Methods and devices for reliable transmission and resource optimization in ultra-dense network intelligent communication

Abstract

This invention discloses a method and apparatus for reliable transmission and resource optimization in intelligent communication for ultra-dense networking. The method includes: establishing an ultra-dense networking system architecture and topology model; establishing an ultra-dense networking communication model and designing orthogonal strategies and interference suppression mechanisms; constructing an optimization problem with minimizing information age as the objective function; decomposing the optimization problem into a constrained master problem and a pricing subproblem and solving them to obtain the resource allocation results of the constrained master problem and the negative cost path results of the pricing subproblem; establishing a redundancy backup resource optimization algorithm based on a column generation algorithm framework based on the resource allocation results of the constrained master problem and the negative cost path results of the pricing subproblem; executing the redundancy backup resource optimization algorithm based on the column generation algorithm framework, outputting the optimization results, and distributing master / backup path configurations and resource allocation strategies to each node. This invention can reduce the average information age of information flows in dense networking with low complexity and improve service transmission capabilities.

Description

Technical Field

[0001] This invention relates to the field of ultra-dense network resource optimization technology, and in particular to a method and apparatus for reliable transmission and resource optimization of intelligent communication in ultra-dense networks. Background Technology

[0002] With the development of 5G-A and 6G technologies, ultra-dense networking has become a core means to improve network capacity. 3GPP proposed the Integrated Access and Backhaul (IAB) network architecture, which achieves seamless coverage in high-density areas through the collaboration of IAB donor nodes and a large number of IAB access + backhaul fusion nodes. Based on in-band backhaul technology, it eliminates the need for additional dedicated fiber optic cables or licensed frequency bands. This meets the requirements of ultra-dense networking for low-cost, flexible deployment, and high bandwidth backhaul networks. In ultra-dense networking scenarios, the large number of user devices and complex service types place stringent demands on communication reliability and latency. However, high-frequency signals suffer from high penetration loss and weak diffraction capabilities, are easily affected by obstructions, and face significant beam interference and co-channel interference issues between dense nodes, making it difficult to guarantee service continuity and timeliness with a single link transmission.

[0003] Current methods for reliable transmission and resource optimization in high-frequency IAB ultra-dense networks generally consider routing and link scheduling separately, neglecting the strong coupling between them and failing to consider redundancy backup mechanisms. In current ultra-dense network system resource optimization, routing and link scheduling are decided separately, ignoring the strong coupling between "routing path selection affecting link load distribution and link resource availability constraining routing decision space," and the complexity of the solution increases exponentially with the number of nodes. This fails to achieve low-complexity resource optimization and reliable transmission, and fails to resolve the contradiction between high-reliability redundancy and efficient resource utilization. For example, Chinese invention patents with publication numbers CN 120711473 A and CN 116056149 A both achieve redundancy by adjusting communication paths to avoid interference, but they do not consider ensuring the orthogonality of backhaul links, making it easy for primary and backup links to share interference sources. Chinese invention patents with publication numbers CN119211943 A, CN 118804005 A, and CN 120034868 A employ static single-hop or double-hop backhaul networks and achieve overall system throughput improvement through proposed optimization methods. However, they do not consider interference suppression or routing and link scheduling optimization, and thus fail to improve the timeliness of system information.

[0004] In summary, ultra-dense networks exhibit dense node density and severe beam overlap. Existing solutions lack targeted beam constraint mechanisms and redundancy considerations. The primary and backup links do not achieve orthogonal separation of node or frequency band resources, meaning that both primary and backup links fail simultaneously when a shared node fails or the frequency band is interfered with. Furthermore, existing resource optimization methods do not dynamically allocate resources based on service timeliness requirements, resulting in high solution complexity in ultra-dense network scenarios and failing to meet real-time scheduling needs. Summary of the Invention

[0005] In view of the above-mentioned shortcomings in the prior art, the method and apparatus for reliable transmission and resource optimization of ultra-dense network intelligent communication provided by the present invention solves the problems of high algorithm complexity, low timeliness of service transmission, insufficient reliability, weak interference suppression capability and low resource utilization in the prior art.

[0006] To achieve the above-mentioned objectives, this invention provides a method for reliable transmission and resource optimization in ultra-dense network intelligent communication, comprising: Establish an ultra-dense network system architecture and topology model; An ultra-dense network communication model is established based on the ultra-dense network system architecture and topology model, and an orthogonal strategy and interference suppression mechanism are designed. Construct an optimization problem with the objective function of minimizing the information age; The optimization problem is decomposed into a restricted master problem and a pricing subproblem and solved to obtain the resource allocation result of the restricted master problem and the negative cost path result of the pricing subproblem. A redundant backup resource optimization algorithm based on the resource allocation results of the constrained master problem and the negative cost path results of the pricing subproblem is established. Execute the redundant backup resource optimization algorithm based on the column generation algorithm framework, output the optimization results, and distribute the primary and backup path configuration and resource allocation strategy to each node.

