Dual-frequency signal adaptive switching transmission method

Abstract

This invention relates to the field of signal transmission technology, specifically to a dual-frequency signal adaptive switching transmission method. The method includes the following steps: synchronously acquiring dual-frequency signals and aligning them spatiotemporally; establishing a three-level priority system; dynamically adjusting the weights of evaluation indicators based on service type; comparing dual-frequency quality evaluation values ​​in real time; issuing switching commands to synchronously adjust radio frequency parameters; buffering service data during the switching period; and retransmitting data after establishing a stable connection. The method also monitors signal quality and transmission status after switching and dynamically adjusts scheduling strategies, evaluation weights, and thresholds. This invention forms a closed-loop optimization mechanism by recording processing time, computing power utilization, and other data, and dynamically optimizing configuration parameters to continuously adapt to actual operating scenarios, further improving the real-time performance and reliability of multi-protocol compatibility.

Description

Technical Field

[0001] This invention relates to the field of signal transmission technology, and more specifically to a dual-frequency signal adaptive switching transmission method. Background Technology

[0002] In existing dual-frequency signal switching transmission technologies, adaptive switching technology encounters data processing blockage due to multiple concurrent tasks. The data analysis of adaptive switching needs to share CPU / memory with service transmission (such as video encoding and positioning calculation). Under high load, the data processing queue is blocked, which leads to delay in switching decisions and may even trigger unexpected switching. In addition, multipath effects may cause inter-symbol interference, and shadow fading may cause slow changes in signal strength, affecting switching. Summary of the Invention

[0003] The purpose of this invention is to provide a dual-frequency signal adaptive switching transmission method to solve the problems mentioned in the background art.

[0004] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0005] The dual-frequency signal adaptive switching transmission method includes the following steps:

[0006] Simultaneously acquire dual-frequency signals and align them in time and space. Use frequency domain equalization to cancel inter-symbol interference, and sliding window filtering to suppress slow signal changes. Extract core features of received power and signal-to-noise ratio.

[0007] Establish a three-level priority system, hierarchical resource scheduling for multiple tasks, and allocate fixed CPU time slices and dedicated memory.

[0008] The weights of evaluation indicators are dynamically adjusted based on the service type, the dual-frequency signal quality evaluation value is calculated, and the threshold ranges of high quality, usable, and unusable are divided.

[0009] Real-time comparison of dual-frequency quality assessment values, and generation of switching or maintenance commands based on several consecutive anti-shake judgments;

[0010] The handover command is issued to adjust the radio frequency parameters synchronously, buffer the service data during the handover period, and resend the data after a stable connection is established.

[0011] Monitor signal quality and transmission status after handover, and dynamically adjust scheduling strategies, evaluation weights, and thresholds.

[0012] Furthermore, the synchronous acquisition and spatiotemporal alignment of dual-frequency signals specifically involves using a dual-frequency radio frequency module that supports the target dual-frequency bands to acquire signals concurrently, calibrating the crystal oscillator through clock synchronization, adding a unified reference timestamp to each frame of signal, pre-measuring the difference in transmission path length between the two frequency band receiving channels and calculating the compensation value, performing time offset correction on one signal frame after acquisition, and aligning the frames according to the characteristics of the signal frame synchronization header.

[0013] Furthermore, the method of offsetting inter-symbol interference through frequency domain equalization specifically involves estimating multipath channel parameters using training sequences based on the subcarrier characteristics of the OFDM transmission architecture, calculating equalization coefficients to compensate for the amplitude and phase of each subcarrier in reverse order to offset inter-carrier interference, and detecting and dynamically adjusting the equalization coefficients in real time to control phase deviation and correct waveform distortion.

[0014] Furthermore, the sliding window filtering suppresses slow signal changes. Specifically, it employs sliding window mean filtering, using a first-in-first-out strategy to update data and calculate the mean of data within the window in real time as the filtered intensity value. It dynamically monitors the rate of change of the original signal to avoid over-filtering, suppresses slow changes caused by shadow fading, and at the same time preserves the rapid fluctuation characteristics of the signal caused by changes in the actual channel state.

[0015] Furthermore, the extraction of core features of received power and signal-to-noise ratio specifically involves extracting core features of received power, signal-to-noise ratio, signal-to-interference-plus-noise ratio, and bit error rate. The extraction process does not occupy service transmission resources and ensures real-time performance. It relies on dedicated memory partitions to cache data in parallel and uses hardware acceleration to shorten processing time. After unifying the feature data format, it is written to the feature cache area of ​​dedicated memory in real time.

