service module comprises a scout task module, a relay task module and an attack task module, wherein the scout task module is responsible for image data acquisition and transmission, the relay task module is responsible for data relay, and the attack task module is responsible for executing attack operation on a target. The micro-service modules interact through a lightweight communication protocol (RESTful API), each micro-service module is reserved with a group of RESful interfaces, other modules can call the interfaces through HTTP requests to realize communication and writing among the modules, and the micro-service modules can quickly create and destroy task subnets according to service requirements. The task sub-network is composed of a plurality of micro-service modules responsible for different functions and used for completing specific tasks, and meanwhile, the micro-service modules support dynamic loading and unloading, so that required modules can be dynamically loaded or no longer required modules can be unloaded in running according to the change of task demands, and the utilization rate and flexibility of self-organizing network system resources are improved. The task sub-network is dynamically created and destroyed according to the service requirements.
More specifically, the micro service module is a logic unit and is composed of a plurality of sub-network nodes, and the sub-network nodes work cooperatively to realize specific network functions and service logic, wherein the sub-network nodes can be common nodes or cluster head nodes.
More specifically, the service execution layer adopts a multi-dimensional label matching cooperative mechanism of the subnet nodes, defines multi-dimensional labels for each subnet node in the self-organizing network, and realizes cross-subnet node resource sharing through the multi-dimensional label matching of the subnet nodes. In addition, independent space-time-frequency resource slices are allocated to each subnet node, so that the resource isolation and efficient utilization among the subnet nodes are ensured.
More specifically, a multidimensional label is defined for each subnet node, the multidimensional label including computing power, communication radius, energy status, task type. And finding the best matched subnet node according to the service requirement and the self-organizing network resource condition by using a label matching algorithm, and distributing independent space-time-frequency resource slices for each subnet node, thereby realizing the resource sharing and cooperative work of the cross-subnet nodes through label matching and improving the resource utilization rate and the task execution efficiency.
More specifically, the space-time-frequency resource slices include space resource slices, time slot resource slices, and spectrum resource slices. The space resource slice utilizes a space multiplexing technology to allocate independent space resources for each subnet node so as to improve the utilization rate of the space resources, the time slot resource slice is used for carrying out resource slicing in a time dimension so as to allocate a specific time slot for each subnet node and avoid resource conflict in time, and the frequency spectrum resource slice is used for allocating independent frequency spectrum resources for each subnet node according to service requirements and frequency spectrum availability.
More specifically, when resource slicing is performed in the time dimension, the system time of the self-organizing network is divided into a plurality of time slots, each time slot has a fixed duration, the time slots are basic time units of resource allocation, and the resources of the self-organizing network are allocated to different network subnet nodes in the time slots, so that the maximization of spectrum efficiency and service isolation are realized.
More specifically, the resource slice of each subnet node is dynamically adjusted according to the change of the service demand in the self-organizing network, so that the resource isolation among the subnet nodes is realized through independent space-time-frequency resource slices, and the resource interference and the conflict are effectively avoided.
It should be noted that the spatial multiplexing technology is a technology for transmitting multiple data streams through different spatial channels on the same frequency by utilizing the characteristics of multipath propagation, and is the prior art, and in addition, the spectrum resource slicing is a spectrum allocation technology commonly used in the field of wireless communication, and specific spectrum slicing processes are not repeated here, which can be understood by those skilled in the art.
It should be noted that, the tag matching algorithm is in the prior art, and the operation and the tag matching process thereof will not be described herein, which will be understood by those skilled in the art.
Furthermore, the layered three-dimensional network plane system also comprises a hardware decoupling and software defining system, and the hardware decoupling and software defining system comprises a reconfigurable hardware platform and a waveform library.
Specifically, the reconfigurable hardware platform provides an operation basis, and the waveform library dynamically loads a protocol stack to adapt to communication requirements.