[0007] Secondly, the present invention also provides a device for reliable transmission and resource optimization of ultra-dense network intelligent communication, comprising: The network topology module is used to establish the architecture and topology model of ultra-dense network systems; The communication link module is used to establish an ultra-dense network communication model based on the ultra-dense network system architecture and topology model, and to design orthogonal strategies and interference suppression mechanisms. The constraint function module is used to construct optimization problems with the objective function of minimizing the information age; The decision-making module is used to decompose the optimization problem into a constrained master problem and a pricing subproblem and solve them to obtain the resource allocation results of the constrained master problem and the negative cost path results of the pricing subproblem; it is also used to establish a redundancy backup resource optimization algorithm based on the resource allocation results of the constrained master problem and the negative cost path results of the pricing subproblem. The resource optimization module is used to execute a redundant backup resource optimization algorithm based on a column generation algorithm framework, output the optimization results, and distribute the primary and backup path configuration and resource allocation strategy to each node.

[0008] The beneficial effects of this invention are as follows: 1. By designing two redundancy backup strategies—node orthogonality and link orthogonality—we ensure that there is no overlap in nodes or resources between the primary and backup links, avoiding cascading failures where "one fails and all fail," and improving the continuity and reliability of service transmission.

[0009] 2. By proposing an interference suppression mechanism based on beam angle constraints, an antenna radiation mode of "horizontal narrow beam + vertical wide beam" is designed to simultaneously constrain the beam angle of any two simultaneously activated links, thereby avoiding main lobe overlap interference, reducing beam interference in ultra-dense networks, and improving link quality.

[0010] 3. By constructing a mathematical model that integrates information age minimization, transmission reliability constraints, and resource efficiency optimization, the system balances service timeliness and resource consumption. Under the premise of meeting the same service requirements, it can reduce the overall network energy consumption or support more high-requirement service access without increasing the total amount of resources, thus significantly enhancing the system's economy and scalability.

[0011] 4. By proposing a low-complexity solution framework based on column generation algorithm, it is suitable for large-scale node scenarios in ultra-dense networking. The solution has low complexity, low cost, and strong scalability. Attached Figure Description

[0012] Figure 1 A flowchart of a method for reliable transmission and resource optimization in ultra-dense network intelligent communication; Figure 2 A schematic diagram of the ultra-dense networking system architecture and topology model provided for this embodiment; Figure 3 This is a schematic diagram of beam interference suppression in ultra-dense network configurations. Figure 4 A comparison chart of average information freshness performance curves for different schemes under the influence of redundant backup configuration of the number of business flows; Figure 5 A comparison chart of average service overhead performance curves for different schemes under the influence of redundant backup configuration of service flow number; Figure 6 A comparison chart of average link throughput performance curves for different schemes under the influence of redundant backup configuration of the number of service flows. Detailed Implementation

[0013] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0014] like Figure 1 As shown, in one embodiment of the present invention, a method for reliable transmission and resource optimization in ultra-dense network intelligent communication includes the following steps: S1. Establish the architecture and topology model of an ultra-dense network system through the network topology module.

[0015] like Figure 2 As shown, the ultra-dense networking system includes one IAB donor node as the core network access point, V IAB nodes as access and backhaul converged nodes, and U user equipment (UE). Figure 2 As shown (V=9, U=10), a transmission link topology of "core network-IAB donor-IAB node-UE" is formed.

[0016] Define the topology graph as follows , where V={0,...,V} represents the IAB donor node and the set of IAB nodes, and the label of the IAB donor node is 0; Represents the set of UEs; Represents access link and backhaul link A set of.

[0017] The IAB donor is responsible for global resource scheduling and topology management, establishing time-frequency resource connections between UEs and IAB nodes, and allocating resource blocks (RBs). IAB nodes are densely deployed within the coverage area, cascading and expanding via wireless links. UEs connect to IAB nodes through access links to obtain communication services. IAB nodes operate in half-duplex mode and share the same resource pool. Each UE generates an information stream and sends it to a nearby IAB node. The ultra-dense network system is configured as a Directed Acyclic Graph (DAG) topology, allowing IAB nodes to use multiple parent nodes, providing more path options.

[0018] To establish end-to-end redundant links, the IAB donor needs to collect basic network attributes, including connectivity between nodes and the association between nodes and wireless links. Link failure probability and interference probability are characterized by prior knowledge and the link-layer model. (Backhaul link) Defined as the set of links between IAB nodes and between the IAB donor and IAB nodes; access links Defined as the set of links between the UE and the IAB node. Access links and backhaul links between two nodes are considered as two different types of links. The relationship between nodes and links can be represented by a matrix. This indicates that its dimension is The link number is ( If node It is a link The starting end Then its value If node It is a link End of Then its value If node Not on the link If it is above, then its value Then the relationship matrix between nodes and links. It can be represented as:

[0019] The UE generates uplink (UL) or downlink (DL) traffic. The IAB donor needs to periodically collect traffic information from the UE to determine each service flow. Key parameters, including data transfer rate requirements End-to-end information age threshold And so on, as well as the source node of each stream. and destination node .use To describe the end-to-end information of flow scheduling, nodes When it is the source node of the stream 1, node When the destination node of the flow -1, otherwise If it is 0, then the end-to-end information of the flow scheduling is... It can be represented as:

[0020] After determining the network topology and the connections on each link, IAB donors send the results to each IAB node to establish end-to-end redundant connections for the information flow.