[0016] Furthermore, the establishment of a three-level priority system and multi-task hierarchical resource scheduling, allocating fixed CPU time slices and dedicated memory, specifically involves defining three levels of task priorities: the first priority is core-related tasks for dual-frequency switching, the second is core business transmission tasks, and the third is non-real-time auxiliary tasks. A hybrid scheduling mechanism is adopted, allocating fixed CPU time slices to the first priority tasks and allowing preemption of lower-priority resources. An independent dedicated memory partition is created to isolate the tasks from other priorities. CPU and memory usage are monitored in real time, and lower-priority tasks are temporarily suspended to ensure the resource needs of the first-priority tasks are met.

[0017] Furthermore, the method of dynamically adjusting the weights of evaluation indicators based on service type, calculating the dual-frequency signal quality evaluation value, and dividing the threshold ranges of high quality, usable, and unusable involves: identifying the service type in real time, configuring the preset weights according to the service call, reading feature data from dedicated memory to calculate the dual-frequency signal quality evaluation value in parallel, dividing the threshold ranges of high quality, usable, and unusable, storing the thresholds in a dynamically rewritable configuration file, linking with the feedback optimization mechanism, and calibrating the thresholds according to the service transmission status and channel conditions after the switchover to adapt to actual service requirements and channel conditions.

[0018] Furthermore, the real-time comparison of dual-frequency quality assessment values, combined with several consecutive anti-jitter judgments, generates a switching or maintenance instruction. Specifically, it relies on dedicated memory to read dual-frequency quality assessment values ​​in real time, compares them in layers according to high-quality, usable, and unusable intervals, maintains transmission and continuously monitors the high-quality interval, initiates switching preparation based on the difference between dual-frequency assessment values ​​in the usable interval, and immediately triggers switching in the unusable interval. Anti-jitter judgment is introduced, and the final instruction is generated only when the switching conditions are met continuously. It relies on the first priority dedicated resources for parallel operation.

[0019] Furthermore, the switching command is issued to simultaneously adjust radio frequency parameters, cache service data during the switching period, and resend data after a stable connection is established. Specifically, the switching command is issued to simultaneously adjust radio frequency parameters, cache core service data during the switching period, and resend data after a stable connection is established. The command is issued through the high-speed control bus, the radio frequency module adjusts and verifies parameters in parallel, a dedicated buffer is enabled to store data without occupying the first priority resource, the backup frequency band connection is detected after the switch, and the switching transmission is completed and resend data in sequence after the standard is met, thereby controlling the total time consumption and service transmission interruption time.

[0020] Furthermore, the monitoring of signal quality and transmission status after the handover, and the dynamic adjustment of scheduling strategies, evaluation weights and thresholds, specifically involves continuously monitoring the signal quality and service transmission status of the new frequency band after the handover, collecting and analyzing data, adjusting evaluation weights and thresholds according to service type when signal or transmission is abnormal, optimizing scheduling strategies when resource scheduling is not up to standard, calibrating anti-jitter parameters after accumulating the number of handovers, and writing the adjusted parameters into the configuration file to form a closed-loop optimization.

[0021] Compared with existing technologies, this invention has the following advantages: The security instructions in this application are set to the highest priority and preemptively executed. Upon detection, normal data processing is immediately paused, and the output is directly parsed, preventing emergency protection instructions from being blocked and ensuring that protection functions are triggered immediately in emergency situations. Real-time monitoring of MCU computing power is achieved; when not saturated, full parallel parsing improves efficiency; when saturated, fixed computing power is allocated to the main control and backup protocols, and the parsing frequency of the backup protocols is adjusted, avoiding increased processing delays caused by computing power saturation. The sliding window size is dynamically adjusted based on the data change rate; in high-dynamic scenarios, a 1-frame window is used for direct output, while in low-dynamic scenarios, 3 frames are maintained for smooth processing, solving the decision-making lag problem under high dynamic conditions and ensuring data stability in normal scenarios. By recording processing time, computing power utilization, and other data and dynamically optimizing configuration parameters, a closed-loop optimization mechanism is formed, continuously adapting to actual operating scenarios and further improving the real-time performance and reliability of multi-protocol compatibility. Attached Figure Description

[0022] Figure 1 This is a flowchart of the dual-frequency signal adaptive switching transmission method of the present invention. Detailed Implementation

[0023] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments and 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.

[0024] This application discloses a dual-frequency signal adaptive switching transmission method, such as... Figure 1 As shown, the steps include:

[0025] Simultaneously acquire dual-frequency signals and align them in time and space. Use frequency domain equalization to cancel inter-symbol interference, and sliding window filtering to suppress slow signal changes. Extract core features of received power and signal-to-noise ratio.

[0026] Establish a three-level priority system, hierarchical resource scheduling for multiple tasks, and allocate fixed CPU time slices and dedicated memory.

[0027] The weights of evaluation indicators are dynamically adjusted based on the service type, the dual-frequency signal quality evaluation value is calculated, and the threshold ranges of high quality, usable, and unusable are divided.