Specifically, the reconfigurable hardware platform is a heterogeneous computing architecture based on an FPGA, an ARM and a radio frequency front end. Specifically, the FPGA provides high-efficiency parallel computing capability and flexible hardware programmability, can accelerate specific signal processing tasks, is a Xilinx UltraScale + series FPGA, an ARM processor is responsible for running control software and executing complex algorithms including an AI decision engine and a resource scheduling algorithm, a four-core ARM Cortex-A72 processor is adopted, the main frequency is 1.5GHz, 4GB DDR4 memory and 64GB eMMC memory are provided, the system can smoothly run a plurality of tasks, a radio frequency front end is integrated with a multimode radio frequency module, the multimode radio frequency module is responsible for transmitting and receiving signals, and the multimode radio frequency module supports 2 GHz-60 GHz full-frequency band coverage.
More specifically, the reconfigurable hardware platform provides hardware support for the operation of the waveform library, and the reconfigurable hardware platform supports the dynamic loading technology of the waveform library, and switches different communication protocol stacks through the cooperative work of the FPGA and the ARM processor.
Specifically, the waveform library contains 50 communication protocol stacks. And switching different communication protocol stacks through a dynamic loading technology. The communication protocol stack comprises an LTE-D2D, wi-Fi 6, TSN time sensitive network.
It should be noted that, the dynamic loading technique is a software engineering policy, which allows a program to load required functional modules or resources as needed at runtime, instead of loading all contents once at the time of starting the program, which is an existing software defined radio and communication technique, and it will be understood by those skilled in the art that the details are not repeated here.
Another aspect of the present invention provides a software defined radio based intelligent management and control method for an ad hoc network dynamic slice architecture, which is implemented according to the following steps:
Step S1, a basic communication layer, an intelligent control layer and a service execution layer are constructed;
And S2, determining the frequency spectrum ranges and the access modes of different planes according to the requirements of the different planes of the self-organizing network.
Specifically, in step S2, the rule for allocating plane-level resources is as follows:
and step S3, monitoring the node state and the link quality, and performing survivability enhancement processing after the node fails or the link terminal.
In the step, the self-organizing network continuously monitors the node state and the link quality, and when a certain node failure or link interruption is detected, a route repair algorithm is immediately triggered, and a link path is recalculated based on the route repair algorithm. After the new link path calculation is completed, the routing table of the affected node is updated, and the update information is broadcast to the relevant nodes. Meanwhile, the affected node caches data, and after a new path is established, the cached data is retransmitted to the target node.
It should be noted that, the route repair algorithm is in the prior art, and a detailed process of performing new path calculation by using the route repair algorithm is not described in detail, which can be understood by those skilled in the art.
When the network node fails or the link is interrupted, the route repair is triggered, and the route repair can finish path recalculation within 100 ms.
Further, in the present invention, the architecture of the basic communication layer, the intelligent control layer, and the service execution layer in step S1 is implemented according to the following steps:
step S1.1, initializing a node of a basic communication layer and constructing topology;
In this step, each node of the ad hoc network starts a bluetooth low energy scanning function, periodically broadcasts a broadcast frame including its own identifier, position, and energy state, and after receiving the broadcast frame, the neighboring node extracts and records information of the transmitting node. Meanwhile, each node of the self-organizing network sends an active detection frame to the neighbor node, and the neighbor node is requested to respond to acquire more detailed neighbor information and confirm the reachability of the neighbor node. Based on the collected neighbor node information, an initial topology map of the ad hoc network is constructed.
In the invention, the step S1.1 is realized according to the following steps:
s1.1.1, constructing a self-discovery and neighbor node list of a self-organizing network node;
Specifically, the nodes in the self-organizing network periodically broadcast a broadcast frame containing self-identification, position and energy state, and after receiving the broadcast frame, the adjacent nodes extract and record information of the transmitting nodes, wherein the information of the nodes comprises the identification, the position and the energy state, and a preliminary adjacent node list is constructed.
Step S1.1.2, initializing a cluster structure and electing cluster head nodes;
After the preliminary cluster structure is formed, each cluster node in the cluster performs voting according to the comprehensive performance index of the cluster head candidate cluster node, and a plurality of cluster nodes with the highest vote count commonly serve as cluster heads to form a multi-center architecture with a plurality of cluster head nodes in the cluster.
S1.2, dynamic architecture and optimization of an intelligent control layer;
step S1.2 is realized according to the following steps:
S1.2.1, initializing and scanning a command node;
Specifically, during the intelligent control layer construction process, the finger control node is activated. The finger control node performs full-network scanning by using a phased array antenna, collects self-organizing network topology information, has high gain and wide beam coverage capacity, and can rapidly acquire state information such as space position information, node residual energy and the like of all nodes in the self-organizing network.