[0021] This step completes the quantitative modeling of the functions and topological relationships of the various entity modules (IAB donor, IAB node, user equipment) within the ultra-dense networking system, and derives the processing flow and parameters of the service flow within the network.

[0022] S2. Establish an ultra-dense network communication model based on the ultra-dense network system architecture and topology model through the communication link module, and design an orthogonal strategy and interference suppression mechanism.

[0023] The ultra-dense network communication model includes the channel propagation model, the link layer model, and the antenna and interference control model.

[0024] The specific method for constructing a channel propagation model is as follows: A channel propagation model is constructed based on 3GPP standards, modeling both line-of-sight and non-line-of-sight propagation scenarios separately. The appropriate standardized channel propagation model is selected based on the physical environment of the network deployment: a macrocell model is used in open areas, while a microcell model is used in dense areas. Open areas typically refer to geographically simple environments, such as deserts and hills, where terrain features have minimal impact on wireless propagation, user distribution is sparse, and service demand is relatively low. Network deployment in these areas may prioritize coverage and cost-effectiveness. For example, satellite communication systems in integrated air-space-ground networks can supplement terrestrial cellular networks, increasing revenue for terrestrial operators and are suitable for emergency communication for individual users in remote areas within the coverage range of gateway stations, as well as supplementary coverage in areas without terrestrial network coverage. Dense areas generally refer to urban areas with densely distributed buildings, complex roads, densely distributed and highly mobile users, and high service demand. Network deployment in these areas requires higher service levels and more complex equipment setups. For example, grid-based atmospheric monitoring stations need to be densely deployed to achieve accurate monitoring of grid-level pollution conditions at the street and community levels, providing precise guidance for regional air pollution control.

[0025] Different configurations are implemented for different link types in the network: the UMA channel model is configured for backhaul links, and the Indoor Office channel model is configured for access links. The signal attenuation characteristics under different scenarios are quantified using path loss formulas, providing a basis for link quality assessment.

[0026] View distance probabilities of different models for:

[0027]

[0028] in,

[0029]

[0030] In the above formula, This represents the line-of-sight probability for the Indoor Office channel model. For the line-of-sight probability of the UMA channel model, This indicates the distance between the transceiver and the receiver. For the height of the UE, The height of the IAB donor or IAB node. This is a correction factor for UE height. Distance from breakpoint For carrier frequency, It is the speed of light.

[0031] This yields the path losses for both models. for:

[0032]

[0033] In the formula, The path loss for the Indoor Office channel model. This represents the path loss in the UMA channel model.

[0034] For the link layer model, the service flow needs to meet the data transmission rate requirements. End-to-end information age threshold To ensure QoS, two link layer models are established to meet these requirements: a first link layer model and a second link layer model, to capture the probability of interruption based on data rate and information age threshold.

[0035] The specific method for constructing the first link layer model is as follows: The effective capacity model of the link is constructed, and its expression is:

[0036] In the formula, The QoS index is Time Link Effective capacity, For business flow end-to-end rate, It is the natural logarithm function. For the expected value calculation, is the base of the natural logarithm; The expression for setting the rate interruption probability constraint is as follows:

[0037] In the formula, Indicates the probability of rate interruption. Indicates the data transmission rate requirement. The threshold for the probability of rate interruption; Based on the link effective capacity model and rate interruption probability constraint, a first-link layer model is constructed, the expression of which is: ; The specific method for constructing the second link layer model is as follows: Construct an information age evolution model, the expression of which is:

[0038] In the formula, Indicates time slot Corresponding information age, Indicates time slot Corresponding information age; For time slots The packet arrival indicator is used when a new packet arrives. ,otherwise ; Construct a delay interruption probability model, the expression of which is:

[0039] In the formula, Indicates the probability of delayed interruption. Indicates link The time delay on, Indicates link The probability of backlog; The second link layer model is constructed based on the information age evolution model and the delay interruption probability model, and its expression is as follows:

[0040] In the formula, This is the information age threshold.

[0041] The ultra-dense network system has a directed acyclic graph topology, allowing IAB nodes to be configured with multiple parent nodes. Therefore, beam overlap may occur. By modeling the antennas and setting constraints, overlapping links are prevented from being activated simultaneously. In addition, the antenna pattern is obtained, the actual antenna gain is obtained, and a link interference model is built to provide global network interference information for IAB donors running this solution. Figure 3 This is a schematic diagram of beam interference suppression in ultra-dense network configurations.

[0042] For the antenna and interference control model, the IAB donor and IAB nodes consist of three sectored antennas with sufficient spatial isolation to limit self-interference. The transmit and receive beams are perfectly aligned between different nodes. Furthermore, the antenna radiation pattern is explicitly simulated to assess interference, as the beams of the interfering transmitter and the intended receiver may overlap. For each link in the scheduling mode, all other links are considered interfering links. According to the antenna model, the antenna radiation pattern is represented as a superposition of elemental radiation patterns. Therefore, the values ​​of the half-power beamwidth and antenna gain depend on the number of antenna elements.