[0028] Real-time comparison of dual-frequency quality assessment values, combined with several consecutive anti-shake judgments, generates switching or maintenance commands;

[0029] The handover command is issued to adjust the radio frequency parameters synchronously, buffer the service data during the handover period, and resend the data after a stable connection is established.

[0030] Monitor signal quality and transmission status after handover, and dynamically adjust scheduling strategies, evaluation weights, and thresholds.

[0031] The synchronous acquisition and spatiotemporal alignment of dual-frequency signals is specifically implemented by using a dual-frequency radio frequency module that supports concurrent acquisition of target dual frequency bands (such as Sub-6GHz and millimeter wave), calibrating the internal crystal oscillator of the module through a high-precision clock synchronization circuit to ensure that the acquisition trigger command for dual-frequency signals is generated synchronously, and adding a unified reference system timestamp to each frame of signal during the acquisition process.

[0032] For example, the timestamp adopts a 64-bit time format with a precision of 0.6μs, which meets the timing requirement of ≤1μs. At the same time, the difference in signal transmission path length between the two frequency band receiving channels in the dual-frequency RF module is measured in advance, and the corresponding transmission delay compensation value is calculated. After data acquisition, the signal frame of one frequency band is time offset corrected according to the compensation value. Then, frame alignment matching is performed according to the synchronization header characteristics of the signal frame to achieve accurate alignment of the dual-frequency signals in the time and spatial dimensions, ensuring that subsequent joint preprocessing operations such as frequency domain equalization and filtering can be carried out based on the same spatiotemporal reference signal data.

[0033] The method of canceling inter-symbol interference through frequency domain equalization is specifically implemented as follows: Based on the same spatiotemporal reference data after spatiotemporal alignment of the dual-frequency signals, and considering the subcarrier characteristics of the Orthogonal Frequency Division Multiplexing (OFDM) transmission architecture, the multipath channels of each dual-frequency signal are first accurately estimated using a preset training sequence to obtain the channel frequency response parameters corresponding to each subcarrier, including the amplitude attenuation and phase offset caused by multipath propagation. Then, according to the minimum mean square error criterion, a dedicated equalization coefficient is calculated for each subcarrier, and the signal amplitude and phase of each subcarrier are reverse-compensated by an equalizer to cancel the mutual interference between subcarriers caused by multipath propagation. During the equalization process, the phase deviation value of each subcarrier is detected in real time, and the equalization coefficient is adjusted through dynamic feedback to ensure that the phase deviation between subcarriers is controlled within 0.5° after the dual-frequency signal is equalized. At the same time, the signal waveform distortion caused by multipath effect is corrected, so that the processed dual-frequency signal maintains clear symbol boundaries, thereby eliminating the impact of inter-symbol interference on subsequent feature extraction and quality assessment from the root.

[0034] The sliding window filtering suppresses slow signal changes. Specifically, based on the original signal strength data after spatiotemporal alignment and frequency domain equalization of the dual-frequency signals, a sliding window mean filtering mechanism with a fixed length of 8 signal frames is used. The duration of each signal frame is consistent with the dual-frequency signal acquisition cycle, ensuring that the time dimension covered by the window matches the signal feature extraction rhythm. The filtering process adopts a first-in, first-out (FIFO) data update strategy. For each newly acquired frame of signal strength data, the earliest frame stored in the window is removed, and the arithmetic mean of the 8 frames within the window is calculated in real time. This mean is used as the filtered signal strength value at the current moment. To ensure the real-time performance of the preprocessing link, the execution time of the filtering algorithm is controlled within 0.2ms. Furthermore, by dynamically monitoring the rate of change of the original signal, over-filtering is avoided to prevent the true trend of signal change from being masked. After filtering, the fluctuation range of the signal strength is verified in real time to ensure that any consecutive 10... The intensity fluctuation of each signal frame does not exceed ±3%, effectively suppressing the slow changes caused by shadow fading, while retaining the rapid fluctuation characteristics of the signal caused by changes in the actual channel state, providing a stable and real data source for the accurate extraction of core features such as received power (RP).

[0035] The extraction of core features for received power and signal-to-noise ratio (SNR) is specifically implemented by performing parallel core feature extraction operations based on high-quality data after spatiotemporal alignment, frequency domain equalization, and sliding window filtering of dual-frequency signals. This ensures that the extraction process does not occupy service transmission resources and meets real-time requirements. When extracting received power (RP), the effective signal amplitude after amplification and demodulation of the dual-frequency signal by the RF module is detected, its squared mean is calculated, and calibration is performed according to the module's preset receive gain and attenuation coefficient to obtain the actual received power value. The unit is uniformly dBm, and the measurement accuracy is controlled within ±0.5 dBm. When extracting signal-to-noise ratio (SNR), the pilot subcarrier signal in the OFDM transmission architecture is used to separate the useful signal power at the pilot position from the background noise power of the non-pilot blank subcarrier. The SNR is calculated by the ratio of the two, and a 16-point moving average method is used to smooth the calculation result to avoid instantaneous noise interference. When extracting signal-to-interference-plus-noise ratio (SINR), the SNR is calculated based on the SNR... Based on computation, spectrum analysis is used to identify the frequency bands and power of co-channel and adjacent-channel interference signals. The interference power is subtracted from the useful signal power before being compared with the noise power to ensure the accuracy of features in interference scenarios. When extracting the bit error rate (BER), the ratio of the number of erroneous symbols to the total number of symbols in the received data is calculated based on the check bits in the service data frames or a preset training sequence. A segmented statistical strategy is adopted, with BER updated every four signal frames to balance real-time performance and statistical accuracy. The entire extraction process relies on cached data in a dedicated memory partition for parallel computation, leveraging hardware acceleration logic to shorten processing time. Simultaneously, the extracted RP, SNR, SINR, and BER feature data are uniformly formatted as 32-bit floating-point numbers and written to the dedicated memory feature cache in real time for direct reading by the subsequent quality assessment module. This ensures that the total latency from data input to feature output is strictly controlled within 1ms, guaranteeing the timeliness and availability of the feature data.

[0036] The proposed system establishes a three-tier priority system, hierarchical resource scheduling for multiple tasks, and allocates fixed CPU time slices and dedicated memory. In practice, this involves first clarifying the specific criteria for the three-tier task priorities. The first priority is limited to tasks directly related to the core switching logic, such as dual-frequency signal feature extraction and switching decision calculation. The second priority is for core business transmission tasks that require real-time assurance, such as video encoding and positioning calculation. The third priority is for auxiliary tasks that do not have strict timeliness requirements, such as log recording, system status statistics, and non-real-time data backup.

[0037] A hybrid scheduling mechanism of "Earliest Deadline First (EDF) and time-slice round-robin" is adopted. The system scheduling cycle is set to 10ms. A fixed CPU time slice is allocated to the first priority task, accounting for no less than 40% (i.e., at least 4ms in a single scheduling cycle). The remaining time slices are allocated among the second and third priority tasks in a round-robin manner. When a new first priority task is triggered, it can directly preempt the resources of the currently executing second and third priority tasks. During the preemption process, the execution status and data of the preempted task are temporarily stored through a context saving mechanism to ensure that there is no data loss or process interruption when the execution is resumed.

[0038] A dedicated physical memory partition, with a capacity of no less than 8MB, is allocated to the dual-frequency signal processing and switching decision modules through a memory management unit. This memory partition is only open to read and write permissions for the first-priority task and is completely isolated from the memory space of the second and third-priority tasks to avoid memory access conflicts and resource contention between different modules. A real-time monitoring mechanism for CPU and memory usage is established every 2ms. When the memory usage of the first-priority task reaches 90% or above, or the CPU usage reaches 80% or above, the third-priority task is temporarily suspended to release its CPU and memory resources. If the resources are still insufficient to meet the needs of the first-priority task, non-critical auxiliary processes in the second-priority task (such as non-real-time frame rate optimization for video encoding and historical data backtracking analysis for positioning calculation) are further temporarily suspended. The released resources are immediately allocated to the first-priority task until its resource usage drops below the threshold, ensuring that dual-frequency switching-related data processing always receives priority resource guarantees.

[0039] The method involves dynamically adjusting the weights of evaluation indicators based on service type, calculating the dual-frequency signal quality evaluation value, and dividing the threshold ranges into high-quality, usable, and unusable categories. Specifically, this involves first establishing a real-time service type identification mechanism. This mechanism dynamically determines the currently running service type using the status identifier bits of the service transmission module (such as video transmission identifier bits and location service identifier bits), with an identification response delay ≤0.5ms, ensuring that the weight adjustment matches the service scenario in real time. Based on the identification results, a preset weight configuration library is called. This configuration library is stored in a parameter partition of dedicated memory, clearly defining the indicator weight allocation for different services. For example, in video transmission services, the SNR weight is 40%, BER weight is 30%, SINR weight is 20%, and RP weight is 10%; in location services, the SINR weight is 40%, RP weight is 30%, SNR weight is 20%, and BER weight is 10%; and general services use a balanced weight (25% for each indicator). The total weight is fixed at 100%, and the configuration can be updated through a feedback optimization mechanism.

[0040] When calculating the quality assessment value, the real-time RP, SNR, SINR, and BER data extracted in the previous steps are read from the feature cache in dedicated memory (data read latency ≤ 0.3ms), and calculated according to the preset formula (RP × ... + SNR× + SINR× + (1-BER)× The calculation is performed by multiplying the result by 100. The calculation process adopts parallel computing logic, relying on the CPU time slice of the first priority task to ensure efficiency. The BER value is constrained to a range of 0-1 to avoid abnormal interference. The calculation result retains integer bits, and the value is strictly limited to 0-100 points. The overall calculation delay is ≤2ms. When dividing the threshold interval, the initial settings are: high-quality interval ≥80 points, usable interval 60-79 points, and unusable interval <60 points. The threshold parameters are stored in a dynamically rewritable configuration file. The calibration logic is linked with the feedback mechanism of other steps. After every 50 handovers, the average transmission delay, packet loss rate and signal stability of the service after the handover are statistically analyzed. If the packet loss rate of the service in the high-quality interval is >0.03% or the delay is >15ms, the lower limit of the high-quality interval is increased by 3 points. If the proportion of unexpected handovers in the usable interval is >10%, the lower limit of the usable interval is increased by 2 points. Each threshold adjustment does not exceed 5 points to ensure that the interval division is accurately adapted to the actual channel status and service requirements.

[0041] In implementation, a fixed-weight linear combination is used to calculate the evaluation value, without considering the correlation between indicators and the real-time dynamic requirements of service quality, such as frame rate fluctuations in video services and instantaneous accuracy requirements in positioning services. Therefore, in a further embodiment, a dynamic weighted evaluation method based on the mutual information between service quality requirements and indicators is introduced, including the following steps:

[0042] The preprocessed real-time signal metrics (RP: received power (dBm), SNR: signal-to-noise ratio (dB), SINR: signal-to-interference-plus-noise ratio (dB), BER: bit error rate) are normalized to eliminate the influence of dimensions, resulting in normalized metrics x∈[0,1]. The normalization formula for positive metrics (RP, SNR, SINR, the larger the value, the better) is as follows: ; where y i The original index value, y i,min This is the theoretical minimum value of the index (e.g., RPmin = -120dBm), y i,max The theoretical maximum value (e.g., RP) max =-30dBm).

[0043] The normalization formula for the negative indicator (BER, the smaller the better) is: Where y4 is the original BER value, y 4,min =10-8 (Target bit error rate) ， y 4,max =10 -3 (Maximum bit error rate).

[0044] The output result is a normalized index vector X = [x1 (RP), x2 (SNR), x3 (SINR), x4 (BER)].

[0045] Input the normalized indicator vector X, the business requirement feature vector Q, and the business type T. First, calculate the mutual information weights of the indicators. The mutual information I(x) between any two metrics is calculated using a sliding window (window length = 10 frames). i , x j This measures the correlation between indicators; the lower the correlation, the higher the weight of the indicator. The calculation formula is:

[0046] ;

[0047] Where I(x) i , x j ) is x i With x j The mutual information (unit: bit) reflects the degree of dependence between the two indicators, I∈ [0,1].

[0048] Introducing the business demand coefficient α i (Bound to service type T), the weight is adjusted based on QoS feature Q, specifically:

[0049] Video service (α1=0.1, α2=0.4, α3=0.2, α4=0.3): ;

[0050] Location services (α1=0.3, α2=0.2, α3=0.4, α4=0.1): ;

[0051] General business (α1=α2=α3=α4=0.25): ;

[0052] Finally, regarding w i Normalization ensures ∑w i =1.

[0053] Considering the diminishing marginal utility of indicators (e.g., excessively high SNR has limited impact on business improvement), the weights are adjusted using an exponential function. The adjustment formula is as follows: ;

[0054] Output dynamically adjusted weight vector .

[0055] Then calculate the signal quality evaluation value. Specifically,

[0056] Input the normalized index vector X and the dynamic correction weight vector , and adopt the weighted geometric mean fusion index to avoid the "averaging" defect of linear fusion and highlight the influence of key indicators. The fusion formula is: ;

[0057] where S is the quality evaluation value (∈[0, 100]), keep the integer part, and the calculation delay ≤ 2 ms.

[0058] Then, based on the dynamic division of the threshold interval under the business QoS constraint. Specifically,

[0059] Input the quality evaluation value S, the service type T, and the service historical QoS data (transmission delay D, packet loss rate L, positioning error E, etc. of the recent 50 handovers).

[0060] The initial thresholds are set as the high-quality interval S≥S1, the available interval S2≤S<S1, and the unavailable interval S<S2. The initial values are S1 = 80 and S2 = 60. Every time 30 handovers are accumulated, adjust the thresholds in combination with the business QoS constraint. For video services, if D>D0 (D0 = 15 ms) or L>L0 (L0 = 0.03%) in the high-quality interval, then S1 = S1 + 2×(D / D0 - 1) (the upper limit ≤ 95); if the frame rate fluctuation ΔF>10% in the available interval, then S2 = S2 + 1×(ΔF / 10% - 1) (the upper limit ≤ 75). For positioning services, if E>E0 (E0 = 1 m) in the high-quality interval, then S1 = S1 + 3×(E / E0 - 1) (the upper limit ≤ 95); if the error bounce rate after handover>15% in the available interval, then S3 = S2 + 2×(bounce rate / 15% - 1) (the upper limit ≤ 75). The adjustment amplitude each time ≤ 5 to ensure stability.

[0061] Finally, output the dynamically updated threshold intervals (high-quality, available, unavailable).

[0062] Compare the real-time dual-frequency quality evaluation values, and generate a switching or maintaining instruction based on continuous anti-shake judgments for several times; specifically in implementation, rely on the feature buffer area of the dedicated memory to read the latest quality evaluation values of the current working frequency band and the standby frequency band in real time. The data reading delay ≤ 0.3 ms to ensure that the comparison is based on real-time data with the same time reference and avoid decision-making deviations caused by data asynchrony.

[0063] The comparison process is executed in layers according to preset logic. For example, if the quality assessment value of the current working frequency band is in the excellent range (≥80 points), an instruction to maintain the current transmission state is directly generated. At the same time, a quality assessment value rereading and secondary comparison are triggered every 50ms to maintain continuous monitoring of the signal status. If the current working frequency band is in the usable range (60-79 points), the difference between the assessment values ​​of the backup frequency band and the current frequency band is calculated. When the difference is ≥10 points, a handover preparation instruction is generated to start parameter preloading preparation before the frequency band handover. If the current working frequency band is in the unusable range (<60 points), a handover execution trigger instruction is generated immediately without waiting for the difference judgment.

[0064] To suppress unexpected handovers caused by instantaneous signal fluctuations, a jitter-reduction mechanism is introduced. Each time the handover condition is met (the difference between the available intervals meets the standard or the unavailable interval is reached), continuous counting begins. The dual-frequency quality assessment value is repeatedly read and the handover condition is verified at fixed 1ms intervals. Only when three consecutive detections meet the corresponding handover condition is the initial command upgraded to the final effective handover command. If any detection fails to meet the condition, the count is immediately reset and accumulation restarts. The entire comparison and judgment process relies on the dedicated CPU time slice of the first priority task for parallel computation, avoiding delays caused by resource contention. The comparison logic operation time is ≤2ms. Based on the data reading delay, the total delay from the start of comparison to the generation of the final command is strictly controlled within 3ms, meeting the timing requirement of a total handover decision link delay ≤7ms. Simultaneously, the count reset mechanism ensures the rigor of the jitter-reduction logic, avoiding instantaneous interference without affecting the rapid response to real channel degradation.

[0065] The handover command is issued synchronously to adjust radio frequency parameters, caches service data during the handover period, and retransmits data after a stable connection is established. Specifically, the final handover command generated by the handover decision module is sent to the dual-band radio frequency module via the system's internal high-speed control bus, with a command transmission delay of ≤0.3ms, ensuring rapid response to the handover action. The dual-band radio frequency module has a built-in parameter configuration unit that, upon receiving the command, performs multi-dimensional radio frequency parameter adjustments in parallel, synchronously switching the carrier frequency (e.g., switching from the Sub-6GHz band to the millimeter-wave band), modulation and coding scheme (dynamically selecting 64-QAM, 16-QAM, or QPSK based on signal quality), and radio frequency transmit power. The parameter writing process employs a hardware-level synchronization mechanism to avoid transient signal distortion caused by asynchronous parameter adjustments. After adjustment, the module's built-in verification circuit confirms the parameter validity, ensuring correct modulation and coding scheme switching, and radio frequency power deviation controlled within ±1dBm. The entire parameter handover time is strictly controlled within 2ms.

[0066] Simultaneously with the switch command, a dedicated buffer area activation mechanism is automatically triggered. This buffer area is an independent memory partition defined in the previous steps, with a capacity of ≥16MB. It is specifically used to cache the data to be transmitted for second-priority core services (video encoding, positioning calculation, etc.). It adopts a first-in-first-out circular buffer structure, storing data sequentially according to the data generation order. The buffer write latency is ≤0.2ms, and it does not occupy the memory resources of the first-priority tasks. The maximum buffer duration is set to 5ms to avoid data overflow. After the dual-band RF module switches to the backup frequency band, it collects the SINR and BER data of this frequency band in real time, continuously detecting 3 times at 0.5ms intervals. If the results of the 3 detections all meet the requirements of SINR≥60 and BER≤... The system immediately cuts off the transmission link of the current frequency band and simultaneously starts service transmission on the backup frequency band by using a control signal to ensure a stable connection.

[0067] The cached data retransmission process is then initiated, reading cached data from the circular buffer in timestamp order. The retransmission rate is consistent with the current service transmission rate (e.g., 10Mbps, 20Mbps, etc.). During the retransmission process, the transmission status is monitored in real time through a data frame sequence number verification mechanism. If packet loss is detected, the corresponding data packet is immediately retransmitted to ensure that no cached data is missed. The entire handover execution process (parameter adjustment, caching, connection detection, and data retransmission) takes ≤8ms, with the service transmission interruption time controlled within 1ms. After retransmission is completed, the caching mechanism is automatically closed, releasing cache resources to ensure that subsequent service transmissions can normally occupy system resources.

[0068] The monitoring of signal quality and transmission status after the switchover involves dynamically adjusting scheduling strategies, evaluation weights, and thresholds. Specifically, after the switchover is completed, a continuous monitoring cycle of 100ms is initiated immediately. Signal quality indicators and service transmission status data of the new operating frequency band are collected synchronously at a frequency of 1ms / time. Signal quality indicators include BER, SINR, RP, and SNR, while service transmission status data covers transmission delay and data packet loss rate. All monitoring data is written to a dedicated memory feedback buffer in real time to ensure that no data is missed and the read latency is ≤0.2ms.

[0069] After the monitoring period ends, the system automatically performs statistical analysis on the data. For example, if the statistical results show BER > If the transmission latency is greater than 20ms or the data packet loss rate is greater than 0.05%, the evaluation weight and threshold range will be adjusted. The weight of key indicators will be increased in a targeted manner according to the service type. For example, the SNR weight will be increased by 10% for video transmission services and the SINR weight will be increased by 10% for positioning services. At the same time, the threshold range will be updated. The lower limit of the high-quality range can be increased by 3-5 points and the lower limit of the usable range can be increased by 2-5 points. Each adjustment will not exceed 5 points to avoid excessive fluctuations in the range that may affect stability. If the execution latency of the first priority task (dual-frequency signal feature analysis, handover decision calculation) is greater than 3ms or the memory usage is frequently close to 90% in a multi-task concurrent scenario, the scheduling strategy will be adjusted. The CPU time slice ratio of the first priority task will be increased from ≥40% to 50%, or the dedicated physical memory capacity will be expanded from ≥8MB to 10MB. At the same time, the suspension triggering conditions of low priority tasks will be optimized, and the CPU usage warning threshold of the first priority task will be lowered from 80% to 75% to release resources in advance. In addition, for every 100 handover operations, the system will calculate the overall handover success rate and the proportion of unexpected handovers. If the proportion of unexpected handovers is greater than 5%, the anti-shake judgment parameters will be further calibrated, adjusting the number of consecutive detections from 3 to 4, or increasing the handover difference threshold from 10 points to 12 points. All adjusted scheduling strategy parameters, evaluation weight coefficients, and threshold range values ​​are written to the dynamically rewritable configuration file in real time through atomic operations, overwriting the original parameters and generating version records. This ensures that the optimized parameters are directly called when the handover process is started next time, forming a closed-loop optimization mechanism of "monitoring, analysis, adjustment, and effectiveness". This continuously adapts to channel changes and service requirements, improving the stability and handover accuracy of long-term operation.

[0070] The above description is merely one specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for adaptive switching transmission of dual frequency signals, characterized in that, Including the following steps: Simultaneously acquire dual-frequency signals and align them in time and space. Use frequency domain equalization to cancel inter-symbol interference, and sliding window filtering to suppress slow signal changes. Extract core features of received power and signal-to-noise ratio. Establish a three-level priority system, hierarchical resource scheduling for multiple tasks, and allocate fixed CPU time slices and dedicated memory. The weights of evaluation indicators are dynamically adjusted based on the service type, the dual-frequency signal quality evaluation value is calculated, and the threshold ranges of high quality, usable, and unusable are divided. Real-time comparison of dual-frequency quality assessment values, combined with several consecutive anti-shake judgments, generates switching or maintenance commands; The handover command is issued to adjust the radio frequency parameters synchronously, buffer the service data during the handover period, and resend the data after a stable connection is established. Monitor signal quality and transmission status after handover, and dynamically adjust scheduling strategies, evaluation weights, and thresholds; The establishment of a three-level priority system and multi-task hierarchical resource scheduling, allocating fixed CPU time slices and dedicated memory, specifically involves defining three levels of task priorities: the first priority is core-related tasks for dual-frequency switching, the second is core business transmission tasks, and the third is non-real-time auxiliary tasks. A hybrid scheduling mechanism is adopted, allocating fixed CPU time slices to the first priority tasks and allowing preemption of lower-priority resources. An independent dedicated memory partition is created to isolate the tasks from other priorities. CPU and memory usage are monitored in real time, and lower-priority tasks are temporarily suspended to ensure the resource needs of the first-priority tasks are met. The process of dynamically adjusting the weights of evaluation indicators based on service type, calculating dual-frequency signal quality evaluation values, and dividing threshold ranges into high-quality, usable, and unusable categories involves: real-time identification of service types, configuration of preset weights according to service calls, reading feature data from dedicated memory to calculate dual-frequency signal quality evaluation values ​​in parallel, dividing threshold ranges into high-quality, usable, and unusable categories, storing the thresholds in a dynamically rewritable configuration file, linking with a feedback optimization mechanism, and calibrating the thresholds based on the service transmission status and channel conditions after switching to adapt to actual service requirements and channel conditions. The real-time comparison of dual-frequency quality assessment values, combined with several consecutive anti-jitter judgments, generates a switching or maintenance command. Specifically, it relies on dedicated memory to read dual-frequency quality assessment values ​​in real time, compares them in layers according to high-quality, usable, and unusable intervals, maintains transmission and continuously monitors the high-quality interval, initiates switching preparation based on the difference between the dual-frequency assessment values ​​in the usable interval, and immediately triggers switching in the unusable interval. Anti-jitter judgment is introduced, and the final command is generated only when the switching conditions are met continuously. It relies on the first priority dedicated resources for parallel operation.