Step S1.2.2, generating an initial control link according to the self-organizing network topology information of step S1.2.1;
Specifically, after the whole network topology information of the self-organizing network is obtained, an initial finger control node-area controller-common node control link is generated.
S1.2.3, identifying cluster head nodes of each cluster in the self-organizing network by the command node, electing a regional controller and configuring;
Specifically, after the initial control link is generated, the finger control node identifies the cluster head node of each cluster in the self-organizing network, and meanwhile, the cluster head node doubles as an area controller and is responsible for managing the communication and resource allocation of the nodes in the cluster. The command node configures the regional controller according to the scale of each cluster, the distribution of the nodes and the service requirement, and the configuration process comprises the steps of distributing control tasks, setting communication parameters and defining management strategies.
S1.2.4, collecting state data of nodes in the cluster by the regional controller and preprocessing;
Specifically, the regional controller collects state data of nodes in the cluster in real time, pre-processes the collected data, extracts key information and compresses the data, and reduces the data quantity transmitted to the command node. The state data of the nodes in the cluster comprise energy consumption, communication load and task execution progress of the nodes, and the preprocessing process comprises data cleaning and feature extraction.
Step S1.2.5, continuously monitoring the state of the self-organizing network;
Specifically, the intelligent control layer continuously monitors the state of the self-organizing network, and after the state of the self-organizing network changes, the command node adjusts the control strategy and generates a new control command. The regulation of the control strategy comprises the steps of reconfiguring the task of the regional controller, regulating the link bandwidth and optimizing the data transmission path, and the self-organizing network state change comprises the joining of nodes, the leaving of nodes, the change of link quality and the fluctuation of service demands.
Step S1.2.6, distributing the control instruction in step S1.2.5 to common nodes in the cluster and executing.
Specifically, the command node generates a new control instruction according to the adjusted control strategy, the control instruction is sent to the regional controller through the phased array antenna, the regional controller converts the control instruction into a specific configuration instruction and distributes the specific configuration instruction to the common nodes in the cluster, the common nodes execute corresponding operations after receiving the configuration instruction and feed back execution results to the regional controller, and the regional controller reports the feedback information to the command node after summarizing the feedback information to form closed-loop control.
And S1.3, dynamic task deployment and management of a service execution layer.
In the step, the command node receives an external command, analyzes the command to obtain task parameters, wherein the task parameters comprise task type, execution time, target area, time delay and bandwidth requirement, and adds the task into a scheduling queue according to task priority. And screening network nodes meeting the conditions according to the resource requirements of the tasks, and finally distributing resources such as communication frequency bands, time slots, bandwidths and the like to the network nodes according to the task types and the network node distribution planning subnet topological structure. After the configuration is completed, the network node enters an active state, and the subnet is put into use to start executing tasks.
In the invention, the step S1.3 is realized according to the following steps:
step S1.3.1, external instruction receiving and analyzing;
specifically, the command node receives an external task command, where the external task command is a command initiated by an entity outside the service execution layer of the ad hoc network, and is used to instruct the ad hoc network to execute a specific task or perform a specific operation. The command node analyzes the task instruction and extracts key task parameters. The task parameters include task type (scout, relay, attack), task execution time window, location information of target area, latency of task, bandwidth requirement.
Step S1.3.2, matching task scheduling with resources;
Specifically, tasks are added into a scheduling queue according to the priority and the emergency degree of the tasks, and then the demands of the tasks on network resources are analyzed according to task parameters, wherein the required network resources comprise spectrum resources, computing power, storage resources and communication bandwidth. And matching the subnet nodes meeting the task resource requirements according to the network resources required by the tasks.
Step S1.3.3, creating and configuring a task subnet;
Specifically, according to the task demand and the screened subnet node distribution in step S1.3.2, a task subnet is created, and an independent space-time-frequency resource slice is allocated to the task subnet. Meanwhile, the command node sends a configuration instruction to the task sub-network. And the subnet node performs corresponding configuration according to the configuration instruction and enters a task ready state. The topology structure of the task subnetwork can adopt star shape and net shape, and the configuration instruction comprises communication parameter setting, task execution program loading and resource allocation.
