A wireless network signal channel intelligent switching method and system

Abstract

This invention discloses a method and system for intelligent switching of wireless network signal channels, relating to the field of wireless communication technology. The method involves sensing at least one other channel besides the current working channel while maintaining communication on the current working channel, identifying and marking at least one inferior channel. This inferior channel meets preset low-quality conditions and is different from the current working channel. This invention does not rely on external positioning information such as GPS / BeiDou; it can complete channel sensing and coding detection using only communication gaps. It can still function normally in tunnel scenarios, fundamentally solving the switching problem when trains pass through tunnels. Furthermore, by using a timing coupling mechanism that first eliminates inferior channels and then performs binary coding detection, the globally optimal channel can be deterministically locked. Theoretical analysis and simulation verification show that the switching success rate can reach over 99.5%, and the switching delay is controlled within 30ms, fully meeting the stringent requirements of 5G-R systems for high-reliability services such as automatic train operation.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and specifically to a method and system for intelligent switching of wireless network signal channels. Background Technology

[0002] Intelligent channel handover of wireless network signals is mainly used to ensure continuous and stable connectivity of mobile communication terminals in high-speed mobile scenarios. It is a crucial component of 5G-R railway mobile communication systems supporting key services such as automatic train operation (ATO) and intelligent dispatching. With the deepening research on wireless handover in high-speed railway scenarios, the research trend of channel handover has gradually shifted from passive response algorithms based on signal strength triggering to active cognitive handover algorithms based on location prediction.

[0003] While current location-prediction-based handover algorithms can use GPS / BeiDou positioning data and along-route channel models to plan handover timing in advance, achieving a reduction in handover interruption time to about 7ms in open road sections, the algorithm faces a fundamental obstacle in tunnel scenarios: GNSS signals are completely blocked inside tunnels.

[0004] Specifically, electromagnetic waves experience severe attenuation when passing through tunnel lining and rock strata. A few meters after the train enters the tunnel entrance, the carrier-to-noise ratio (C / N0) of the received GNSS signal drops sharply from over 50 dB-Hz in open environments to below 20 dB-Hz, far below the 35 dB-Hz threshold required for receiver positioning. At this point, the receiver cannot extract valid time and ephemeris information from the navigation message, and pseudorange measurement errors exceed the order of hundreds of meters, causing the positioning module to completely lose lock. Without positioning input, the position prediction algorithm cannot determine the train's current location, cannot determine the relative distance to the next base station, and cannot calculate the handover trigger window; the entire handover decision chain is broken at the first stage. Although engineering practice has attempted to introduce inertial navigation systems (INS) or wheel speed odometers for dead reckoning to compensate for GNSS interruptions, such solutions suffer from cumulative error problems. In tunnels tens of kilometers long, dead reckoning errors without external correction will accumulate at a rate of about 0.1% to 1% of the mileage. Traveling 10 kilometers can result in a positioning error of 10 to 100 meters, while the positioning accuracy required for high-speed rail switching is within the meter level.

[0005] Therefore, existing location prediction schemes struggle to restore handover decision-making capabilities within tunnels, creating a long-standing unresolved "tunnel handover blind spot" problem. When a train travels through a tunnel at 350 km / h, this blind spot directly leads to a significant decrease in handover success rate, failing to meet the stringent requirements of 5G-R systems for a handover success rate of no less than 99.5% and a handover latency of no more than 30ms. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for intelligent switching of wireless network signal channels to address the shortcomings in the prior art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for intelligent switching of wireless network signal channels, comprising: S1. While maintaining communication on the current working channel, sense at least one other channel besides the current working channel, identify and mark at least one inferior channel, wherein the inferior channel meets a preset low-quality condition and is different from the current working channel. S2. For the remaining channel set after excluding the inferior channels, perform bit-by-bit probing according to the binary encoding rules. Each bit of probing corresponds to measuring the aggregated channel quality of a group of channels. The aggregated channel quality refers to the comprehensive statistical value of the quality indicators of each channel in the group. The target channel number is obtained by decoding based on the results of each probing. S3. Determine whether the target channel corresponding to the target channel number is better than the current working channel. If so, perform channel switching; otherwise, maintain the current channel.

[0008] In a preferred embodiment, the preset low-quality condition includes at least one of the following: Interference intensity is higher than the first threshold, signal-to-noise ratio is lower than the second threshold, received signal strength is lower than the third threshold, channel utilization is higher than the fourth threshold, packet loss rate is higher than the fifth threshold, or it is marked as a prohibited channel based on prior information; Furthermore, when quality data from multiple other channels are detected, the channel with the worst quality is marked as the inferior channel, and the inferior channel is unique.