2. The dual-frequency signal adaptive switching transmission method according to claim 1, characterized in that, The synchronous acquisition of dual-frequency signals and spatiotemporal alignment specifically involves using a dual-frequency radio frequency module that supports the target dual-frequency band to acquire signals concurrently, calibrating the crystal oscillator through clock synchronization, and adding a unified reference timestamp to each frame of signal. The difference in transmission path length between the two frequency band receiving channels is measured in advance and the compensation value is calculated. After acquisition, time offset correction is performed on one signal frame, and frame alignment is performed based on the synchronization header characteristics of the signal frame.

3. The dual-frequency signal adaptive switching transmission method according to claim 1, characterized in that, The method of offsetting inter-symbol interference through frequency domain equalization specifically involves estimating multipath channel parameters through training sequences based on the subcarrier characteristics of the OFDM transmission architecture, calculating equalization coefficients to compensate for the amplitude and phase of each subcarrier in reverse order to offset inter-carrier interference, detecting and dynamically adjusting the equalization coefficients in real time, controlling phase deviation, and correcting waveform distortion.

4. The dual-frequency signal adaptive switching transmission method according to claim 1, characterized in that, The sliding window filtering suppresses slow changes in the signal. Specifically, it employs sliding window mean filtering, updates data using a first-in-first-out strategy, calculates the mean of the data within the window in real time as the filtered intensity value, dynamically monitors the rate of change of the original signal to avoid over-filtering, suppresses slow changes caused by shadow fading, and at the same time preserves the rapid fluctuation characteristics of the signal caused by changes in the actual channel state.

5. The dual-frequency signal adaptive switching transmission method according to claim 1, characterized in that, The extraction of core features of received power and signal-to-noise ratio specifically involves extracting core features of received power, signal-to-noise ratio, signal-to-interference-plus-noise ratio, and bit error rate. The extraction process does not occupy service transmission resources and ensures real-time performance. It relies on dedicated memory partitions to cache data in parallel and uses hardware acceleration to shorten processing time. After unifying the feature data format, it is written to the feature cache area of ​​dedicated memory in real time.

6. The dual-frequency signal adaptive switching transmission method according to claim 1, characterized in that, The process involves issuing a handover command to simultaneously adjust radio frequency parameters, caching service data during the handover period, and retransmitting data after a stable connection is established. Specifically, the handover command is issued to simultaneously adjust radio frequency parameters, cache core service data during the handover period, and retransmit data after a stable connection is established. Commands are issued through a high-speed control bus, the radio frequency module adjusts and verifies parameters in parallel, a dedicated buffer is enabled to store data without occupying first-priority resources, the connection of the backup frequency band is detected after the handover, and the handover transmission is switched and retransmitted in sequence after the standard is met, controlling the total time consumption and service transmission interruption time.

7. The dual-frequency signal adaptive switching transmission method according to claim 1, characterized in that, The monitoring of signal quality and transmission status after the handover, and the dynamic adjustment of scheduling strategies, evaluation weights and thresholds, specifically involves continuously monitoring the signal quality and service transmission status of the new frequency band after the handover, collecting and analyzing data, adjusting evaluation weights and thresholds according to service type when signal or transmission is abnormal, optimizing scheduling strategies when resource scheduling is not up to standard, calibrating anti-jitter parameters after accumulating the number of handovers, and writing the adjusted parameters into the configuration file to form a closed-loop optimization.

Images