A method for interference coordination communication for high-density scenarios

Abstract

The application relates to the technical field of communication, in particular to a high-density scene-oriented interference coordination communication method, aiming to solve the problems of low spectrum efficiency and excessive signaling overhead caused by multi-dimensional composite interference in an ultra-dense network. The method comprises the following steps: constructing a three-dimensional interference feature matrix in the time domain, the frequency domain and the space domain to accurately depict the interference; only broadcasting high-interference resource identifiers through a sparse interaction protocol to reduce the signaling load; jointly optimizing time-frequency resource blocks and beamforming directions to dynamically avoid interference areas; adaptively adjusting the beam null position based on the compression sensing result to suppress the interference source; and introducing a closed-loop verification mechanism to trigger strategy updating according to the actual throughput and the signal-to-interference-and-noise ratio. Through the above technical scheme, high-precision interference coordination, low-overhead scalable coordination and millisecond-level dynamic adaptation are realized, the ultra-reliable and low-latency service is guaranteed, and the system spectrum efficiency and the edge user performance are significantly improved.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and more specifically to an interference coordination communication method for high-density scenarios. Background Technology

[0002] As fifth-generation and future mobile communication systems evolve towards high-density deployment, architectures such as ultra-dense networks (UDNs) and massive MIMO (Multiple-Input Multiple-Output) have become key means to improve network capacity and coverage performance. In such high-density scenarios, the dense deployment of numerous access nodes leads to a significant increase in inter-cell interference, severely restricting spectrum efficiency and user service quality. Interference coordination, as a core technology for mitigating such problems, aims to suppress harmful interference from neighboring cells while ensuring the transmission of useful signals through resource scheduling and signal processing strategies.

[0003] Interference coordination methods for high-density wireless environments require dynamic adaptation and joint optimization across multiple dimensions, including the time, frequency, and spatial domains. However, existing interference coordination mechanisms still face multiple challenges in dealing with complex interference environments characterized by high density and high dynamics. Existing technologies typically rely on static or semi-static coordination strategies, making it difficult to respond in real-time to rapidly changing user distribution and channel states. While some solutions introduce channel state information for precoding or power control, these become impractical in high-density scenarios due to the dramatic increase in feedback overhead. Other methods only implement coordination in a single dimension (such as the time domain), lacking the ability to jointly perceive and process multi-dimensional interference sources, resulting in low resource utilization efficiency. Furthermore, traditional coordination mechanisms are mostly based on tightly coupled base station cooperation, which can easily trigger signaling storms in dense networks and is difficult to reconcile with the differentiated needs of various access nodes, such as macro, micro, and small stations, in heterogeneous networks.

[0004] Therefore, there is an urgent need for an interference coordination communication method for high-density scenarios, which can achieve efficient, low-overhead and highly adaptable interference management through multi-dimensional resource collaboration, lightweight interaction mechanisms and intelligent interference identification, so as to support the performance evolution requirements of future high-density wireless communication systems. Summary of the Invention

[0005] The purpose of this invention is to provide an interference coordination communication method for high-density scenarios, which can effectively solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: An interference coordination communication method for high-density scenarios includes the following specific steps: At each access node, a three-dimensional interference feature matrix is ​​constructed based on historical scheduling logs, neighboring cell resource occupancy status, and user channel quality feedback. This matrix includes time, frequency, and spatial dimensions. The three-dimensional interference feature matrix is ​​used to characterize the intensity and source of interference experienced by each resource unit in the current cell. Each access node broadcasts only the high-interference resource identifiers and their corresponding interference levels in its interference feature matrix that exceed a preset threshold to its neighboring nodes through a preset sparse interaction protocol, thus avoiding full information exchange. Based on the received high-interference resource identifiers of neighboring cells and its own interference feature matrix, a multi-objective optimization algorithm is used to jointly allocate time-frequency resource blocks and beamforming directions, so that the scheduling of users in this cell avoids high-interference time-space-frequency regions, while minimizing the reverse interference caused to neighboring cells. Based on the current user distribution and the location of the interference source, the beam main lobe pointing and null position of the large-scale multiple input multiple output antenna array are dynamically adjusted to concentrate the useful signal energy in the direction of the target user and form a deep suppression null in the direction of the interference source. At the end of each scheduling cycle, the actual received signal-to-interference-plus-noise ratio and throughput data are collected to evaluate the effectiveness of the current interference coordination strategy. If the performance indicators are lower than the preset threshold, the system will return to rebuild the interference perception model and update the coordination strategy.

[0007] In some embodiments, the process of constructing the three-dimensional interference feature matrix includes: dividing the system bandwidth into several subcarrier groups as frequency domain units using scheduling subframes as time units, and combining the spatial beam index of the large-scale multiple-input multiple-output system as spatial domain units to form a three-dimensional grid structure; for each grid unit, the ratio of the average received power from neighboring users to the target channel gain of the users in the current cell is calculated as the interference intensity index of that unit.

[0008] In some embodiments, the sparse interaction protocol specifies that: only when the interference intensity index of a resource unit exceeds a preset threshold, its identifier and interference level are encoded into information of a predetermined bit length and broadcast; the broadcast period is a preset interval of multiple scheduling subframes, and it is only sent to adjacent access nodes whose geographical distance is less than a predetermined distance.

[0009] In some embodiments, the objective function of the multi-objective optimization algorithm is a weighted sum of maximizing the total system throughput and minimizing the cross-cell interference power. The weight coefficients are dynamically adjusted according to the current network load. The total system throughput is calculated by Shannon's formula, and the cross-cell interference power is obtained by weighted summation of the occupancy of high-interference resource identifiers in neighboring cells.

[0010] In some embodiments, the beamforming uses sparse channel estimation results based on compressed sensing to determine the spatial angle of the interference source, and uses the minimum mean square error criterion to calculate the optimal precoding vector, so that while the gain in the direction of the target user is not lower than a preset gain threshold, the radiated power attenuation in the direction of the interference source is greater than or equal to a preset attenuation threshold.

[0011] In some embodiments, the preset threshold for performance verification includes: the average throughput of edge users is not lower than a preset throughput threshold, or the overall signal-to-interference-plus-noise ratio of the cell is not lower than a preset signal-to-interference-plus-noise ratio threshold; if any condition is not met for multiple consecutive scheduling cycles, the current strategy is determined to be invalid and model reconstruction is triggered.

[0012] In some embodiments, the method is applicable to heterogeneous network environments, wherein access nodes include macro base stations, micro base stations, and small base stations, and each type of node adopts a differentiated strategy when performing broadcast operations: macro base stations broadcast to all neighboring cells, micro base stations broadcast only to the same type or macro base stations, and small base stations respond only to query requests from macro base stations.

[0013] In some embodiments, the method further includes: during the initial network deployment phase, generating an interference pattern library for typical high-density scenarios through offline simulation, wherein the interference pattern library stores several typical interference feature matrix templates; during the operation phase, matching the interference feature matrix constructed in real time with the pattern library, and if the similarity is greater than a preset similarity threshold, directly invoking the preset coordination strategy of the corresponding template.

[0014] In some embodiments, the method further includes: assigning an interference tolerance level label to each user, the label being set according to the service type; and prioritizing resource isolation for users with low tolerance levels during resource allocation to ensure they are not affected by high interference.

[0015] In some embodiments, the low-tolerance level users include ultra-reliable low-latency communication service users, and the high-tolerance level users include voice service users; when constructing a feasible resource candidate set, all spatiotemporal frequency units with interference levels not lower than the acceptable lower limit in the high-interference resource identifiers of neighboring cells are excluded, and resource units with interference intensity indicators exceeding the local threshold within the cell are additionally excluded for low-tolerance level users.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) Breaking through the limitations of single-dimensional coordination: By constructing a three-dimensional interference perception model that combines time domain, frequency domain, and spatial domain, it can accurately identify composite interference sources generated by multiple cells in high-density scenarios, and realize three-dimensional characterization and localization of interference. Compared with traditional methods that rely solely on time domain blanking or frequency domain static division, the accuracy of interference identification is significantly improved. Dynamic resource joint optimization: The timing and frequency selection and beam direction are considered simultaneously during the resource allocation stage, avoiding the suboptimal solution problem of allocating resources first and then adjusting the beam in the traditional scheme, which greatly improves the spectrum efficiency, especially in densely populated user areas.

[0017] (2) Lightweight collaboration mechanism: adopts a sparse interaction protocol, only exchanges high interference resource identifiers instead of full channel state information, which greatly reduces the signaling load between base stations. In ultra-dense deployment scenarios, the signaling overhead is controlled within a low proportion of the total system overhead, effectively avoiding signaling storms; Heterogeneous network compatibility design: Adapts to heterogeneous architectures where macro stations, micro stations, and small stations coexist through a layered broadcast strategy. No unified coordination center is required, and plug-and-play deployment is supported. Network scalability is significantly enhanced, making it suitable for complex and high-density scenarios such as urban hotspots and sports venues.

[0018] (3) Real-time policy update mechanism: Introducing closed-loop verification based on actual performance indicators, when the network environment changes suddenly (such as users moving quickly or sudden traffic) and coordination fails, the model can be automatically reconstructed. The response delay meets the millisecond-level real-time requirements, which is far superior to the second-level response of the traditional semi-static coordination scheme. Business-aware resource assurance: Differentiated scheduling based on user interference tolerance levels ensures service quality for critical services such as ultra-reliable low latency, significantly reduces the standard deviation of throughput fluctuations for edge users, and significantly improves user experience consistency. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 This is an overall flowchart of an interference coordination communication method for high-density scenarios proposed in this invention; Figure 2 This is a schematic diagram illustrating the core principle framework of the multi-dimensional interference perception model and lightweight collaborative negotiation mechanism in this invention. Figure 3 This is a flowchart illustrating the joint optimization logic of dynamic resource avoidance strategy generation and adaptive beamforming in this invention. Figure 4 This is a schematic diagram of the multi-level interaction relationship and data flow of the closed-loop performance verification and strategy update mechanism in this invention. Detailed Implementation Example

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0021] like Figure 1 As shown, an interference coordination communication method for high-density scenarios includes the following specific steps: Step 1: At each access node, based on historical scheduling logs, neighboring cell resource occupancy status, and user channel quality feedback, construct a three-dimensional interference feature matrix containing time, frequency, and spatial dimensions. The three-dimensional interference feature matrix is ​​used to characterize the intensity and source of interference experienced by each resource unit in the current cell. Step 2: Each access node broadcasts only the high interference resource identifiers and their corresponding interference levels in its interference feature matrix that exceed a preset threshold to its neighboring nodes through a preset sparse interaction protocol, thus avoiding full information exchange. Step 3: Based on the received high-interference resource identifiers of neighboring cells and its own interference feature matrix, a multi-objective optimization algorithm is used to jointly allocate time-frequency resource blocks and beamforming directions, so that the scheduling of users in this cell avoids high-interference time-frequency regions, while minimizing the reverse interference caused to neighboring cells. Step 4: Based on the current user distribution and the location of the interference source, dynamically adjust the beam main lobe pointing and null position of the large-scale multiple input multiple output antenna array to concentrate the useful signal energy in the direction of the target user and form a deep suppression null in the direction of the interference source. Step 5: At the end of each scheduling cycle, collect the actual received signal-to-interference-plus-noise ratio and throughput data, evaluate the effectiveness of the current interference coordination strategy, and if the performance index is lower than the preset threshold, trigger a return to rebuild the interference perception model and update the coordination strategy.

[0022] In some embodiments, the process of constructing the three-dimensional interference feature matrix includes: dividing the system bandwidth into several subcarrier groups as frequency domain units using scheduling subframes as time units, and combining the spatial beam index of the large-scale multiple-input multiple-output system as spatial domain units to form a three-dimensional grid structure; for each grid unit, the ratio of the average received power from neighboring users to the target channel gain of the users in the current cell is calculated as the interference intensity index of that unit.

[0023] In some embodiments, the sparse interaction protocol specifies that: only when the interference intensity index of a resource unit exceeds a preset threshold, its identifier and interference level are encoded into information of a predetermined bit length and broadcast; the broadcast period is a preset interval of multiple scheduling subframes, and it is only sent to adjacent access nodes whose geographical distance is less than a predetermined distance.

[0024] In some embodiments, the objective function of the multi-objective optimization algorithm is a weighted sum of maximizing the total system throughput and minimizing the cross-cell interference power. The weight coefficients are dynamically adjusted according to the current network load. The total system throughput is calculated by Shannon's formula, and the cross-cell interference power is obtained by weighted summation of the occupancy of high-interference resource identifiers in neighboring cells.

[0025] In some embodiments, the beamforming uses sparse channel estimation results based on compressed sensing to determine the spatial angle of the interference source, and uses the minimum mean square error criterion to calculate the optimal precoding vector, so that while the gain in the direction of the target user is not lower than a preset gain threshold, the radiated power attenuation in the direction of the interference source is greater than or equal to a preset attenuation threshold.

[0026] In some embodiments, the preset threshold for performance verification includes: the average throughput of edge users is not lower than a preset throughput threshold, or the overall signal-to-interference-plus-noise ratio of the cell is not lower than a preset signal-to-interference-plus-noise ratio threshold; if any condition is not met for multiple consecutive scheduling cycles, the current strategy is determined to be invalid and model reconstruction is triggered.

[0027] In some embodiments, the method is applicable to heterogeneous network environments, wherein access nodes include macro base stations, micro base stations, and small base stations, and each type of node adopts a differentiated strategy when performing broadcast operations: macro base stations broadcast to all neighboring cells, micro base stations broadcast only to the same type or macro base stations, and small base stations respond only to query requests from macro base stations.

[0028] In some embodiments, the method further includes: during the initial network deployment phase, generating an interference pattern library for typical high-density scenarios through offline simulation, wherein the interference pattern library stores several typical interference feature matrix templates; during the operation phase, matching the interference feature matrix constructed in real time with the pattern library, and if the similarity is greater than a preset similarity threshold, directly invoking the preset coordination strategy of the corresponding template.

[0029] In some embodiments, the method further includes: assigning an interference tolerance level label to each user, the label being set according to the service type; and prioritizing resource isolation for users with low tolerance levels during resource allocation to ensure they are not affected by high interference.

[0030] In some embodiments, the low-tolerance level users include ultra-reliable low-latency communication service users, and the high-tolerance level users include voice service users; when constructing a feasible resource candidate set, all spatiotemporal frequency units with interference levels not lower than the acceptable lower limit in the high-interference resource identifiers of neighboring cells are excluded, and resource units with interference intensity indicators exceeding the local threshold within the cell are additionally excluded for low-tolerance level users.

[0031] like Figure 2As shown, in the above-mentioned interference coordination communication method for high-density scenarios, step 1, constructing a multi-dimensional interference perception model, includes: at each access node, based on historical scheduling logs, neighboring cell resource occupancy status, and user channel quality feedback, constructing a three-dimensional interference feature matrix containing time, frequency, and spatial dimensions. This three-dimensional interference feature matrix is ​​used to characterize the intensity and source of interference experienced by each resource unit within the current cell. Specifically, the process of constructing the three-dimensional interference feature matrix uses scheduling subframes as the time unit, dividing the system bandwidth into several subcarrier groups as frequency domain units, and combining them with the spatial beam index of a large-scale multiple-input multiple-output system as spatial domain units, forming a three-dimensional grid structure. For each grid unit, the ratio of the average received power from neighboring cell users to the target channel gain of the user in this cell is calculated as the interference intensity index for that unit.

[0032] In implementation, each access node is equipped with a local interference sensing module. This module periodically retrieves historical scheduling logs from the physical layer scheduler. The logs include allocation records for each resource block in the past N scheduling subframes (N is a positive integer, typically 8 to 32), the corresponding user identifier, transmit power, and actual received signal-to-interference-plus-noise ratio (SINR) feedback value. Simultaneously, this module receives resource occupancy status summary information broadcast by neighboring access nodes via the X2 interface or a dedicated backhaul link. This summary information includes scheduling indications and approximate transmit power levels for neighboring cells on the same time-frequency resources. Furthermore, user equipment (UE) periodically reports channel quality indication (CQI), precoding matrix indication (PMI), and rank indication (RI) via the uplink control channel. This information is used to deduce the target channel gain.

[0033] The data structure of the three-dimensional interference feature matrix is ​​defined as a three-dimensional array M[t, f, s], where t ∈ {0,1, ..., T-1} represents the time dimension, corresponding to T consecutive scheduling subframes; f ∈ {0, 1, ..., F-1} represents the frequency domain dimension, corresponding to F subcarrier groups, each subcarrier group consisting of 12 consecutive subcarriers, conforming to the Physical Resource Block (PRB) definition of the fifth-generation mobile communication system (5G NR); s ∈ {0, 1, ..., S-1} represents the spatial domain dimension, corresponding to S predefined spatial beam directions, the value of S being determined by the physical structure of the massive MIMO antenna array, typically 16, 32, or 64. For each three-dimensional grid cell (t, f, s), its interference intensity index I(t, f, s) is calculated as follows: First, the downlink received power of the user u in this cell scheduled at time t, frequency band f, and beam s is extracted from the historical scheduling log. Secondly, identify the set V of users scheduled by neighboring cells on the same spatiotemporal frequency unit from the resource occupancy status of neighboring cells; then, for each v ∈ V, utilize its reported CQI and known transmit power. v, combined with the path loss model, estimates the interference power it generates at the receiver in this cell. Finally, calculate the total interference power. and order If the cell is not scheduled by this cell, I(t, f, s) is set to 0 or a very small default value. Each element of the matrix M is stored as a floating-point number and is periodically updated with a sliding window, discarding the oldest time slice data and incorporating the observations from the latest scheduling period.

[0034] In the aforementioned interference coordination communication method for high-density scenarios, step 2 performs lightweight collaborative negotiation, including: each access node broadcasts, through a preset sparse interaction protocol, only to neighboring nodes the high-interference resource identifiers and their corresponding interference levels that exceed a preset threshold in its interference feature matrix, avoiding full information exchange and thus reducing signaling overhead. Specifically, the sparse interaction protocol stipulates that: only when the interference intensity index of a resource unit exceeds the preset threshold is its identifier and interference level encoded into information of a predetermined bit length and broadcast; the broadcast period is a preset interval of multiple scheduling subframes, and it is only sent to neighboring access nodes with a geographical distance less than a predetermined distance, in order to limit the signaling propagation range.

[0035] In engineering implementation, each access node maintains an interference threshold θ, which can be configured as a fixed value (e.g., 3.0, corresponding to a 10 dB interference-to-signal ratio) or dynamically adjusted according to the current network load. When any element I(t, f, s) in the three-dimensional interference feature matrix M > θ, the spatiotemporal frequency unit (t, f, s) is marked as a high-interference resource unit. The high-interference resource identifier is represented by a triplet. Composition, in which , The number of subframes corresponding to the broadcast period (typically 4, 8, or 16) is used to compress time-dimensional information. The interference level L is divided into multiple discrete levels based on the numerical range of I(t, f, s), for example: L=1 (low interference, 1.0 < I ≤ 2.0), L=2 (medium interference, 2.0 < I ≤ 4.0), L=3 (high interference, I > 4.0). Each high-interference resource identifier and its interference level are encoded into a fixed-length message unit, with a typical encoding scheme as follows: occupies 2 bits, f occupies 6 bits (supports up to 64 frequency domain units), s occupies 6 bits (supports up to 64 spatial domain units), and L occupies 2 bits, for a total of 16 bits.

[0036] Broadcast operations are performed by the coordination negotiation module of the access node. This module maintains a list of neighboring cells, and each neighboring cell entry in the list includes its physical location coordinates, type (macro base station, micro base station, or small base station), and maximum communication distance. At the start of each broadcast cycle, the module iterates through all high-interference resource units, generates corresponding encoded messages, and filters receivers based on the geographical distance between neighboring cells: receivers are selected only if the Euclidean distance between the current node and its neighboring cells is equal. Only when the time is right will the message be added to the sending queue of that neighboring cell. The value is set differently depending on the node type: for macro base stations, Set to twice the cell radius; for micro base stations, Set to 500 meters; for small base stations, The distance is set to 100 meters. Messages are sent via a dedicated lightweight signaling channel (such as the PC5 interface in 5G NR or a custom UDP / IP tunnel) using an unacknowledged best-effort transmission mode to further reduce overhead. The receiving node parses the received high-interference resource identifiers and stores them in its local neighbor cell interference buffer for subsequent resource allocation.

[0037] like Figure 3 As shown, in the above-mentioned interference coordination communication method for high-density scenarios, step 3, generating a dynamic resource avoidance strategy, includes: based on the received high-interference resource identifiers of neighboring cells and its own interference feature matrix, using a multi-objective optimization algorithm to jointly allocate time-frequency resource blocks and beamforming directions, so that user scheduling in this cell avoids high-interference spatio-frequency regions, while minimizing reverse interference to neighboring cells. Specifically, the objective function of the multi-objective optimization algorithm is to maximize the total system throughput and minimize the weighted sum of cross-cell interference power. The weight coefficients are dynamically adjusted according to the current network load. The total system throughput is calculated using the Shannon formula, and the cross-cell interference power is obtained by weighted summation of the occupancy of high-interference resource identifiers in neighboring cells.

[0038] In practice, the resource avoidance strategy generation module first integrates two types of input data: one is the three-dimensional interference feature matrix M of the current node, and the other is the set of high-interference resource identifiers received and cached from neighboring cells. For each user u to be scheduled, the module constructs a feasible resource candidate set R_u, which excludes all spatiotemporal frequency units that satisfy any of the following conditions: (1) ( (2) (This is the internal interference threshold of this cell). And the corresponding interference level ( The acceptable disturbance level is the lower bound, typically 2). Subsequently, the module establishes a mixed-integer programming problem, with decision variables... Let J be a binary variable representing whether user u is assigned to cell (t, f, s). The objective function J is defined as: The first item is the total system throughput, and B is the bandwidth (in Hz) of each subcarrier group. The transmit power allocated to user u, For the target channel gain, For noise power spectral density, The first term is the interference power of the cell (derived from M[t,f,s]); the second term is the cross-cell interference penalty term. The interference weight for cell (t,f,s) is equal to the number of times the cell appears in the high interference flag of the neighboring cell multiplied by the corresponding interference level L. This is a weighting coefficient, with a value range of [0,1], based on the current network load. Dynamic adjustment: when , (Focusing on throughput); when (Focus on interference suppression).

[0039] The constraints include: (1) Each user u must be allocated at least one resource unit in each scheduling cycle; (2) The same time-frequency unit cannot be allocated to multiple users (single-user MIMO scenario) or multiple stream users (orthogonality must be satisfied in multi-user MIMO scenario); (3) The total transmit power must not exceed the maximum power of the base station. This optimization problem is solved within a time limit using heuristic algorithms (such as genetic algorithms or simulated annealing), and the output is the optimal resource allocation scheme and the corresponding beam index s for each user.

[0040] like Figure 4 As shown, in the above-mentioned interference coordination communication method for high-density scenarios, step 4 implements adaptive beamforming, which includes: dynamically adjusting the main lobe pointing and null position of the large-scale multiple-input multiple-output antenna array according to the current user distribution and the location of the interference source, concentrating the useful signal energy in the direction of the target user, and forming a deep suppression null in the direction of the interference source. Specifically, the beamforming uses the sparse channel estimation results based on compressed sensing to determine the spatial angle of the interference source, and uses the minimum mean square error criterion to calculate the optimal precoding vector, so that while the gain in the direction of the target user is not lower than a preset gain threshold, the radiated power attenuation in the direction of the interference source is greater than or equal to a preset attenuation threshold.

[0041] In terms of engineering details, the adaptive beamforming module first obtains the sparse channel impulse response (CIR) from the channel estimation unit. This CIR is reconstructed from a finite number of pilot symbols using compressed sensing techniques. Its non-zero taps correspond to the main multipath components, each containing the angle of arrival (AoA) and complex gain information. The direction of the target user is determined by the AoA of its dominant path, denoted as . The direction of the interference source is determined by analyzing the spatial index 's' in the high-interference resource identifier of the neighboring cell and combining it with the beam pattern of the antenna array in this cell, to obtain the set of interference incident angles through reverse mapping. .

[0042] The precoding vector w is calculated based on the minimum mean square error (MMSE) criterion, and its optimization objective is: in, Let θ be the equivalent channel vector for the target user, and a(θ) be the steering vector of the antenna array. For noise variance, The preset zero-dimple depth threshold is used (typically -20 dB). This problem is solved using the Lagrange multiplier method or the convex optimization toolbox, and the obtained w ensures that... Directional gain (γ is a preset gain threshold, such as 0 dB), and at the same time in all The k-direction satisfies the zero trap constraint. The calculated precoding vector is loaded into the precoding module of the baseband processor and takes effect in the next scheduling cycle.

[0043] In the aforementioned interference coordination communication method for high-density scenarios, step 5, which performs closed-loop performance verification and policy update, includes: at the end of each scheduling cycle, collecting actual received signal-to-interference-plus-noise ratio (SIR) and throughput data, evaluating the effectiveness of the current interference coordination policy, and if the performance indicators are lower than a preset threshold, triggering a return to step 1 to reconstruct the interference perception model and update the coordination policy. Specifically, the preset thresholds for performance verification include: the average throughput of edge users is not lower than a preset throughput threshold, or the overall SIR of the cell is not lower than a preset SIR threshold; if neither condition is met for multiple consecutive scheduling cycles, the current policy is deemed invalid, triggering model reconstruction.

[0044] At the implementation level, the performance monitoring module collects actual throughput reports and SINR measurements from user equipment at the end of each scheduling cycle. Edge users are defined as the bottom 10% of users in terms of throughput within the cell. The module calculates the average throughput of edge users. and the overall average of the community The preset thresholds include: ( Throughput thresholds determined by business needs, such as 1 Mbps) and ( (Typical value is 0 dB). The verification logic is: if and If the counter increments by 1, then counter C is incremented by 1; otherwise, C is reset to 0. ( If the number of consecutive failure cycles to be tolerated is typically 3, then the strategy is deemed to have failed, triggering a return to step 1.