Step S1.3.4, executing tasks by the task sub-network;
Specifically, after configuration is completed, the task subnetwork begins executing tasks. Meanwhile, the command node monitors the task execution state in real time through the area controller, wherein the task execution state comprises the data transmission progress, the node energy consumption and the task completion condition.
And S1.3.5, completing the task and recycling the resources.
Specifically, when the task goal is reached or the task execution time is over, it is determined that the task is completed. After the task is completed, the command node issues a resource recovery command, and the subnet nodes in the task subnet release the allocated resources, including frequency spectrum, time slot, calculation resources and the like. After the resources are recovered, the subnet nodes return to an idle state and can be reused by other tasks.
Thus, various embodiments of the present invention have been described in detail. In order to avoid obscuring the concepts of the invention, some details known in the art have not been described. How to implement the solutions of the invention herein will be fully apparent to those skilled in the art from the above description, the scope of which is defined by the appended claims.
Claims (10)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202511059444.7A CN120825384A (en) | 2025-07-30 | 2025-07-30 | Dynamic slicing architecture and control method of self-organizing networks based on software-defined radio |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202511059444.7A CN120825384A (en) | 2025-07-30 | 2025-07-30 | Dynamic slicing architecture and control method of self-organizing networks based on software-defined radio |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN120825384A true CN120825384A (en) | 2025-10-21 |
Family
ID=97365609
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202511059444.7A Pending CN120825384A (en) | 2025-07-30 | 2025-07-30 | Dynamic slicing architecture and control method of self-organizing networks based on software-defined radio |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN120825384A (en) |
-
2025
- 2025-07-30 CN CN202511059444.7A patent/CN120825384A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR102235763B1 (en) | Multi-access edge computing based Heterogeneous Networks System | |
| US11838862B2 (en) | Coordinated target wake time (TWT) service for localized wireless neighborhoods | |
| Huang et al. | Software-defined wireless mesh networks: architecture and traffic orchestration | |
| US7742436B2 (en) | Distributed networking agent and method of making and using the same | |
| US11824726B2 (en) | Systems and methods for communication network customization | |
| JP7607557B2 (en) | Organizer and interconnection fabric mapper for virtual wireless base stations - Patents.com | |
| Das et al. | WLC30-4: static channel assignment in multi-radio multi-channel 802.11 wireless mesh networks: issues, metrics and algorithms | |
| US10433191B2 (en) | Channel management in a virtual access point (VAP) | |
| Dawaliby et al. | Network slicing optimization in large scale lora wide area networks | |
| Wang et al. | Autonomous traffic offloading in heterogeneous ultra-dense networks using machine learning | |
| CN110012475A (en) | A kind of the Slice framework and its construction method of ad-hoc self-organizing network | |
| CN104080093A (en) | Spectrum sensing and dynamic channel binding method, device and system | |
| CN117793657A (en) | Air-ground integrated network architecture and implementation method based on SDN and AI technology | |
| CN120166377A (en) | A method, system, device and medium for layered multi-subgroup self-organizing network of unmanned aerial vehicles | |
| CN111491301A (en) | Spectrum management device, electronic device, wireless communication method, and storage medium | |
| CN120825384A (en) | Dynamic slicing architecture and control method of self-organizing networks based on software-defined radio | |
| CN119865825A (en) | Multi-access data access method for farm monitoring and operation unmanned aerial vehicle | |
| Bartoli et al. | AI based network and radio resource management in 5G HetNets | |
| Zhang et al. | Analysis of mobile communication network architecture based on SDN | |
| US11617088B2 (en) | Real-time RF spectrum allocation and optimization in multi-function, co-located, interacting heterogeneous networks | |
| Kooshki et al. | Multi-Architecture COexistence Enabling Network Framework for 5G and Beyond Mobile Systems | |
| Wang et al. | Reinforcement learning based scheduling for heterogeneous UAV networking | |
| CN114666844A (en) | Intelligent, simple, efficient and fully-decoupled network architecture facing 6G | |
| CN112637066A (en) | Network slicing and path selection optimization method and system for power internet of things | |
| CN116545509B (en) | Unmanned aerial vehicle bee colony network communication system and control method thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |