A PDCCH channel optimization processing system based on NR architecture

Abstract

The application relates to the technical field of communication, and discloses a PDCCH channel optimization processing system based on an NR architecture, which comprises a channel state information acquisition module used for acquiring dynamic channel characteristic parameters; a dynamic resource allocation module used for generating a resource block allocation strategy and screening a target resource block set based on the parameters through a reinforcement learning algorithm; an interference coordination module used for calculating an interference covariance matrix and generating interference suppression beam weights; a multi-user detection module used for carrying out serial interference cancellation through a non-orthogonal multiple access technology; and modules such as adaptive modulation and coding, channel estimation optimization, time-frequency synchronization, control information encapsulation, power allocation and a feedback mechanism. The modules work cooperatively, effectively solve the problems of the PDCCH channel under the NR architecture in terms of resource allocation, interference coordination, multi-user detection and the like, improve the performance of a communication system, and meet the requirements of 5G and future mobile communication networks.

Description

Technical Field

[0001] This invention relates to the field of communication technology, specifically to a PDCCH channel optimization processing system based on an NR architecture. Background Technology

[0002] With the widespread application and continuous development of 5G technology, the performance requirements for communication systems are increasing. In the 5G NR architecture, the PDCCH, as a key component, is responsible for transmitting downlink control information, such as scheduling and power control commands, and its performance directly affects the communication quality and efficiency of the entire system. However, the PDCCH channel under the current NR architecture faces many challenges.

[0003] In terms of channel state information acquisition, the wireless channel environment is complex and ever-changing. Factors such as multipath delay and Doppler shift cause channel characteristics to change continuously. Traditional channel state information acquisition methods struggle to obtain this dynamically changing information in real time and accurately, leaving subsequent resource allocation and signal processing without a reliable basis. For example, in high-speed mobile scenarios, Doppler shift causes frequency shifts in the signal, leading to a deterioration in received signal quality, and inaccurate channel state information cannot provide effective support for solving this problem.

[0004] Resource allocation is equally challenging. Under limited time-frequency resources, efficiently allocating PDCCH candidate resource blocks to different users while simultaneously meeting low latency requirements is a pressing problem. Existing resource allocation algorithms often fail to adequately consider dynamic channel changes and real-time user demands, leading to low resource utilization and compromised service quality for some users. For example, in scenarios with sudden surges in service requests, a large number of users simultaneously request resources; traditional algorithms may be unable to allocate resources quickly and reasonably, resulting in increased latency and data transmission interruptions.

[0005] Interference is also a significant factor limiting the performance of PDCCH channels. In multi-user communication environments, interference can easily occur between adjacent PDCCH resource blocks, reducing the reliability of signal transmission. This is especially true in densely populated cell deployment scenarios, where neighboring cell interference is more severe. Current interference coordination techniques are insufficient to effectively suppress interference, leading to increased bit error rate and deteriorated communication quality. For example, when multiple cells use the same frequency band simultaneously, neighboring cell interference can reduce the signal-to-noise ratio of the received signal, affecting the user's correct reception of control information.

[0006] In multi-user detection, as the number of users continues to increase, the performance of traditional multi-user detection technologies gradually declines when processing user signals in overlapping PDCCH resource blocks. Serial interference cancellation algorithms are complex and prone to error propagation, resulting in low detection accuracy and failing to meet the needs of large-scale user access. For example, in IoT scenarios, a large number of devices access the network simultaneously, and traditional multi-user detection technologies struggle to accurately separate the signals of different users, affecting the accuracy of data transmission.

[0007] Channel estimation and synchronization are also critical aspects affecting PDCCH channel performance. Due to the time-varying nature of wireless channels, channel estimation errors accumulate over time, leading to demodulation errors. Simultaneously, inaccurate time-frequency synchronization can cause phase and frequency deviations in the received signal, further degrading communication quality. Existing channel estimation and synchronization methods have poor adaptability in complex channel environments and cannot meet the requirements for high-precision communication.

[0008] Regarding control information encapsulation and power allocation, current technologies cannot dynamically adjust the PDCCH transmit power according to users' real-time needs and priorities, resulting in insufficient signal strength or wasted power for some users. Furthermore, the encapsulation method of control information can also affect transmission efficiency and reliability. For example, when transmitting data from high-priority users, the inability to increase transmit power in a timely manner can lead to data transmission failure.

[0009] Regarding feedback mechanisms, existing mechanisms suffer from issues such as untimely response and unreasonable retransmission scheduling when handling ACK / NACK feedback from user equipment. This leads to increased data transmission latency and reduces overall system performance. For example, when a user equipment sends a NACK, retransmission scheduling cannot be performed quickly and accurately, affecting data real-time performance. Summary of the Invention

[0010] The purpose of this invention is to provide a PDCCH channel optimization processing system based on an NR architecture to solve the problems mentioned in the background art.