[0043] The specific method for constructing the antenna and interference control model is as follows: The model of a single antenna element is constructed, and its expression is as follows:

[0044] In the formula, This represents the radiation pattern of a single antenna element. This represents the gain of a single antenna element. The horizontal angle between the signal transmission direction and the reception direction. The vertical angle between the signal transmission direction and the reception direction. Indicates horizontal gain attenuation. Indicates vertical gain attenuation. Indicates the maximum front-to-back ratio of a single antenna element; The array model of the IAB nodes is constructed as follows:

[0045] In the formula, This represents the radiation pattern of the array antenna elements. This represents the correlation coefficient between antennas. This indicates the number of horizontal elements in the antenna array. Indicates the number of vertical elements in the antenna array; The antenna gain in the main transmission direction and the antenna gain in the interference direction are calculated based on the array model using IAB nodes; the antenna gain in the main transmission direction is... :

[0046] The gain in the direction of interference is :

[0047] In the formula, Represents a node With nodes The horizontal angle between them Represents a node With nodes The vertical angle between them; The overall link interference model is constructed based on the main transmission direction gain and the interference direction gain, and its expression is:

[0048] In the formula, Represents a node Total interference within the interference range express At every moment The set of nodes within the interference range, For nodes The transmission power, For nodes With nodes Channel gain between Represents a node With nodes Path loss between Represents a node With nodes The distance between them; The link is calculated based on the total link interference model. The signal-to-noise ratio at a given point is expressed as:

[0049] in, Indicates link Signal-to-noise ratio at the location Represents a node The transmission power, For nodes With nodes Channel gain between node With nodes Path loss between For nodes With nodes The distance between them This indicates the link scheduling strategy at this time; This indicates the thermal noise power of the system; The beam main lobe overlap constraint condition is constructed, and its expression is as follows:

[0050] In the formula, , , All are node indexes, representing IAB donor nodes or IAB nodes; For link With Link The angle between them This represents the threshold value for the angle between links.

[0051] Meanwhile, the IAB donor determines end-to-end redundancy routing policies based on policy generation. Two main redundancy routing policies are considered: link orthogonal policy and node orthogonal policy. The link orthogonal policy allows access link sets and backhaul link sets for the same service flow to share intermediate nodes, while the node orthogonal policy does not allow access link sets and backhaul link sets for the same service flow to share any nodes. Of the two policies, the link orthogonal policy can be considered a special form of the node orthogonal policy (sharing nodes but isolating resources), and the system can dynamically select the appropriate policy based on service reliability requirements.

[0052] In this embodiment, the scheduling mechanism based on the aforementioned redundant routing strategy specifically includes a node redundancy mechanism and a link redundancy mechanism. Node redundancy mechanism: Multiple IAB nodes are densely deployed within the coverage area, employing redundant deployment to form spatially overlapping coverage. When an IAB node fails or channel quality deteriorates, it can quickly switch to a redundant node. Link redundancy mechanism: Link redundancy includes access link redundancy and backhaul link redundancy. Access link redundancy is achieved by configuring multiple IAB access nodes for the UE, allowing the UE to establish connections with multiple IAB nodes simultaneously and dynamically select the transmission link based on link quality. Backhaul link redundancy is achieved by establishing multiple backhaul paths. IAB nodes can establish backhaul links through different IAB nodes; when a backhaul link is interrupted due to interference or failure, it automatically switches to a backup backhaul link.

[0053] The technical effect that this step can achieve is to complete the construction of the ultra-dense network communication model and design orthogonal strategies and interference suppression mechanisms. Specifically, it involves establishing a link-layer communication model to provide a theoretical basis for link evaluation, and designing orthogonal strategies and interference suppression mechanisms to provide a theoretical basis for redundancy scheduling constraints.

[0054] S3. Construct an optimization problem with the objective function of minimizing information age using the constraint function module.

[0055] This includes the objective function and constraints; The constraints include the main constraints on link scheduling and the orthogonality constraints on redundant links; To meet the safety redundancy requirements of train services, each service is configured with a main link. and backup links Furthermore, the two links must be orthogonal (without overlapping links) to avoid simultaneous link interruptions. Mathematically, this can be expressed as: .

[0056] The specific orthogonality constraint of redundant links is as follows:

[0057] In the formula, As the primary link scheduling decision variable, Indicates business flow At any moment via link transmission, Represents the backup link scheduling decision variable. Indicates business flow At any moment via link transmission; The main constraints of link scheduling include: Half-duplex constraint: The half-duplex characteristic of IAB nodes stems from the hardware limitations of their radio frequency (RF) modules: at any given time, the node's RF unit can only operate in either "receive" or "transmit" mode, and cannot perform bidirectional transmission simultaneously. In ultra-dense network scenarios, this constraint manifests specifically as follows: when an IAB child node receives UE data through the access link, its backhaul link must be in a dormant state and cannot send data to other IAB nodes or IAB donors; when an IAB child node uploads data to its parent node through the backhaul link, the access link must suspend service and cannot receive UE access requests. Violating this constraint can lead to internal signal interference (self-interference) within the node, which can significantly reduce link throughput in severe cases.

[0058] Its expression is:

[0059] In the formula, Represents a set of business flows. Represents nodes Other connected nodes, For scheduling decision variables, Indicates business flow At any moment via link transmission; For scheduling decision variables, Indicates business flow At any moment via link Transmission; V represents the IAB donor and the set of IAB nodes; Link resource exclusivity constraints: In ultra-dense networking systems, link resource exclusivity constraints are particularly important, meaning that within any time slot t, a physical link (including access links and backhaul links) can only support one service flow. Providing services must not allow multiple streams to share the same link, otherwise it will lead to data conflicts, a surge in packet loss rate, and seriously affect the reliability of critical business operations.

[0060] Its expression is:

[0061] In the formula, This represents the set of access links and backhaul links in the architecture and topology model of an ultra-dense network system.

[0062] Multi-beam transceiver constraints: The multi-beam transceiver capability of IAB nodes (especially parent nodes acting as data aggregation points, such as ground IAB donors or upper-layer IAB nodes) has a physical upper limit. Any IAB node... In any time slot Within the beam, the total number of simultaneously active receiving or transmitting links must not exceed the beam concurrency limit. This constraint is particularly critical for the parent node, as it needs to connect to multiple child nodes. If the concurrent link limit is exceeded, it will lead to beam main lobe overlap and a surge in interference, seriously affecting the reliability of service transmission.

[0063] Its expression is:

[0064] In the formula, This is the upper limit for beam concurrency; Link scheduling order constraint: Ensures that links in the path are activated in the correct order (e.g., child node links are activated first, parent node links are activated later), avoiding data transmission disorder. Its expression is:

[0065] In the formula, For scheduling decision variables, Represents information flow At any moment via link transmission; Link capacity constraints ensure that the total transmission capacity of active links meets the flow's requirements; their expression is:

[0066] In the formula, Indicates the main link. The length of the subframe. For channel capacity, This represents the probability of link interruption.

[0067] The objective function is as follows:

[0068] In the formula, The age of the business flow information received by the recipient. The information age of the backup link.

[0069] The technical effect achieved by this step is to model the average information age of the business flow and the business requirements, redundancy backup and resource constraints within the system, thereby deriving the problem of minimizing information age.

[0070] S4. The optimization problem is decomposed into a constrained master problem and a pricing subproblem by the decision-making module and then solved to obtain the resource allocation result of the constrained master problem and the negative cost path result of the pricing subproblem. The objective of the primary problem of selecting the optimal redundancy configuration is to choose a primary-backup link combination from all feasible configurations, minimizing the freshness of train safety information while satisfying constraints such as interference and resource exclusivity. The objective function is to minimize the maximum value of the freshness of the primary-backup link, given a set... The main problem can be represented as:

[0071] At this point, the objective function is a non-linear function that includes a max function, therefore auxiliary variables are used. Let represent the maximum information age in the primary and backup paths. To ensure that the information age meets the requirements, the objective function is transformed into a linear function, expressed as:

[0072] Accordingly, information age constraints are added for any time slot. Heliu :

[0073] Furthermore, the end-to-end constraint for the flow is added as follows:

[0074] In the formula, This represents the maximum age of the information in the primary and backup paths. Represents the link-node topology matrix of the system. An end-to-end information matrix representing the information flow.

[0075] In this embodiment, the initial connection selection algorithm is used to map the connection selection problem to the shortest path selection problem, and then Dijkstra's algorithm is used to solve this constrained problem. This algorithm is used because it is a widely used routing algorithm. Note that such algorithms can be replaced by other algorithms. Dijkstra's algorithm consists of three steps. First, each link is obtained according to the aforementioned steps. First, determine the parameters. Second, randomly select a subset of the flows. Third, execute Dijkstra's algorithm to find the parameters for each flow. The two orthogonal shortest paths between the source and destination can be mapped to the flow. Two candidate initial connections. Since the goal of the restricted master problem is to minimize the information age of the traffic, the link metric is designed to be based on the traffic. The required scheduling time is the sum of the routing metrics and the link metrics. To obtain the link metrics, it is assumed that each connection is assigned to a flow with equal probability, and the sending power of nodes is evenly distributed across candidate connections. To ensure that candidate connections meet the constraints, each path is further checked for compliance. If a path does not meet all constraints, it is removed and replaced with a new shortest path. Otherwise, it is added to the set. In the middle. Due to Each flow in the set selects two end-to-end connections that satisfy the constraints ("link scheduling main constraint and redundant link orthogonality constraint" and "flow end-to-end constraint"), thus the initial connection selection algorithm guarantees the set The initial connections in the network satisfy the above constraints.

[0076] The pricing subproblem, based on the dual variables of the main problem, generates new redundant link configurations to reduce the objective value of the main problem. The core is calculating the reduced cost (RC) and finding new configurations with negative RC. Based on the strong duality theorem of linear programming, the following steps are taken: , are the dual variables of the main constraints on link scheduling and the orthogonality constraints on redundant links, respectively, and they represent the following meanings: Indicates link In the time slot marginal cost of resources or penalty coefficient for occupation, Represents a node During the period Penalty coefficient for violating half-duplex state, Represents a node During the period Penalty factor for beam overuse Indicates time period Penalty coefficient for information exceeding age limits Indicates a violation of the link or flow Penalty coefficient for effective capacity constraints This represents the protection coefficient for the business connection.