[0045] Furthermore, the method is applicable to heterogeneous network environments, where access nodes include macro base stations, micro base stations, and small base stations. Each type of node employs a differentiated broadcasting strategy during step 2: macro base stations broadcast to all neighboring cells, micro base stations broadcast only to similar or macro base stations, and small base stations only respond to query requests from macro base stations, thus achieving hierarchical coordination. During the initial network deployment phase, an interference pattern library for typical high-density scenarios is generated through offline simulation. This library stores several typical interference feature matrix templates. During the runtime phase, the real-time constructed interference feature matrix is ​​matched with the pattern library. If the similarity exceeds a preset similarity threshold, the preset coordination strategy of the corresponding template is directly invoked, accelerating the strategy generation process. An interference tolerance level label is assigned to each user. This label is set according to the service type, such as high tolerance for voice services and low tolerance for ultra-reliable low-latency services. During resource allocation in step 3, priority is given to isolating low-tolerance users to ensure they are not affected by high interference.

[0046] Specific application example: Consider a high-density deployment scenario in a sports stadium, including 1 macro base station (covering the entire stadium), 8 micro base stations (covering the stands), and 32 small base stations (covering VIP boxes). The system bandwidth is 100MHz, divided into 50 PRBs (F=50), with a scheduling subframe length of 1ms (T=16 corresponding to a 16ms window), and a 64-element massive MIMO antenna array (S=32 beams). During a certain scheduling period, the macro base station detects high interference (I>4.0) in PRBs 20-30 in the beam 15-20 direction, and therefore broadcasts an identifier (...). =0, f=25, s=18, L=3). After receiving the data, the micro base station avoids this spatiotemporal frequency region during resource allocation. Simultaneously, an Ultra-Reliable Low-Latency Communication (URLLC) user (with low interference tolerance) is preferentially allocated to the clean resources of PRB 5-10 and beam 5. The beamforming module, based on channel estimation, creates a -25 dB null in the θ=120° direction to suppress interference from neighboring cells. Performance verification revealed that the edge user throughput was below 500 kbps for three consecutive cycles, triggering model reconstruction and rebuilding the interference-aware model. Example

[0047] In another embodiment, the method is applied to a heterogeneous network in an urban hotspot area. Macro base stations are deployed at street intersections, micro base stations are installed on lampposts, and small base stations are embedded indoors in shopping malls. The differentiated broadcasting strategy is implemented as follows: macro base stations broadcast high-interference flags to all neighboring cells (including micro and small base stations) within a 500-meter radius; micro base stations only broadcast to other micro base stations and macro base stations, ignoring small base stations; small base stations do not actively broadcast, but only reply with local interference information when they receive a specific query request from a macro base station (such as "Please report the interference situation of PRB 40-45"). The interference pattern library is generated offline through Monte Carlo simulation and includes 10 typical scenario templates such as "peak traffic at intersections" and "shopping mall weekend promotions". During online operation, the similarity between the real-time interference matrix and the templates is calculated using cosine similarity. If the similarity is >0.85, a preset strategy is directly loaded, reducing the strategy generation latency from 50ms to 5ms. Example

[0048] In the third embodiment, the method is integrated into the baseband processing unit (BBU) of a 5G base station. The BBU's software architecture includes five functional modules: an interference sensing module, a cooperative negotiation module, a resource optimization module, a beamforming module, and a performance verification module. These modules interact through shared memory and message queues. The interference sensing module updates the three-dimensional matrix every 1 ms; the cooperative negotiation module performs a broadcast every 4 ms; the resource optimization module completes the solution within 2 ms; the beamforming module utilizes FPGA to accelerate precoding calculations; and the performance verification module triggers an evaluation every 10 ms. The entire process meets the 1 ms scheduling requirement of ultra-low latency communication (uRLLC), and the signaling overhead is controlled within 3% of the total overhead.