[0011] To achieve the above objectives, the present invention provides the following technical solution: a PDCCH channel optimization processing system based on an NR architecture, the system comprising:

[0012] The channel state information acquisition module is used to acquire the wireless channel state information of the user equipment in real time. The wireless channel state information includes multipath delay, Doppler frequency shift, and singular value decomposition eigenvalues ​​of the channel matrix, and the wireless channel state information is marked as dynamic channel characteristic parameters.

[0013] The dynamic resource allocation module is used to generate an allocation strategy for PDCCH candidate resource blocks based on the dynamic channel characteristic parameters within a preset time-frequency resource window using a reinforcement learning algorithm, and to filter a set of target resource blocks that meet low latency constraints.

[0014] The specific method for selecting the set of target resource blocks that meet the low latency constraint is as follows:

[0015] Retrieve historical PDCCH resource block allocation records from the local database, extract the average latency feature value of each resource block, and construct a latency feature matrix.

[0016] The latency optimization weights for each candidate resource block are calculated using the Q-learning algorithm;

[0017] If the latency optimization weight of a resource block is greater than the preset threshold θ, then the resource block is added to the target resource block set.

[0018] Preferably, the dynamic resource allocation module further includes:

[0019] The interference coordination module is used to calculate the interference covariance matrix between adjacent PDCCH resource blocks based on the spatial location information of user equipment, and to generate interference suppression beam weights through singular value decomposition.

[0020] The specific method for calculating the interference covariance matrix is ​​as follows:

[0021] Obtain the emission signal vector s of each resource block in the target resource block set. m and neighboring cell interference signal vector j n ;

[0022] Constructing the interference covariance matrix Where σ 2 Let I be the noise power, and I be the identity matrix. s m The conjugate transpose of M; M is the number of resource blocks currently occupied by the serving user in the target resource block set, and N is the number of PDCCH signal sources that generate interference in the neighboring cells;

[0023] Eigenvalue decomposition is performed on R, and the eigenvector corresponding to the largest eigenvalue is extracted as the interference suppression beam weight.

[0024] Preferably, the interference coordination module further includes:

[0025] The dynamic beamforming module is used to adjust the phase and amplitude parameters of the PDCCH transmit beam based on the interference suppression beam weights. The specific adjustment method is as follows:

[0026] Obtain the azimuth angle φ and elevation angle θ of the target user equipment from the local database, and calculate the initial beam pointing vector v0 = [ej2πφ ,e j2πθ ] T ;

[0027] The beam weights w are optimized using the gradient descent algorithm, such that the objective function f(w) = ||w||. H ·v0‖ 2 -λ‖w H Maximize ·R·w‖, where λ is the interference suppression factor, w H Let w be the conjugate transpose of w, and v0 be the initial beam pointing vector, calculated from the azimuth angle φ and elevation angle θ of the user equipment. H ·R·w represents the weighting effect of the beam weight vector w on the interference covariance matrix R.

[0028] Preferred options also include:

[0029] The multi-user detection module is used to cancel serial interference in user signals in overlapping PDCCH resource blocks based on non-orthogonal multiple access technology. The specific method is as follows:

[0030] The initial detection signal is obtained by performing linear minimum mean square error estimation on the received signal y.

[0031] Calculate residual interference Where H1 is the first user channel matrix;

[0032] Iteratively execute the above steps until the signal of the Kth user satisfies ||r|| K || 2 <∈, where ∈ is the preset error threshold.

[0033] Preferably, the multi-user detection module further includes:

[0034] The adaptive modulation and coding module is used to dynamically select the modulation and coding strategy (MCS) based on the channel quality indicator (CQI). The specific selection method is as follows:

[0035] Obtain the bit error rate (BER) curves corresponding to each MCS level from the local database, and fit the relationship function f between CQI and BER. BER (c);

[0036] The binary search method is used to find the condition f. BER (c) The maximum MCS level ≤ γ, where γ is the upper limit of the BER tolerated by the system.

[0037] Preferred options also include:

[0038] The channel estimation optimization module is used to predict and compensate for CSI using a convolutional neural network (CNN). The specific method is as follows:

[0039] Construct a training dataset containing historical CSI sequences, input it into a CNN network for feature extraction, and output a predicted CSI matrix.

[0040] Calculate CSI compensation coefficient If η > δ, then real-time channel re-estimation is triggered, where ‖·‖ F δ represents the Frobenius norm, used to measure matrix differences, and δ is the compensation coefficient threshold.

[0041] Preferably, the channel estimation optimization module further includes:

[0042] The time-frequency synchronization module is used to generate a synchronization signal from the Zadov-Chu sequence. The specific method is as follows:

[0043] Generate baseband sequence z(n) = e -jπμn(n+1) / L , where μ is the root index and L is the sequence length;

[0044] Cyclic correlation is performed on the received signal to extract the peak position as the time-frequency offset estimate.

[0045] Preferably, it also includes: a control information encapsulation module, used to map downlink control information (DCI) to PDCCH resource elements, the specific method of which is as follows:

[0046] Perform CRC appending and polar code encoding on the DCI bitstream to generate the encoded codeword c;

[0047] The subcarrier index set of the target resource block is mapped to c through orthogonal frequency division multiplexing modulation.

[0048] Preferably, the control information encapsulation module further includes:

[0049] The power allocation module is used to dynamically adjust the PDCCH transmit power according to user priority. The specific method is as follows:

[0050] Retrieve user priority weights from the local database and calculate the power allocation ratio. Where p u For the power allocation ratio of the u-th user, P total For the total transmit power, ω u Let represent the priority weight of the u-th user, where U is the total number of users.

[0051] Preferably, the system further includes:

[0052] The feedback mechanism module is used to receive ACK / NACK feedback from user equipment via the uplink control channel PUCCH. Its specific processing method is as follows:

[0053] Parse the mixed automatic retransmission request process ID in the feedback information. If it is NACK, retrieve the original data packet from the local cache and trigger retransmission scheduling.

[0054] Compared with the prior art, the beneficial effects of the present invention are:

[0055] In terms of channel state information acquisition and utilization, the channel state information acquisition module acquires dynamic channel characteristic parameters in real time, such as multipath delay, Doppler shift, and singular value decomposition eigenvalues ​​of the channel matrix. This enables the system to accurately grasp the real-time changes of the wireless channel, providing a reliable basis for subsequent resource allocation and signal processing.

[0056] The dynamic resource allocation module, based on dynamic channel characteristic parameters, utilizes reinforcement learning algorithms to generate allocation strategies for PDCCH candidate resource blocks within a preset time-frequency resource window, and filters a set of target resource blocks that meet low-latency constraints. This module constructs a latency feature matrix by retrieving historical PDCCH resource block allocation records from a local database and calculates latency optimization weights using a Q-learning algorithm, enabling more rational resource allocation. In scenarios with sudden service disruptions, it significantly improves resource utilization, ensuring service quality for different users, especially for latency-sensitive services such as high-definition video calls and vehicle-to-everything (V2X) communication, effectively reducing stuttering and latency.

[0057] The interference coordination module effectively suppresses neighboring cell interference by calculating the interference covariance matrix between adjacent PDCCH resource blocks and generating interference suppression beam weights. It constructs the interference covariance matrix by obtaining the transmitted signal vectors of each resource block in the target resource block set and the neighboring cell interference signal vectors, and then extracts the eigenvector corresponding to the largest eigenvalue through singular value decomposition as the interference suppression beam weights. In dense cell deployment scenarios, this interference coordination technique significantly reduces the signal error rate and improves the reliability of signal transmission.

[0058] The dynamic beamforming module adjusts the phase and amplitude parameters of the PDCCH transmit beam based on the interference suppression beam weights. It calculates the initial beam pointing vector by acquiring the azimuth and elevation angles of the target user equipment, and optimizes the beam weights using a gradient descent algorithm to maximize the objective function. This further enhances the interference suppression capability, while improving the directionality and coverage of signal transmission, reducing signal blind spots in complex environments, and improving the user experience.

[0059] The multi-user detection module, based on non-orthogonal multiple access (NOMA) technology, eliminates serial interference in user signals within overlapping PDCCH resource blocks. By estimating linear minimum mean square error and iteratively calculating residual interference until a preset error threshold is met, the accuracy of multi-user detection is effectively improved. In large-scale user access scenarios such as the Internet of Things (IoT), compared to traditional multi-user detection technologies, it can accurately separate signals from more users, ensuring the accuracy of data transmission.

[0060] The adaptive modulation and coding module dynamically selects the modulation and coding strategy (MCS) based on the channel quality indicator (CQI). By acquiring the bit error rate (BER) curves corresponding to each MCS level, fitting the relationship function between CQI and BER, and using a binary search method to search for the maximum MCS level that satisfies the system's tolerance BER upper limit, dynamic optimization of the modulation and coding strategy is achieved. This allows the system to adjust the transmission mode in real time according to channel quality. When channel quality is good, it can increase the data transmission rate; when channel quality is poor, it can ensure transmission reliability and reduce the BER.