[0077] The formula for calculating cost reduction is: in, The inherent overhead of quantifying the network resources occupied by a path consists of link transmission costs and is defined as the reciprocal of the link capacity. ; These are all decision variables, representing the correlation variables between the main constraints of link scheduling and the orthogonality constraints of redundant links corresponding to path P. Node nIn the time slot t When activated by path P, , , ; For path P in time slot t Information about age increment variables; For link l For business flow f The effective capacity penalty cost for path P to provide services.

[0078] The pricing subproblem, which is the result of decomposing the optimization problem, is as follows:

[0079] In the formula, This indicates a reduction in the cost of solving the link. To solve for the route path length, This indicates the maximum path length. The effective channel capacity represents the information flow. This indicates the probability of link interruption.

[0080] The technical effect of this step is to decompose the optimization problem into a constrained main problem and a pricing sub-problem. In the constrained problem, the joint optimization results of routing and link scheduling under the initial route candidate set are derived. In the pricing sub-problem, a new path that satisfies redundancy backup and resource constraints is searched.

[0081] S5. Based on the resource allocation results of the constrained master problem and the negative cost path results of the pricing subproblem, a redundant backup resource optimization algorithm based on the column generation algorithm framework is established through the decision-making module.

[0082] The pricing subproblem is used to find new candidate connections that reduce the information age of the objective function. To obtain these connections, the solutions to the dual problem of the primal problem are used to determine the metrics of the links and nodes.

[0083] More specifically, the subproblems employ a "hierarchical search + cost adjustment" solution strategy, with the base cost as the starting point. To measure the initial path set, the constrained K-short-circuit algorithm is used to generate the top 10 shortest paths as candidate primary paths, ensuring the resource efficiency of the initial solution. For each candidate primary path, the current dual variable is substituted to calculate the corrected cost, ensuring consistency with the optimization direction of the main problem. For each surviving primary path, the constrained K-short-circuit algorithm is run again to generate backup paths, and the path with the largest reduction in RC is selected as the optimal backup path. For each link, the RC value is calculated. If RC < 0, it indicates that the configuration can improve the solution to the main problem. The same procedure is executed on each node. For each traffic, the constrained K-short-circuit algorithm is used to select the path with the lowest new routing metric, where the routing metric is the sum of the node metric and the link metric. When the connection exploration algorithm cannot find a new connection, the joint optimization algorithm of routing planning and link scheduling terminates. After iterative convergence, the solution to the main problem is the optimal redundant network configuration, which includes: primary and backup link routing planning that satisfies orthogonal constraints in redundant backup transmission; time slot scheduling strategy that satisfies service requirements and resource constraints such as half-duplex for each link; and minimizing the average information age of the service flow.

[0084] S6. Execute the redundant backup resource optimization algorithm based on the column generation algorithm framework through the resource optimization module, output the optimization results, and distribute the primary and backup path configuration and resource allocation strategy to each node.

[0085] The IAB donor, based on the proposed algorithm, outputs optimization results and distributes primary / backup path configurations and resource allocation strategies to each node. It periodically updates the network status and adjusts the optimization scheme accordingly. IAB nodes execute data transmission and reception according to instructions, monitor path quality in real time, and ensure continuous service transmission. The network status is periodically updated, and the algorithm is rerun to dynamically adjust the optimization scheme. The technical effect of this step is that the IAB donor executes a redundancy backup resource optimization algorithm based on a column generation algorithm framework, outputs optimization results based on the proposed algorithm, distributes primary / backup path configurations and resource allocation strategies to each node, and dynamically adjusts the optimization scheme.

[0086] To verify that the method proposed in this invention can reduce the average information age of information flow under dense networking with low complexity and improve service transmission capability, the effect of this invention is illustrated using the parameter configuration in Table 1 as an example.

[0087] Table 1

[0088] Under the simulation parameter settings shown in Table 1, two comparison schemes—one based on optimal latency and the other based on a minimum weight directed tree—were set up to compare the performance of the proposed method with those in this specification. Simulations were performed to obtain the average service overhead performance curve, average link throughput performance curve, and average link throughput performance curve of the proposed method under the influence of the number of service flows and different redundancy backup configurations, as shown in the figures below. Figure 4 , Figure 5 and Figure 6 As shown, the proposed method can reduce the average information age of information flow under dense networking with low complexity and improve service transmission capabilities.

[0089] Figure 4 The graph shows the average information age performance under the influence of redundant backup configuration for the number of service flows. As can be seen from the graph, the proposed solution achieves a lower information age compared to the latency-optimal solution and the minimum weight directed tree solution. With the increase in the number of service flows, the information age increases due to system resource constraints. The node orthogonal mechanism improves information age performance compared to the link orthogonal mechanism, thanks to the synchronous resource allocation mechanism of all links in the IAB node. This mechanism minimizes transmission delays caused by half-duplex conflicts, thus ensuring the timeliness of information.