[0009] In a preferred embodiment, the sensing in S1 is performed in any of the following ways: During communication gaps without interrupting current communication, perform timing scans of other channels or use independent radio frequency links for parallel monitoring.

[0010] In a preferred embodiment, the bit-by-bit probing in S2 includes: Let the total number of available channels be N, and the number of channels in the remaining channel set be M = N-1; assign an L-bit binary code to each of the M channels, where L is the smallest integer satisfying 2^L ≥ M; For each bit, the channels encoded as 1 form the first subset, and the channels encoded as 0 form the second subset. The aggregated channel quality of the two subsets is then measured. Select the coded bit value corresponding to the subset with better aggregation channel quality as the decision result for that bit; The L-bit decision results are combined and decoded to obtain the target channel number.

[0011] In a preferred embodiment, the aggregation channel quality is calculated using any of the following methods: The average interference intensity, average signal-to-noise ratio, average received signal strength, average channel utilization, or weighted fusion score of all channels within the subset; The measurement is achieved by synchronously monitoring multiple channels in a subset within the same time period, or by sequentially hopping frequencies within a set short time slot and calculating the average value. Furthermore, when M is not equal to an integer power of 2, it can be padded to 2^L by introducing a virtual channel, or the probe subset can be dynamically partitioned using a non-complete binary tree method.

[0012] In a preferred embodiment, before step S2, if the number of channels in the remaining channel set is less than or equal to a preset small channel number threshold, then the complete quality index of each of the remaining channels is measured directly and the best one is selected as the target channel to replace the bit-by-bit probing. The threshold for the number of small channels is 2, 3, or 4.

[0013] In a preferred embodiment, S4 is also included: The result of this channel handover is recorded in the historical database, and the channel quality assessment parameters are updated to optimize the decision threshold for subsequent sensing and the measurement threshold for the aggregated channel quality. The decision threshold is independent of the first to fifth thresholds.

[0014] In a preferred embodiment, the criterion for determining whether the target channel is superior to the current working channel in step S3 is a comparison with hysteresis: Channel switching is performed when the overall quality score of the target channel is higher than that of the current working channel and exceeds the preset hysteresis threshold; otherwise, the current channel is maintained.

[0015] In a preferred embodiment, it further includes: Based on historical probe data or prior information about channel quality, the probe order of bit-by-bit probes in S2 is adjusted, prioritizing the probe of coded bits that contribute more to the decision.

[0016] The present invention also provides a wireless network signal channel intelligent switching system, comprising: The sensing module is used to sense at least one other channel besides the current working channel while maintaining communication on the current working channel, and to identify and mark the inferior channel. The detection module is used to perform bit-by-bit detection on the remaining channel set after excluding inferior channels, according to the binary encoding rules, to obtain the aggregated channel quality measurement results of each detection. The decision module is used to decode the target channel number based on the measurement results, determine whether the target channel is better than the current working channel, and output a switching decision. The execution module is used to perform channel switching or maintain the current channel based on the switching decision.

[0017] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention does not rely on external positioning information such as GPS / BeiDou. It can complete channel sensing and coding detection using only communication gaps. Compared with existing solutions that completely fail due to GNSS signal interruption in tunnels, this invention can still work normally in tunnel scenarios, fundamentally solving the handover problem when trains pass through tunnels. Furthermore, through a timing coupling mechanism that first eliminates inferior channels and then performs binary coding detection, this invention can deterministically lock the globally optimal channel. Theoretical analysis and simulation verification show that the handover success rate can reach over 99.5%, and the handover delay is controlled within 30ms, fully meeting the stringent requirements of 5G-R systems for high-reliability services such as automatic train operation.

[0018] This invention eliminates a low-quality channel in the first stage, and the second stage of coded probing is performed on the reduced channel set. Taking a total of 17 channels as an example, direct binary coded probing requires 5 probes, while this scheme, after eliminating one low-quality channel, leaves 16 channels and requires only 4 probes, reducing the number of probes by 1. Combined with extremely low-overhead communication gap sensing, this achieves efficient allocation of probe resources and improved handover efficiency. Furthermore, by dynamically optimizing the decision threshold through historical feedback learning and combining it with a hysteresis comparison criterion, ping-pong handover is effectively prevented. Simultaneously, it supports adaptive optimization of the probe sequence based on prior information, further ensuring the long-term stability of the communication link. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0020] Figure 1 This is a flowchart of the method of the present invention.

[0021] Figure 2 This is a system block diagram of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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.

[0023] Prior information refers to historical data or preset information related to channel quality that the system has already obtained before this handover decision, including but not limited to: historical packet loss rate, historical signal-to-noise ratio, historical interference intensity, tunnel section information predicted based on train operation schedule, and a list of prohibited channels marked based on past handover experience.

[0024] Temporally coupled mode: This refers to the specific temporal relationship between steps S1 and S2. S1 must be executed before S2, and the result of S1 provides the input conditions for S2 (the set of remaining channels after excluding inferior channels). The execution of S2 depends on the output of S1. There is no possibility of parallel execution or reversal between the two.

[0025] Hysteresis threshold: This refers to the hysteresis comparison value set in the handover decision to avoid frequent handovers caused by minor fluctuations in channel quality. Specifically, a handover is only triggered when the difference between the overall quality score of the target channel and the score of the current operating channel is greater than this threshold value; otherwise, it is considered equivalent, and the current channel is maintained. This threshold value can be preset or dynamically adjusted.

[0026] Aggregated channel quality: defined as "the comprehensive statistical value of the channel quality indicators within this group". Specifically, statistical methods such as average, weighted sum, and median can be used to compress multiple channel quality indicators (such as signal-to-noise ratio, interference intensity, received signal strength, etc.) of a channel subset into a representative value, which is used to compare the overall quality of different subsets.

[0027] Example 1, please refer to Figure 1 As shown in this embodiment, a method for intelligent switching of wireless network signal channels includes: S1. While maintaining communication on the current working channel, sense at least one other channel besides the current working channel, identify and mark at least one inferior channel, wherein the inferior channel meets a preset low-quality condition and is different from the current working channel. S2. For the remaining channel set after excluding the inferior channels, perform bit-by-bit probing according to the binary encoding rules. Each bit of probing corresponds to measuring the aggregated channel quality of a group of channels. The aggregated channel quality refers to the comprehensive statistical value of the quality indicators of each channel in the group. The target channel number is obtained by decoding based on the results of each probing. S3. Determine whether the target channel corresponding to the target channel number is better than the current working channel. If so, perform channel switching; otherwise, maintain the current channel.

[0028] As described in S1-S3 above, the wireless network signal channel intelligent switching method provided in this embodiment, through three core steps and an optional feedback learning step S4, addresses the long-standing technical problem pointed out in the background technology that "the complete obstruction of GNSS signals in the tunnel causes the failure of the existing location prediction switching scheme," and adopts a completely different technical path from the existing technology.

[0029] Specifically, traditional location prediction-based schemes rely on real-time GPS / BeiDou positioning data. Within a few meters of the tunnel entrance, the signal carrier-to-noise ratio drops sharply to below 20dB-Hz, resulting in complete loss of lock and making it impossible to provide the location information required for switching decisions. In contrast, step S1 requires no external positioning input and can quickly sense other channels by utilizing the communication gap of the current working channel, and mark at least one poor-quality channel (such as the worst-quality channel), thereby reducing the search space from all N channels to N-1 channels.

[0030] Based on this, step S2 performs binary-coded bit-by-bit probing on the reduced set of remaining channels. By associating the binary code of each channel with the subset aggregation quality measurement results, it is possible to... The target channel number is deterministically decoded within the first detection. Because S1 and S2 are executed in a time-coupled manner—first eliminating inferior channels to reduce the detection space, and then encoding the detection within that reduced space—this method simultaneously achieves the dual technical effects of deterministic optimal channel positioning and low detection overhead, and the entire process is completely independent of satellite positioning information. Theoretical analysis and simulation verification show that this method can meet the stringent requirements of 5G-R systems for a handover success rate of no less than 99.5% and a handover delay of no more than 30ms, fundamentally solving the "handover blind spot" problem when trains pass through tunnels, and providing reliable communication guarantees for critical services such as high-speed railway automatic driving.

[0031] To more clearly illustrate the technical solution of the present invention, the following describes each step and its optional implementation methods in detail with reference to several specific embodiments.

[0032] In one specific implementation, the sensing in step S1 can be a rapid energy detection of some or all other channels other than the current working channel.

[0033] For example, in a high-speed rail tunnel scenario, the signal quality of the current operating channel may plummet due to multipath fading. In this case, the system can sequentially sample other candidate channels for several microseconds to tens of microseconds without interrupting communication, obtaining the Received Signal Strength Indication (RSSI) or noise floor of each channel. Through this rapid sensing, the system can preliminarily determine which channels are in a significantly degraded state with minimal overhead (typically less than 1 ms).

[0034] Its advantages are: it can complete the initial screening of inferior channels without relying on external positioning information, using only communication gaps, which reduces the search space for subsequent accurate detection, and avoids long-term interruptions caused by full-band scanning.

[0035] In another specific implementation, identifying and marking at least one inferior channel in step S1 can specifically mean that after sensing the quality data of multiple other channels, the system marks the channel with the worst quality as an inferior channel.

[0036] For example, if the signal-to-noise ratio (SNR) of channel A is 5dB, that of channel B is 12dB, and that of channel C is 8dB, then channel A is marked as a poor-quality channel. This marking operation is equivalent to removing a deterministically unusable channel from the subsequent binary-coded probing, thus simplifying the problem of searching for the optimal channel in the entire channel set to locating the optimal channel in the reduced set. Theoretical analysis shows that removing each poor-quality channel can reduce the number of binary-coded probing attempts by at most one, and enhance the determinism of the handover decision.

[0037] In one specific implementation, the preset low-quality conditions may include at least one of the following: interference intensity exceeding a first threshold, signal-to-noise ratio below a second threshold, received signal strength below a third threshold, channel utilization exceeding a fourth threshold, packet loss rate exceeding a fifth threshold, or being marked as a prohibited channel based on prior information. The first threshold (interference intensity) can be set to a value between -70dBm and -90dBm; the second threshold (signal-to-noise ratio) can be set to a value between 5dB and 15dB; the third threshold (received signal strength) can be set to a value between -85dBm and -65dBm; the fourth threshold (channel utilization) can be set to a value between 60% and 80%; and the fifth threshold (packet loss rate) can be set to a value between 5% and 15%. These thresholds can be pre-configured or dynamically adjusted according to the actual network environment.

[0038] For example, in a tunnel scenario, the signal-to-noise ratio threshold can be set to 10dB (this value can be adjusted according to actual measurements within the range of 5dB to 15dB). When the instantaneous signal-to-noise ratio of a channel is lower than this threshold, it is judged as poor quality. The interference intensity threshold can be set to -80dBm, and the channel utilization threshold can be set to 70%.

[0039] In another specific implementation, the marking of inferior channels in step S1 can also be aided by combining historical handover data with current environmental information. For example, the system can maintain a historical channel quality rating table, recording indicators such as the average signal-to-noise ratio and average packet loss rate of each channel over a period of time. When a train passes through the same tunnel multiple times, the system can predict, based on historical data, that certain channels are likely to be in an inferior state in specific sections of the tunnel. Therefore, in the sensing phase of S1, these channels are prioritized for rapid energy detection, and their low-quality condition threshold is lowered (e.g., the signal-to-noise ratio threshold is lowered by 3dB), making these channels easier to mark as inferior channels. This predictive mechanism can further accelerate the identification speed of inferior channels, especially in scenarios where the tunnel environment is relatively stable, reducing the number of actual measurements and improving the timeliness of handover decisions.

[0040] In another specific implementation, the sensing in S1 can be performed in any of the following ways: Method 1 involves timing scans of other channels during communication gaps without interrupting current communication. This method is suitable for low-cost devices with a single radio frequency link. Method 2 involves parallel listening using independent radio frequency links. This method is suitable for high-performance devices with multi-radio frequency capabilities and can further shorten the sensing time.

[0041] Regardless of the method used, the sensing process does not occupy the communication bandwidth of the current working channel, thus ensuring service continuity.

[0042] In another specific embodiment, the binary coding detection of the remaining channel set in step S2 can be implemented as follows: Suppose that the remaining channel set contains M channels. First, assign an L-bit binary code to each channel, where L is the smallest integer that satisfies 2^L≥M.

[0043] For example, if M=12, then L=4; if M=5, then L=3. Then, for the i-th bit (i from 1 to L), the system groups all channels in the remaining channel set that are encoded as "1" into a first subset, and all channels encoded as "0" into a second subset, and measures the aggregated channel quality of these two subsets respectively. The aggregated channel quality can refer to the average signal-to-noise ratio, average received signal strength, or weighted composite score of all channels within the subset. During measurement, the system can use FPGA or SDR hardware to synchronously and rapidly sample multiple channels in the subset within the same measurement period, or it can sequentially hop and listen within extremely short consecutive time slots. By comparing the measurement results, the system selects the encoded bit value (0 or 1) corresponding to the subset with the better aggregate quality as the decision result for that bit. This process is repeated until all L bits have been decided, and finally, the L bit decision results are merged and decoded to uniquely determine the target channel number.

[0044] The beneficial effect of this mechanism is that it only requires L probes (usually L≤L). This method can deterministically locate the optimal channel from M channels, which is several times more efficient than the traditional M-times scanning method. At the same time, since each detection is a parallel evaluation of the overall quality of a group of channels, the statistical stability of the measurement results is higher and less susceptible to interference from instantaneous fluctuations.

[0045] In another specific implementation, when the system has multiple independent radio frequency links, the bit-by-bit probing in step S2 can be performed in parallel.

[0046] Specifically, the system can allocate multiple bits of the L-bit binary code to different radio frequency links for simultaneous probing, with each radio frequency link responsible for measuring the aggregated channel quality of its corresponding subset.

[0047] For example, when L=4 and the system has two RF links, bits 1 and 2 can be assigned to the first link, and bits 3 and 4 to the second link. The two links perform their respective subset aggregation measurements within the same time period, thereby reducing the total detection time from L consecutive time slots to ⌈L / 2⌉ time slots. This parallel detection method is suitable for communication equipment with multiple antennas or transceivers (such as high-speed rail vehicle-mounted radios or high-end industrial gateways), and can further compress the total time spent on handover decisions without increasing the number of detections, meeting the ultra-high reliability requirements (e.g., within 15ms) of scenarios with extreme handover latency requirements.

[0048] Specifically, S1 and S2 are executed in a time-coupled manner: S1 first eliminates inferior channels to reduce the search space, and S2 only performs binary-coded probing on the reduced set of remaining channels. This cascaded architecture of elimination followed by probing enables the handover decision to simultaneously possess deterministic optimal channel positioning capability and low probing overhead characteristics, and the entire handover decision process does not depend on external satellite positioning information.

[0049] Specifically, since after S1 eliminates at least one inferior channel, the remaining number of channels M is equal to the total number of available channels N minus 1, therefore the number of bit-by-bit probes L in S2 = The number of times the binary code probe is performed directly on all N channels is not higher than the number of times the probe is performed directly on all N channels. Furthermore, the number of probes decreases by one when the remaining number of channels is a power of 2 (i.e., N = 2^k + 1). For example, with a total number of channels N = 17, direct detection requires 5 probes. =5), while this scheme eliminates one poor-quality channel and leaves 16 channels, so S2 only needs 4 probes ( =4), reducing the number of detections by 1, and the sensing overhead of S1 is much smaller than that of a single detection, thus achieving effective allocation of detection resources and improved switching efficiency. This combined effect cannot be achieved independently by a single technical feature.

[0050] In another specific implementation, the aggregated channel quality can be calculated using any of the following methods: average interference intensity, average signal-to-noise ratio, average received signal strength, average channel utilization, or weighted fusion score for all channels within the subset.

[0051] For example, weighted fusion scoring can assign different weights to different quality indicators and then calculate a weighted sum. The measurement can be achieved by simultaneously monitoring multiple channels in a subset within the same time period (using a broadband receiver to sample multiple channels simultaneously), or by sequentially hopping frequencies within a set short time slot and calculating the average value. When M is not an integer power of 2 (e.g., M=12, 2^3=8, 2^4=16, M is between these two), virtual channel padding can be used: several virtual channels are introduced, and their aggregated quality is preset to the lowest value (e.g., set to negative infinity or the minimum possible score), making the total number reach 2^L, and then the normal binary coded probing process is executed; or a non-complete binary tree method is used to dynamically divide the probing subset, dynamically adjusting the number of channels in the first and second subsets at certain bits to make them as balanced as possible. Both methods ensure the completeness of the probing process, and the decoded target channel number always corresponds to the real channel. This processing method solves the problem that coded probing cannot be directly applied when the number of channels is not exactly a power of 2 in practical applications, expanding the applicability of this method.

[0052] In one specific implementation, when the number of remaining channels M is not equal to an integer power of 2 (e.g., M=12), a non-complete binary tree approach can be used for probe subset partitioning. The specific steps are as follows: Step 1: Calculate L= =4, the total number of detections is 4.

[0053] Step 2: For the first bit (most significant bit), divide the M channels into two subsets as evenly as possible. Let M=12, then allocate 6 channels to the first subset (coded with 1 bit) and 6 channels to the second subset (coded with 0 bits). Measure the aggregated channel quality of the two subsets respectively, and select the coded bit value corresponding to the better one.

[0054] Step 3: For the second bit, instead of forcibly dividing according to the weight definition of binary encoding, it is dynamically re-divided based on the size of the remaining candidate channel set: assuming the decision result of the first bit is 1, the remaining candidate channels are the 6 channels of the first subset; these 6 channels are divided into two subsets (3 channels each) as evenly as possible according to the encoding rule of the current bit. The bit is then determined after measurement.

[0055] Step 4: Repeat the above dynamic partitioning process until L probes are completed. During each partitioning, if the subset sizes are unequal, a difference of one channel is allowed.

[0056] Step 5: Combine the L-bit decision results (each bit corresponds to 0 or 1) in sequence to obtain the target channel's coding number, and then map it back to the actual channel identifier.

[0057] This incomplete binary tree partitioning method ensures that L probes can be completed under any value of M, and that the two subsets of each probe are basically the same size, thus guaranteeing the balance and comparability of the measurement results.

[0058] In another specific implementation, the weighting coefficient of the weighted fusion score in the calculation method of the aggregated channel quality can be dynamically adjusted according to the current channel quality fluctuation level.

[0059] For example, the system can monitor the instantaneous signal-to-noise ratio (SNR) change rate of the current operating channel in real time. If the change rate is high (indicating the channel is in a fast fading environment), the weight of the average SNR in the aggregation score is increased, while the weight of the average interference intensity is decreased. Conversely, if the change rate is low (the channel is relatively stable), the weight of the interference intensity is appropriately increased to more accurately identify persistent interference. This dynamic weight adjustment mechanism can learn and adapt online, making the aggregation quality evaluation standard used in each decision of binary coded probing more closely match the current channel characteristics, thereby improving the accuracy of optimal channel determination.

[0060] In another specific implementation, before step S2, if the number M of the remaining channel set is less than or equal to a preset small channel number threshold (e.g., 2, 3, or 4), then all remaining channels can be precisely measured one by one (e.g., measuring the full signal-to-noise ratio, throughput, etc. of each channel), and then the optimal one can be selected as the target channel to replace the bit-by-bit probing. This processing method avoids the fixed overhead of coded probing in a small search space, resulting in higher overall switching efficiency. The small channel number threshold can be flexibly configured according to the total number of channels and real-time requirements of the actual application scenario, preferably 2, 3, or 4.

[0061] In another specific implementation, the system can also dynamically select whether to skip binary coded probing or adopt a simpler probing mode based on historical handover frequency or channel congestion. For example, if the system has very few handovers in the past hour (e.g., less than 3 times), indicating that the current working channel is stable and environmental interference is minimal, then when a handover is triggered, the aforementioned small channel number substitution scheme (precise measurement one by one) can be directly adopted, even if the number of remaining channels is greater than the small channel number threshold, simplified probing can still be selected. Conversely, if the handover frequency is high (e.g., more than 10 times / hour), the system enables full binary coded probing from S1 to S2 to more accurately locate the optimal channel. This hierarchical probing strategy, while ensuring reliability, further optimizes the balance between probing overhead and handover accuracy based on dynamic changes in the scenario.

[0062] In another specific embodiment, the method may also include an optional feedback learning step S4.

[0063] Specifically, after each channel handover is completed, the system can record the results of this handover (including the channel identifiers before and after the handover, the threshold used for the handover decision, the measured performance after the handover, etc.) in the historical database, and use this historical data to update the channel quality assessment parameters.

[0064] For example, the decision thresholds for low-quality conditions in step S1 (such as interference intensity thresholds, signal-to-noise ratio thresholds, etc.) and the thresholds for aggregated channel quality measurements in step S2 (such as the weighting coefficient of the average value) can be dynamically adjusted. The adjustment parameters for the decision thresholds are independent of the first to fifth thresholds mentioned above to avoid mutual interference. This feedback learning mechanism enables the system to adapt to environmental changes (such as the impact of temperature and humidity in different seasons within the tunnel on signal attenuation), continuously self-optimizing as operating time increases, and gradually improving the accuracy of switching decisions.

[0065] In another specific implementation, the first to fifth thresholds are not fixed values, but can be adaptively adjusted based on the results of closed-loop verification.

[0066] Specifically, in step S6 (closed-loop verification), the system compares the actual performance after the handover (such as throughput and latency) with the predicted performance before the handover. If the actual performance is significantly worse than the predicted performance, it indicates that the current threshold setting is either too aggressive (resulting in the failure to eliminate inferior channels in a timely manner) or too conservative (resulting in the premature elimination of available channels). The system can then fine-tune the relevant thresholds according to a predetermined step size (e.g., ±1dB each time) until the deviation between the actual and predicted performance after the handover converges to a preset allowable range. This adaptive threshold adjustment mechanism complements the feedback learning step in S4; one focuses on parameter drift caused by long-term environmental changes (feedback learning), while the other focuses on the immediate correction of the single decision threshold (closed-loop adjustment), jointly improving the robustness and adaptability of the system.

[0067] In another specific implementation, the criterion for determining whether the target channel is superior to the current working channel in step S3 can be a comparison with hysteresis.

[0068] Specifically, the system can pre-set a hysteresis threshold (e.g., a 3dB or 5% score difference). Channel switching is only triggered when the overall quality score of the target channel is higher than that of the current operating channel and exceeds the hysteresis threshold; otherwise, even if the target channel score is slightly higher but does not meet the threshold difference, the system maintains the current channel. This hysteresis comparison criterion effectively prevents frequent ping-pong switching when channel quality fluctuates critically, thereby further improving the stability of the communication link.

[0069] For example, when a high-speed train passes through a tunnel exit, the signal may fluctuate briefly. The hysteresis threshold can filter out these minor changes that do not affect the actual communication quality, thus avoiding unnecessary handover overhead.

[0070] In another specific implementation, it may also include: adjusting the detection order of bit-by-bit detection in S2 based on historical detection data or prior information about channel quality, prioritizing the detection of coded bits that contribute more to the decision.

[0071] For example, if the system discovers from historical records that certain coding bits (such as the highest bit) are more effective at distinguishing channel quality differences, these higher bits can be probed first to obtain more significant decision information earlier. This allows the probe to be terminated earlier in some cases (e.g., when the aggregate quality of one subset is extremely high while that of another subset is extremely poor, the impact of subsequent lower bits on the final decision can be ignored). This adaptive optimization of the probe order can further reduce the average number of probes and accelerate the convergence speed of the handover decision. Its beneficial effect is that, in scenarios with high real-time requirements (such as automatic train operation), it can compress the handover decision time as much as possible while ensuring positioning accuracy.

[0072] In another specific implementation, the sensing in step S1 may not be limited to one-time sensing, but may be repeatedly triggered when the quality of the current working channel continues to decline.

[0073] For example, when the packet loss rate of the current working channel exceeds the threshold three times consecutively, the system can automatically initiate a complete S1 to S3 process. This triggering mechanism enables on-demand switching, avoiding unnecessary power consumption and spectrum occupation caused by periodic scanning.

[0074] Example 2, please refer to Figure 2 As shown in this embodiment, a wireless network signal channel intelligent switching system includes: The sensing module is used to sense at least some of the other channels while maintaining communication on the current working channel, and to identify and mark inferior channels; The detection module is used to perform bit-by-bit detection on the remaining channel set after excluding inferior channels, according to the binary encoding rules, to obtain the aggregated channel quality measurement results of each detection. The decision module is used to decode the target channel number based on the measurement results, determine whether the target channel is better than the current working channel, and output a switching decision. The execution module is used to perform channel switching or maintain the current channel based on the switching decision.

[0075] In one embodiment, the sensing module may have a built-in signal acquisition unit and a defective marking unit. The signal acquisition unit may be equipped with a single-radio time-division scanning circuit or a dual-radio parallel monitoring circuit to adapt to the two sensing methods in Embodiment 1. The defective marking unit may pre-store a first threshold to a fifth threshold and a list of prohibited channels, filter defective channels according to multi-dimensional judgment rules, and enter the defective channel number into the blocking list.

[0076] This module can perform preliminary channel screening and filter out invalid and inferior channels, simplifying the computational burden on the backend detection module. The overall effect of the combination of hardware and logic is that it can automatically complete the initial channel screening without manual intervention, and is suitable for wireless networks in various scenarios such as home and industrial environments.

[0077] In one embodiment, the detection module may have a built-in coding allocation unit, a group detection unit, and a data statistics unit. The coding allocation unit can adaptively identify the number of remaining channels, automatically generate binary codes, and determine the number of coding bits. The group detection unit can divide the channels into subsets according to coding rules and collect subset channel data synchronously or in a time-division manner. The data statistics unit can be equipped with multiple aggregation algorithms and can switch between average calculation and weighted fusion calculation modes as needed.

[0078] This module relies on binary encoding to achieve efficient detection and accurate collection of aggregated channel quality data. The overall effect is to reduce the complexity of detection operations, reduce hardware computing power consumption, and adapt to low-cost embedded wireless devices.

[0079] In one embodiment, the decision-making module may include a built-in decoding unit, a quality comparison unit, and a hysteresis determination unit. The decoding unit is used to concatenate multi-bit encoded results and parse them to obtain a unique target channel number. The quality comparison unit can retrieve real-time channel scoring data to compare the parameters of the target channel and the currently operating channel. The hysteresis determination unit may include an editable hysteresis threshold to avoid frequent handover issues. The module's function is to complete data parsing and intelligent determination, and output accurate handover commands. The overall combined effect is to avoid human error in judgment and improve the rationality and stability of handover decisions.

[0080] In one embodiment, the execution module may include a built-in channel switching unit and a status feedback unit. The channel switching unit may employ a soft switching method, with a 2ms buffer time reserved between switching intervals to reduce data packet loss. The status feedback unit can transmit the switching result, switching time, and post-switching channel quality data back to the sensing module, providing data support for database iterative optimization. The module's function is to accurately execute switching actions and complete status feedback. The overall combined effect is to achieve closed-loop system control, cooperate with other modules to complete adaptive optimization, and improve the long-term operating quality of the wireless network.

[0081] The above description is merely a 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 intelligent switching of wireless network signal channels, characterized in that, include: S1. While maintaining communication on the current working channel, sense at least one other channel besides the current working channel, identify and mark at least one inferior channel, wherein the inferior channel meets a preset low-quality condition and is different from the current working channel. S2. For the remaining channel set after excluding the inferior channels, perform bit-by-bit probing according to the binary encoding rules. Each bit of probing corresponds to measuring the aggregated channel quality of a group of channels. The aggregated channel quality refers to the comprehensive statistical value of the quality indicators of each channel in the group. The target channel number is obtained by decoding based on the results of each probing. S3. Determine whether the target channel corresponding to the target channel number is better than the current working channel. If so, perform channel switching; otherwise, maintain the current channel.

2. The method of claim 1, wherein: The preset low-quality conditions include at least one of the following: Interference intensity is higher than the first threshold, signal-to-noise ratio is lower than the second threshold, received signal strength is lower than the third threshold, channel utilization is higher than the fourth threshold, packet loss rate is higher than the fifth threshold, or it is marked as a prohibited channel based on prior information; Furthermore, when quality data from multiple other channels are detected, the channel with the worst quality is marked as the inferior channel, and the inferior channel is unique.

3. The method for intelligent switching of wireless network signal channels according to claim 1, characterized in that: The sensing in S1 is performed in any of the following ways: During communication gaps without interrupting current communication, perform timing scans of other channels or use independent radio frequency links for parallel monitoring.

4. The method for intelligent switching of wireless network signal channels according to claim 1, characterized in that: The bit-by-bit detection in S2 includes: Let the total number of available channels be N, and the number of channels in the remaining channel set be M = N-1; assign an L-bit binary code to each of the M channels, where L is the smallest integer satisfying 2^L≥M; For each bit, the channels encoded as 1 form the first subset, and the channels encoded as 0 form the second subset. The aggregated channel quality of the two subsets is then measured. Select the coded bit value corresponding to the subset with better aggregation channel quality as the decision result for that bit; The L-bit decision results are combined and decoded to obtain the target channel number.

5. The method for intelligent switching of wireless network signal channels according to claim 4, characterized in that: The aggregation channel quality is calculated using any of the following methods: The average interference intensity, average signal-to-noise ratio, average received signal strength, average channel utilization, or weighted fusion score of all channels within the subset; The measurement is achieved by synchronously monitoring multiple channels in a subset within the same time period, or by sequentially hopping frequencies within a set short time slot and calculating the average value. Furthermore, when M is not equal to an integer power of 2, it can be padded to 2^L by introducing a virtual channel, or the probe subset can be dynamically partitioned using a non-complete binary tree method.

6. The method for intelligent switching of wireless network signal channels according to claim 1, characterized in that: Before S2, if the number of channels in the remaining channel set is less than or equal to a preset small channel number threshold, then the complete quality index of each of the remaining channels is measured directly and the best one is selected as the target channel to replace the bit-by-bit probing. The threshold for the number of small channels is 2, 3, or 4.

7. The method for intelligent switching of wireless network signal channels according to claim 1, characterized in that: Also includes S4: The results of this channel handover are recorded in the historical database, and the channel quality assessment parameters are updated to optimize the decision threshold for subsequent sensing and the measurement threshold for the aggregated channel quality. The decision threshold is independent of the first to fifth thresholds.

8. The method for intelligent switching of wireless network signal channels according to claim 1, characterized in that: In step S3, the criterion for determining whether the target channel is superior to the current working channel is a comparison with hysteresis. Channel switching is performed when the overall quality score of the target channel is higher than that of the current working channel and exceeds the preset hysteresis threshold; otherwise, the current channel is maintained.

9. The method for intelligent switching of wireless network signal channels according to claim 1, characterized in that: Also includes: Based on historical probe data or prior information about channel quality, the probe order of bit-by-bit probes in S2 is adjusted, prioritizing the probe of coded bits that contribute more to the decision.

10. A wireless network signal channel intelligent switching system, used to implement the wireless network signal channel intelligent switching method according to any one of claims 1-9, characterized in that, include: The sensing module is used to sense at least one other channel besides the current working channel while maintaining communication on the current working channel, and to identify and mark the inferior channel. The detection module is used to perform bit-by-bit detection on the remaining channel set after excluding inferior channels, according to the binary encoding rules, to obtain the aggregated channel quality measurement results of each detection. The decision module is used to decode the target channel number based on the measurement results, determine whether the target channel is better than the current working channel, and output a switching decision. The execution module is used to perform channel switching or maintain the current channel based on the switching decision.

Images