[0061] The channel estimation optimization module uses a convolutional neural network (CNN) to predict and compensate for channel indexes (CSI). A training dataset containing historical CSI sequences is constructed, input into the CNN for feature extraction, and the output is a predicted CSI matrix. CSI compensation coefficients are calculated, and if the coefficients exceed a threshold, real-time channel re-estimation is triggered. This effectively reduces channel estimation errors and improves the accuracy and real-time performance of channel estimation.

[0062] The time-frequency synchronization module generates a synchronization signal using the Zadov-Chu sequence, performs cyclic correlation operations on the received signal, and extracts the peak position as the time-frequency offset estimate. This method improves the accuracy of time-frequency synchronization, effectively corrects the phase and frequency deviations of the received signal, and ensures correct signal reception and processing.

[0063] The control information encapsulation module maps downlink control information (DCI) to PDCCH resource elements. Through operations such as CRC appending, polar code encoding, and orthogonal frequency division multiplexing modulation, it ensures the reliability and efficiency of control information transmission. The power allocation module dynamically adjusts the PDCCH transmit power based on user priority, improving power utilization efficiency. When transmitting data from high-priority users, it can promptly increase the transmit power to ensure successful data transmission. The feedback mechanism module receives ACK / NACK feedback from user equipment via the uplink control channel PUCCH, parses the hybrid automatic repeat request process ID, and if it is NACK, retrieves the original data packet from the local buffer and triggers retransmission scheduling. This enables the system to respond promptly to user feedback, optimize retransmission scheduling, reduce data transmission latency, and improve overall system performance. Attached Figure Description

[0064] Figure 1This is a schematic diagram illustrating the working principle of the PDCCH channel optimization processing system described in this invention.

[0065] Figure 2 This is a schematic diagram illustrating the working principle of the interference coordination module.

[0066] Figure 3 This is a schematic diagram illustrating the working principle of the multi-user detection module.

[0067] Figure 4 This is a schematic diagram illustrating the working principle of the channel estimation optimization module. Detailed Implementation

[0068] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0069] Please see Figure 1-4 This invention provides a technical solution: a PDCCH channel optimization processing system based on an NR architecture, the system comprising:

[0070] The channel state information acquisition module is responsible for acquiring real-time wireless channel state information from user equipment. This information includes multipath delay, Doppler shift, and singular value decomposition (SVD) eigenvalues ​​of the channel matrix. After acquisition, this information is marked as dynamic channel characteristic parameters for use by subsequent modules. Multipath delay reflects the time delay difference caused by different transmission paths during signal propagation, which affects signal reception quality and stability. Doppler shift is caused by changes in the received signal frequency due to user equipment movement, significantly impacting frequency synchronization in communication. The SVD eigenvalues ​​of the channel matrix characterize the channel's transmission properties, providing crucial information for resource allocation.

[0071] Dynamic resource allocation module: Based on the aforementioned dynamic channel characteristic parameters, it operates within a preset time-frequency resource window. It generates allocation strategies for PDCCH candidate resource blocks using a reinforcement learning algorithm. This algorithm intelligently explores the optimal resource allocation method based on the current channel state and historical experience. Subsequently, the dynamic resource allocation module also needs to screen the set of target resource blocks that meet low-latency constraints. The specific screening steps are as follows:

[0072] Historical PDCCH resource block allocation records are retrieved from the local database. These records contain relevant information about past resource block allocations. Next, the average latency characteristics of each resource block are extracted. These characteristics reflect the latency of different resource blocks during past usage. These average latency characteristics are then organized to construct a latency feature matrix, which provides the data foundation for subsequent calculations.

[0073] The latency optimization weights for each candidate resource block are calculated using the Q-learning algorithm. Q-learning is a reinforcement learning-based optimization algorithm that finds the optimal decision-making strategy through continuous trial and learning. In this invention, the latency optimization weights calculated using this algorithm measure the potential of each candidate resource block to meet low latency constraints.

[0074] A preset threshold is set. If the latency optimization weight of a resource block is greater than this threshold, it indicates that the resource block performs well in meeting low latency requirements. In this case, the resource block is added to the target resource block set. Through this series of filtering operations, the resulting target resource block set can better meet the system's low latency requirements and improve the efficiency and quality of data transmission.

[0075] The present invention will be further described below with reference to Examples 1 to 6:

[0076] Example 1:

[0077] In this invention, the dynamic resource allocation module includes an interference coordination module, which performs interference processing based on the spatial location information of the user equipment. First, the interference covariance matrix between adjacent PDCCH resource blocks is calculated. Then, the transmit signal vector s of each resource block in the target resource block set is obtained. m and neighboring cell interference signal vector j n , where s m This represents the transmission signal vector of the m-th resource block in the target resource block set. The value of m is determined by the number M of resource blocks currently occupied by the serving user in the target resource block set; j n This represents the interference signal vector of the nth PDCCH signal source in the neighboring cell. The value of n is determined by the number N of PDCCH signal sources in the neighboring cell.

[0078] Next, construct the interference covariance matrix. In this formula, σ 2 represents noise power, which reflects the noise intensity in the communication environment. The greater the noise power, the stronger the interference to signal transmission; I is the identity matrix, whose function is to preserve certain properties in matrix operations. s mThe conjugate transpose of is often used in matrix operations to handle related calculations of complex matrices; M is the number of resource blocks currently occupied by the serving user in the target resource block set, which determines the accumulation range of the transmitted signal vector when calculating the interference covariance matrix; N is the number of PDCCH signal sources that generate interference in the neighboring cells, which determines the accumulation range of the interference signal vector.

[0079] After constructing the interference covariance matrix, eigenvalue decomposition is performed on R, and the eigenvector corresponding to the largest eigenvalue is extracted as the interference suppression beam weight. The interference suppression beam weight obtained in this way can effectively suppress interference from neighboring cells and improve the quality of signal transmission.

[0080] The interference coordination module also includes a dynamic beamforming module, which adjusts the phase and amplitude parameters of the PDCCH transmit beam based on the interference suppression beam weights. It obtains the azimuth angle φ and elevation angle θ of the target user equipment from the local database and calculates the initial beam pointing vector v0 = [e j2πφ ,e j2πθ ] T Here, the azimuth angle φ and the elevation angle θ determine the position and orientation of the target user equipment in space. The initial beam pointing vector v0 calculated using these two angles provides the initial direction for subsequent beamforming optimization.

[0081] Then, the beam weights w are optimized using the gradient descent algorithm, so that the objective function f(w) = ||w|| H ·v0‖ 2 -λ‖w H Maximize ·R·w‖. In this objective function, λ is the interference suppression factor, which balances the relationship between beam pointing and interference suppression. A larger value results in stronger interference suppression, but may negatively impact beam pointing accuracy; w H v0 is the conjugate transpose of w; v0 is the initial beam pointing vector calculated earlier; w H ·R·w represents the weighting effect of the beam weight vector w on the interference covariance matrix R. By adjusting the beam weight vector w, the weighting effect on the interference covariance matrix can be changed, thereby achieving the purpose of interference suppression and beam pointing optimization.

[0082] Through the above interference coordination and beamforming optimization measures, the system can more effectively cope with neighboring cell interference, improve the reliability and stability of signal transmission, and provide users with better communication service quality.

[0083] Example 2:

[0084] The main function of this embodiment is to enhance the system's ability to detect multi-user signals and to dynamically adjust the modulation and coding strategy according to the channel quality, thereby improving the system's spectral efficiency and the accuracy of data transmission.

[0085] The multi-user detection module, based on non-orthogonal multiple access (NOMA) technology, performs serial interference cancellation on user signals in overlapping PDCCH resource blocks. First, linear minimum mean square error (LMSE) estimation is performed on the received signal y to obtain the initial detection signal x1. The received signal y is a mixed signal containing multiple user signals and noise. Through LSE estimation, the signal of the first user can be extracted from the mixed signal to a certain extent, thus obtaining the initial detection signal.

[0086] Next, the residual interference r = y - H1x1 is calculated, where H1 is the channel matrix of the first user. This formula is used to calculate the remaining interference signal after the first user signal is detected. This process is repeated iteratively until the signal of the Kth user satisfies ||r||. K || 2 <∈, where ∈ represents a preset error threshold. When the norm of the residual interference is less than the preset error threshold, the detection of the Kth user signal is considered to have reached a certain accuracy requirement, and iteration can stop. Through this serial interference cancellation method, multiple user signals in overlapping PDCCH resource blocks can be gradually separated, improving the system's ability to detect multiple user signals.

[0087] The multi-user detection module also includes an adaptive modulation and coding module, which dynamically selects the modulation and coding scheme (MCS) based on the channel quality indicator (CQI). Bit error rate (BER) curves for each MCS level are obtained from a local database; these curves reflect the BER performance of different MCS levels under different channel conditions. By analyzing and processing these curves, a relationship function f between CQI and BER is fitted. BER (c), where c represents the Channel Quality Indicator (CQI).

[0088] Then, a binary search is performed to find the condition f. BER (c) The maximum MCS level ≤ γ, where γ is the upper limit of the BER tolerated by the system. The bisection method is an efficient search algorithm that gradually approaches the maximum MCS level that meets the conditions by continuously dividing the search interval into two. When the bit error rate corresponding to the found MCS level does not exceed the upper limit of the BER tolerated by the system, this MCS level is the most suitable modulation and coding strategy under the current channel quality. Through this adaptive modulation and coding method, the system can adjust the modulation and coding strategy in real time according to changes in channel quality, improving the system's spectral efficiency while ensuring data transmission accuracy.

[0089] Example 3:

[0090] The channel estimation optimization module uses a convolutional neural network (CNN) to predict and compensate for channel state indices (CSI). First, a training dataset containing historical CSI sequences is constructed. These sequences record channel state information at different times in the past, which is crucial for training the CNN network. The training dataset is then input into the CNN network for feature extraction. CNNs have powerful feature extraction capabilities, able to extract valuable feature information from complex historical CSI sequences. After processing by the network, the predicted CSI matrix H is output.

[0091] Next, calculate the CSI compensation coefficient. In this formula, |·| F η represents the Frobenius norm, used to measure matrix discrepancies, accurately reflecting the degree of difference between two matrices. H represents the actual channel state information matrix, which is the CSI matrix predicted by a CNN network. If η > δ, where δ is the compensation coefficient threshold, it indicates a significant difference between the predicted CSI matrix and the actual channel state information matrix, triggering real-time channel re-estimation. This method allows for timely detection and correction of deviations between the predicted CSI matrix and the actual situation, improving the accuracy of channel state information.

[0092] The channel estimation and optimization module also includes a time-frequency synchronization module, which generates a synchronization signal using the Zadov-Chu sequence. First, a baseband sequence z(n) = e is generated. -jπμn(n+1) / L , where μ is the root index, which determines the characteristics of the Zadov-Chu sequence; different root indices will generate different sequences; L is the sequence length, which affects the performance and synchronization accuracy of the synchronization signal; j is the imaginary unit.

[0093] After generating the baseband sequence, a cyclic correlation operation is performed on the received signal. This operation effectively detects the correlation between signals. The peak position is extracted as an estimate of the time-frequency offset through the cyclic correlation operation. Time-frequency offset causes a deviation between the frequency and time of the received signal, affecting communication quality. By accurately estimating the time-frequency offset, the received signal can be adjusted accordingly, achieving system time-frequency synchronization and ensuring correct signal reception and processing.

[0094] Through channel estimation optimization and time-frequency synchronization measures, the system can obtain channel state information more accurately, achieve time-frequency synchronization, and improve the reliability and stability of communication.

[0095] Example 4:

[0096] The Control Information Encapsulation (DCI) module is responsible for mapping downlink control information (DCI) to PDCCH resource elements. First, the DCI bitstream undergoes CRC appending and polar code encoding to generate the encoded codeword c. CRC appending is used to detect errors during data transmission, improving data reliability; polar code encoding is an advanced channel coding technique that improves coding efficiency and error correction capabilities. After these two processes, the resulting encoded codeword c is more suitable for transmission in the channel.

[0097] Then, orthogonal frequency division multiplexing (OFDM) modulation maps c to the subcarrier index set S of the target resource block. OFDM is a highly efficient modulation technique that maps coded codewords to different subcarriers, enabling parallel transmission and improving spectral efficiency. In this way, downlink control information can be accurately mapped to PDCCH resource elements, preparing for subsequent signal transmission.

[0098] The control information encapsulation module also includes a power allocation module, which dynamically adjusts the PDCCH transmit power based on user priorities. User priority weights, reflecting the importance of different users in the system, are retrieved from a local database. The power allocation ratio is then calculated. In this formula, p u The power allocation ratio for the u-th user determines the proportion of the total transmission power that the u-th user can obtain; P total ω represents the total transmit power, which is the total transmit power that the system can provide; u Let represent the priority weight of the u-th user. The larger the weight, the higher the priority of the user. U represents the total number of users, which is used to calculate the sum of the priority weights of all users.

[0099] By dynamically adjusting the transmission power based on user priority, the system can prioritize the communication quality of high-priority users, while rationally allocating power resources, improving the system's power utilization efficiency, and ensuring stable operation under different user needs.

[0100] Example 5:

[0101] The feedback mechanism module receives ACK / NACK feedback from user equipment via the uplink control channel PUCCH. When user equipment correctly receives data, it sends ACK feedback via PUCCH; if an error occurs during data reception, it sends NACK feedback.

[0102] After receiving feedback information, the feedback mechanism module first parses the Hybrid Automatic Repeat Request (HARQ) process ID. The HARQ process ID identifies different data transmission processes; by parsing this ID, the system can accurately determine which data transmission process encountered a problem. If the parsed feedback is NACK, it indicates that an error occurred in this data transmission, and the user equipment failed to receive the data correctly. At this point, the system retrieves the original data packet from its local cache, which stores copies of previously sent data to ensure rapid data retrieval when retransmission is needed. After retrieving the original data packet, the system triggers retransmission scheduling, resending the original data packet to the user equipment, thereby ensuring accurate data reception and improving the reliability and accuracy of data transmission. Through this feedback mechanism, the system can promptly detect and correct errors in the data transmission process, ensuring communication quality and stability. In practical applications, such as high-speed mobile vehicle communication environments, signals are easily interfered with, increasing the probability of data transmission errors; in such cases, the feedback mechanism plays a particularly important role. When a vehicle passes through an area with signal obstruction at high speed while driving, some data may be lost or received incorrectly. The feedback mechanism can quickly detect these problems and ensure that the vehicle can receive the data completely through retransmission scheduling, maintaining smooth communication and providing reliable communication support for applications such as intelligent driving.

[0103] Example 6:

[0104] The purpose of this embodiment is to explain in detail how to use the Q-learning algorithm to calculate the latency optimization weight of each candidate resource block, so that the system can more accurately select target resource blocks that meet the low latency constraints and improve the efficiency and performance of PDCCH channel resource allocation.

[0105] During system operation, when the dynamic resource allocation module needs to filter the set of target resource blocks that meet the low latency constraints, it will use the Q-learning algorithm to calculate the latency optimization weight of each candidate resource block.

[0106] First, the system initializes the Q-learning algorithm. This includes defining the state space, action space, and reward function. The state space mainly consists of current wireless channel state information, the status of allocated resource blocks, and latency characteristics from historical PDCCH resource block allocation records. This information comprehensively reflects the current operating state of the system, providing a basis for algorithm decisions. For example, multipath delay and Doppler shift in the wireless channel state affect signal transmission latency, while previous resource block allocation and latency characteristics help the algorithm understand which resource blocks have historically exhibited good low-latency performance.

[0107] The action space refers to all possible operations for selecting a resource block from the candidate resource blocks. Each action corresponds to selecting a specific candidate resource block.

[0108] The design of the reward function is crucial, as it directly influences the learning direction of the algorithm. In this embodiment, the reward function is primarily set based on the impact of selecting a candidate resource block on system latency. If selecting a candidate resource block significantly improves system latency, bringing it close to or satisfying the low-latency constraint, a large positive reward is given; conversely, if selecting the resource block increases latency, moving it away from the low-latency target, a negative reward is given. For example, if selecting a resource block results in actual latency lower than the historical average latency and within the system's acceptable low-latency range, the algorithm receives a positive reward, encouraging it to select similar resource blocks more frequently in subsequent decisions.

[0109] After initialization, the algorithm begins iterative learning. In each iteration, the system selects an action from the action space based on the current state, i.e., selects a candidate resource block. The action selection strategy can employ an ε-greedy strategy. This strategy randomly selects an action with a certain probability (ε) to explore new resource block selection methods; it selects the action with the largest Q-value in the current state with a probability of (1-ε), i.e., selects the resource block currently considered optimal. In this way, the algorithm can fully explore different resource block selection possibilities while also utilizing existing learning experience to select a better resource block.

[0110] After selecting an action, the system executes it and observes the new state and the reward obtained. Based on the new state and reward, the algorithm updates the Q-value. The Q-value represents an estimate of the long-term cumulative reward for performing a certain action in a given state. As the number of iterations increases, the Q-value is continuously updated and optimized, and the algorithm gradually learns which action to choose in different states to obtain the maximum reward, i.e., finding the optimal resource block selection strategy.

[0111] After numerous iterative learning iterations, the algorithm converges to a relatively stable state. At this point, each candidate resource block corresponds to an optimized Q-value, which serves as a reference for latency optimization weights. The system evaluates and filters candidate resource blocks based on these latency optimization weights. Candidate resource blocks with higher weights indicate greater potential in meeting low latency constraints and are more likely to be selected into the target resource block set.

[0112] In this way, the Q-learning algorithm can make full use of historical data and current channel state information in complex wireless communication environments to calculate reasonable delay optimization weights for each candidate resource block, helping the system to more efficiently select target resource blocks that meet low delay requirements and improve the resource allocation performance and data transmission efficiency of the PDCCH channel.

[0113] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0114] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A PDCCH channel optimization processing system based on NR architecture, characterized by, Comprise: Channel state information acquisition module, for real-time acquisition of user equipment wireless channel state information, the wireless channel state information includes multipath time delay, doppler shift, singular value decomposition eigenvalue of channel matrix, and the wireless channel state information is marked as dynamic channel characteristic parameter; Dynamic resource allocation module, for based on the dynamic channel characteristic parameter, in the preset time-frequency resource window, through the reinforcement learning algorithm generates the allocation strategy of PDCCH candidate resource block, and filters the target resource block set satisfying low delay constraint; The target resource block set satisfying low delay constraint, its specific screening method is: From the local database, obtain the historical PDCCH resource block allocation record, extract the average delay eigenvalue of each resource block, and construct the delay characteristic matrix; Through the Q-learning algorithm, the delay optimization weight of each candidate resource block is calculated; If the time delay optimization weight of a resource block is greater than a preset threshold The resource block is added to the target resource block set.

2. The system of claim 1, wherein, The dynamic resource allocation module further comprises: Interference coordination module, for based on the spatial position information of user equipment, calculate the interference covariance matrix between adjacent PDCCH resource blocks, and generate interference suppression beam weight through singular value decomposition; The specific method for calculating the interference covariance matrix is: acquire a transmit signal vector of each resource block in the target resource block set and a neighboring cell interference signal vector ; Constructing an interference covariance matrix wherein is the noise power, is the identity matrix; denotes the conjugate transpose; is the number of resource blocks occupied by the currently served users in the target resource block set, is the number of PDCCH signal sources in the neighboring cell that cause interference; right Eigenvalue decomposition is performed, and the eigenvector corresponding to the largest eigenvalue is extracted as the interference suppression beam weight.

3. The system of claim 2, wherein, The interference coordination module further comprises: Dynamic beamforming module, for adjusting the phase and amplitude parameters of PDCCH transmitting beam according to interference suppression beam weight, and the specific adjustment method is: Obtaining an azimuth angle and an elevation angle of a target user equipment from a local database and a pitch angle , calculating an initial beam pointing vector ; Optimizing beam weights by a gradient descent algorithm such that an objective function is maximized, where is an interference suppression factor, is the conjugate transpose of is an initial beam pointing vector, computed by the user equipment azimuth and elevation angles, denotes a beam weight vector acting on a weighted contribution of the interference covariance matrix .

4. The system of claim 1, wherein, Further comprise: Multi-user detection module, for based on non-orthogonal multiple access technology, the user signal in the overlapping PDCCH resource block is carried out serial interference cancellation, and the specific method is: To receive a signal Performing linear minimum mean square error estimation to obtain an initial detection signal ; Computing residual interference wherein is a first user channel matrix; The computing residual interference step is iteratively performed until the signal of the th user satisfies , a preset error threshold.

5. The system of claim 4, wherein, The multi-user detection module further comprises: Adaptive modulation and coding module, for dynamically selecting modulation and coding strategy MCS according to channel quality indicator CQI, and the specific selection method is: Obtain the bit error rate (BER) curve corresponding to each MCS level from the local database, and fit the relationship function between CQI and BER ; The maximum MCS level satisfying is searched by dichotomy, where is the upper limit of the BER tolerated by the system.

6. The system of claim 1, wherein, Further comprise: Channel estimation optimization module, for through convolutional neural network CNN, the specific method for predicting and compensating CSI is: Construct a training data set containing historical CSI sequences, input a CNN network for feature extraction, and output a predicted CSI matrix ; Computing CSI compensation coefficients , if then trigger real-time channel re-estimation, where, denotes the Frobenius norm, which measures the difference between matrices, denotes the actual channel state information matrix, is a compensation coefficient threshold value.

7. The system of claim 6, wherein, The channel estimation optimization module further comprises: Time-frequency synchronization module, for generating synchronization signal through Zadoff-Chu sequence, and the specific method is: Generating base sequences wherein is a root index, is a sequence length; The peak position is extracted as time-frequency offset estimation value by performing cyclic correlation operation on the received signal.

8. The system of claim 1, wherein, Further comprise: Control information packaging module, for mapping downlink control information DCI to PDCCH resource element, and the specific method is: CRC attachment and polar code encoding are performed on the DCI bit stream to generate an encoded codeword ; mapping the data to a set of subcarriers of a target resource block by orthogonal frequency division multiplexing modulation to a set of subcarrier indices of a target resource block .

9. The system of claim 8, wherein, The control information packaging module further comprises: Power allocation module, for dynamically adjusting PDCCH transmitting power according to user priority, and the specific method is: Obtain user priority weight from local database, calculate power allocation ratio wherein is the power allocation ratio for the u-th user, is the total transmit power, is the priority weight for the u-th user, and U is the total number of users.

10. The system of claim 1, wherein, Further comprise: Feedback mechanism module, for receiving ACK / NACK feedback of user equipment through uplink control channel PUCCH, and the specific processing method is: The hybrid automatic repeat request process ID in the feedback information is parsed, if it is NACK, the original data packet is extracted from the local cache and the retransmission scheduling is triggered.

Images