[0090] Figure 5 The graph shows the average service overhead performance of the proposed scheme under the influence of redundant backup configuration for the number of service flows. It can be observed that the proposed scheme has lower service overhead than the latency-optimal scheme and the minimum-weighted directed tree scheme, demonstrating the advantages of the proposed scheduling mechanism. Furthermore, as the number of flows increases, the average service overhead of the proposed scheme, the latency-optimal scheme, and the minimum-weighted directed tree scheme also increases accordingly. The performance difference between the link orthogonal strategy and the node orthogonal strategy stems from their different coordination mechanisms: link orthogonality only achieves resource isolation at the link level, while node orthogonality requires additional synchronization of node-level constraints, which introduces additional overhead.

[0091] Figure 6 The graph shows the average link throughput performance of the proposed solution under the influence of redundant backup configuration for the number of service flows. Compared to the latency-optimal solution and the minimum-weight directed tree solution, the proposed solution achieves a higher average link throughput. The average link throughput of the proposed solution increases with the number of service flows. The node orthogonal mechanism improves throughput compared to the link orthogonal mechanism due to its node-level coordination mechanism: this mechanism avoids fragmented resource usage between node transmit and receive links, thereby achieving more efficient spatial spectrum reuse.

[0092] In summary, the intelligent communication reliable transmission and resource optimization method for ultra-dense networking described in this invention establishes an optimization problem by constructing an ultra-dense networking system model and designing an orthogonal redundancy strategy. This method can reduce the average information age of information flow under dense networking with low complexity and improve service transmission capabilities.

[0093] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent modifications or variations made by those skilled in the art based on the disclosure of the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope set forth in the claims.

Claims

1. A method for reliable transmission and resource optimization in ultra-dense network intelligent communication, characterized in that, include: Establish an ultra-dense network system architecture and topology model; An ultra-dense network communication model is established based on the ultra-dense network system architecture and topology model, and an orthogonal strategy and interference suppression mechanism are designed. Construct an optimization problem with the objective function of minimizing the information age; The optimization problem is decomposed into a restricted master problem and a pricing subproblem and solved to obtain the resource allocation result of the restricted master problem and the negative cost path result of the pricing subproblem. A redundant backup resource optimization algorithm based on the resource allocation results of the constrained master problem and the negative cost path results of the pricing subproblem is established. Execute the redundant backup resource optimization algorithm based on the column generation algorithm framework, output the optimization results, and distribute the primary and backup path configuration and resource allocation strategy to each node; The orthogonal strategy includes node orthogonal strategy and link orthogonal strategy; the link orthogonal strategy allows the access link set and backhaul link set of the primary and backup paths of the same information flow to share intermediate nodes; the node orthogonal strategy does not allow the access link set and backhaul link set of the primary and backup paths of the same information flow to share nodes other than the source node and the destination node. The optimization problem with minimizing information age as the objective function includes the objective function and constraints. The constraints include the main constraints on link scheduling and the orthogonality constraints on redundant links; The specific orthogonality constraint of redundant links is as follows: In the formula, As the primary link scheduling decision variable, Indicates business flow At any moment via link transmission, Represents the backup link scheduling decision variable. Indicates business flow At any moment via link transmission; The main constraints of link scheduling include: The expression for a half-duplex constraint is: In the formula, Represents a set of business flows. Represents nodes Other connected nodes, For scheduling decision variables, Indicates business flow At any moment via link transmission; For scheduling decision variables, Indicates business flow At any moment via link Transmission; V represents the IAB donor node and the set of IAB nodes; The expression for the link resource exclusivity constraint is: In the formula, This represents the set of access links and backhaul links in the architecture and topology model of an ultra-dense network system. The multi-beam transceiver constraint is expressed as follows: In the formula, This is the upper limit for beam concurrency; The link scheduling order constraint is expressed as follows: In the formula, For scheduling decision variables, Indicates business flow At any moment via link transmission; Link capacity constraints are expressed as follows: In the formula, Indicates the main link. The length of the subframe. For channel capacity, This represents the probability of link interruption. The objective function is as follows: In the formula, The age of the business flow information received by the recipient. The information age of the backup link.

2. The method according to claim 1, characterized in that, The ultra-dense network communication model includes a channel propagation model, a link layer model, and an antenna and interference control model; The specific method for constructing a channel propagation model is as follows: A channel propagation model is constructed based on the 3GPP standard, and the channel propagation model is used to model line-of-sight propagation scenarios and non-line-of-sight propagation scenarios respectively. Choose the corresponding standardized channel propagation model based on the physical environment of the network deployment: use the macro cell model as the channel propagation model in open areas and the micro cell model as the channel propagation model in dense areas. Differentiated configurations are made for different link types in the network: the UMA channel model is configured for backhaul links, and the Indoor Office channel model is configured for access links.

3. The method according to claim 2, characterized in that, The link layer model includes the first link layer model and the second link layer model; The specific method for constructing the first link layer model is as follows: The effective capacity model of the link is constructed, and its expression is: In the formula, The QoS index is Time Link Effective capacity, For business flow end-to-end rate, It is the natural logarithm function. For the expected value calculation, is the base of the natural logarithm; The expression for setting the rate interruption probability constraint is as follows: In the formula, Indicates the probability of interruption. Indicates the data transmission rate requirement. The threshold for the probability of rate interruption; Based on the link effective capacity model and rate interruption probability constraint, a first-link layer model is constructed, the expression of which is: ； The specific method for constructing the second link layer model is as follows: Construct an information age evolution model, the expression of which is: In the formula, Indicates time slot Corresponding information age, Indicates time slot Corresponding information age; For time slots The packet arrival indicator is used when a new packet arrives. ,otherwise ; Construct a delay interruption probability model, the expression of which is: In the formula, Indicates link The time delay on, Indicates link The probability of backlog; The second link layer model is constructed based on the information age evolution model and the delay interruption probability model, and its expression is as follows: In the formula, This is the information age threshold.

4. The method according to claim 3, characterized in that, The specific method for constructing the antenna and interference control model is as follows: The model of a single antenna element is constructed, and its expression is as follows: In the formula, This represents the radiation pattern of a single antenna element. This represents the gain of a single antenna element. The horizontal angle between the signal transmission direction and the reception direction. The vertical angle between the signal transmission direction and the reception direction. Indicates horizontal gain attenuation. Indicates vertical gain attenuation. Indicates the maximum front-to-back ratio of a single antenna element; The array model of the IAB nodes is constructed as follows: In the formula, This represents the radiation pattern of the array antenna elements. This represents the correlation coefficient between antennas. This indicates the number of horizontal elements in the antenna array. Indicates the number of vertical elements in the antenna array; The antenna gain in the main transmission direction and the antenna gain in the interference direction are calculated based on the array model using IAB nodes; the antenna gain in the main transmission direction is... : The gain in the direction of interference is : In the formula, Represents a node With nodes The horizontal angle between them Represents a node With nodes The vertical angle between them; The overall link interference model is constructed based on the main transmission direction gain and the interference direction gain, and its expression is: In the formula, Represents a node Total interference within the interference range express At every moment The set of nodes within the interference range, For nodes The transmission power, For nodes With nodes Channel gain between Represents a node With nodes Path loss between Represents a node With nodes The distance between them; The link is calculated based on the total link interference model. The signal-to-noise ratio at a given point is expressed as: in, Indicates link Signal-to-noise ratio at the location Represents a node The transmission power, For nodes With nodes Channel gain between node With nodes Path loss between For nodes With nodes The distance between them This indicates the link scheduling strategy at this time; This indicates the thermal noise power of the system; The beam main lobe overlap constraint condition is constructed, and its expression is as follows: In the formula, , , All are node indexes, representing IAB donor nodes or IAB nodes; For link With Link The angle between them The threshold value represents the angle between links; V represents the IAB donor node and the set of IAB nodes.

5. The method according to claim 1, characterized in that, The restricted master problem after decomposing the optimization problem is specifically: In the formula, This represents the maximum age of the information in the primary and backup paths. Represents the link-node topology matrix of the system. An end-to-end information matrix representing the information flow; The pricing subproblem, which is decomposed from the optimization problem, is as follows: In the formula, This indicates a reduction in the cost of solving the link. To solve for the route path length, This indicates the maximum path length. The effective channel capacity represents the information flow. This indicates the probability of link interruption.

6. The method according to claim 5, characterized in that, The redundancy backup resource optimization algorithm based on the column generation algorithm framework is as follows: Based on the solution results of the restricted master problem, the dual variables of each constraint in the optimization problem with the objective function of minimizing information age are obtained; The obtained dual variables are substituted into each candidate path of the negative cost path result of the pricing subproblem to calculate the corrected cost, and the consistency of the path with the optimization direction of the main problem is evaluated based on the corrected cost. For each surviving main path that passes the evaluation, the algorithm for solving the pricing subproblem is run again to generate a set of candidate backup paths. The path that reduces costs the most in the candidate backup path set is selected as the optimal backup path. For each surviving primary path and its corresponding optimal backup path, calculate the cost reduction value. If the cost reduction is less than 0, then the surviving primary path and its corresponding optimal backup path are used as the solution to update the restricted master problem. When there is no combination of surviving primary path and optimal backup path with a cost reduction of less than 0, the redundant backup resource optimization algorithm ends. Among them, the solution to the main problem after the redundancy backup resource optimization algorithm is the optimal redundancy network configuration, including: primary and backup link routing planning that satisfies orthogonality constraints in redundant backup transmission, time slot scheduling strategy that satisfies constraints for each link, and minimizing the average information age of the service flow.

7. An apparatus based on the ultra-dense network intelligent communication reliable transmission and resource optimization method according to any one of claims 1 to 6, characterized in that, include: The network topology module is used to establish the architecture and topology model of ultra-dense network systems; The communication link module is used to establish an ultra-dense network communication model based on the ultra-dense network system architecture and topology model, and to design orthogonal strategies and interference suppression mechanisms. The constraint function module is used to construct optimization problems with the objective function of minimizing the information age; The decision-making module is used to decompose the optimization problem into a constrained master problem and a pricing subproblem and solve them to obtain the resource allocation results of the constrained master problem and the negative cost path results of the pricing subproblem; it is also used to establish a redundancy backup resource optimization algorithm based on the resource allocation results of the constrained master problem and the negative cost path results of the pricing subproblem. The resource optimization module is used to execute a redundant backup resource optimization algorithm based on a column generation algorithm framework, output the optimization results, and distribute the primary and backup path configuration and resource allocation strategy to each node.

Images