[0049] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

[0050] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for interference coordination communication in high-density scenarios, characterized in that, Includes the following steps: At each access node, a three-dimensional interference feature matrix is ​​constructed based on historical scheduling logs, neighboring cell resource occupancy status, and user channel quality feedback. This matrix includes time, frequency, and spatial dimensions. The three-dimensional interference feature matrix is ​​used to characterize the intensity and source of interference experienced by each resource unit in the current cell. Each access node broadcasts only the high-interference resource identifiers and their corresponding interference levels in its interference feature matrix that exceed a preset threshold to its neighboring nodes through a preset sparse interaction protocol, thus avoiding full information exchange. Based on the received high-interference resource identifiers of neighboring cells and its own interference feature matrix, a multi-objective optimization algorithm is used to jointly allocate time-frequency resource blocks and beamforming directions, so that the scheduling of users in this cell avoids high-interference time-space-frequency regions, while minimizing the reverse interference caused to neighboring cells. Based on the current user distribution and the location of the interference source, the beam main lobe pointing and null position of the large-scale multiple input multiple output antenna array are dynamically adjusted to concentrate the useful signal energy in the direction of the target user and form a deep suppression null in the direction of the interference source. At the end of each scheduling cycle, the actual received signal-to-interference-plus-noise ratio and throughput data are collected to evaluate the effectiveness of the current interference coordination strategy. If the performance indicators are lower than the preset threshold, the system will return to rebuild the interference perception model and update the coordination strategy.

2. The interference coordination communication method for high-density scenarios according to claim 1, characterized in that, The process of constructing the three-dimensional interference feature matrix includes: using scheduling subframes as time units, dividing the system bandwidth into several subcarrier groups as frequency domain units, and combining them with the spatial beam index of the large-scale multiple-input multiple-output system as spatial domain units to form a three-dimensional grid structure; for each grid unit, the ratio of the average received power from neighboring users to the target channel gain of the users in the current cell is calculated as the interference intensity index of that unit.

3. The interference coordination communication method for high-density scenarios according to claim 1, characterized in that, The sparse interaction protocol stipulates that: only when the interference intensity index of a resource unit exceeds a preset threshold, its identifier and interference level are encoded into information of a predetermined bit length and broadcast; the broadcast period is a preset interval of multiple scheduling subframes, and it is only sent to adjacent access nodes whose geographical distance is less than a predetermined distance.

4. The interference coordination communication method for high-density scenarios according to claim 1, characterized in that, The objective function of the multi-objective optimization algorithm is to maximize the total system throughput and minimize the cross-cell interference power. The weight coefficients are dynamically adjusted according to the current network load. The total system throughput is calculated by Shannon's formula, and the cross-cell interference power is obtained by weighted summation of the occupancy of high interference resource identifiers in neighboring cells.

5. The interference coordination communication method for high-density scenarios according to claim 1, characterized in that, The beamforming uses sparse channel estimation results based on compressed sensing to determine the spatial angle of the interference source, and uses the minimum mean square error criterion to calculate the optimal precoding vector, so that while the gain in the direction of the target user is not lower than a preset gain threshold, the radiated power attenuation in the direction of the interference source is greater than or equal to a preset attenuation threshold.

6. The interference coordination communication method for high-density scenarios according to claim 1, characterized in that, The preset thresholds for performance verification include: the average throughput of edge users is not lower than a preset throughput threshold, or the overall signal-to-interference-plus-noise ratio (SIR) of the cell is not lower than a preset SIR threshold; if any condition is not met for multiple consecutive scheduling cycles, the current strategy is deemed to be invalid and model reconstruction is triggered.

7. The interference coordination communication method for high-density scenarios according to claim 1, characterized in that, The method is applicable to heterogeneous network environments, where access nodes include macro base stations, micro base stations, and small base stations. Each type of node adopts a differentiated strategy when performing broadcast operations: macro base stations broadcast to all neighboring cells, micro base stations broadcast only to the same type or macro base stations, and small base stations only respond to query requests from macro base stations.

8. The interference coordination communication method for high-density scenarios according to claim 1, characterized in that, The method further includes: in the initial network deployment phase, generating an interference pattern library for typical high-density scenarios through offline simulation, wherein the interference pattern library stores several typical interference feature matrix templates; in the operation phase, matching the interference feature matrix constructed in real time with the pattern library, and if the similarity is greater than a preset similarity threshold, directly calling the preset coordination strategy of the corresponding template.

9. The interference coordination communication method for high-density scenarios according to claim 1, characterized in that, The method further includes: assigning an interference tolerance level label to each user, the label being set according to the service type; and prioritizing resource isolation for users with low tolerance levels during resource allocation to ensure they are not affected by high interference.

10. The interference coordination communication method for high-density scenarios according to claim 9, characterized in that, The low-tolerance level users include ultra-reliable low-latency communication service users, and the high-tolerance level users include voice service users; when constructing a feasible resource candidate set, all spatiotemporal frequency units with interference levels not lower than the acceptable lower limit in the high interference resource identifiers of all neighboring cells are excluded, and resource units with interference intensity indicators exceeding the local threshold within the cell are additionally excluded for low-tolerance level users.

Images