A method for suppressing interference between the N79 band and the Wi-Fi 5G band in 5GCPE
By combining bandpass filters and hardware isolation of antenna polarization direction in 5GCPE devices, along with cross-chip state synchronization and time-domain synchronization, an internal interference fingerprint matrix is generated, and the transmit power and resource scheduling are dynamically adjusted. This solves the mutual interference problem when the N79 band and the Wi-Fi 5G band work concurrently, and improves communication stability and efficiency.
Patent Information
- Application Number
- CN202611446661.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-17
- Publication Date
- 2026-11-17
AI Technical Summary
In existing technologies, when 5GCPE devices operate concurrently in the N79 band and the Wi-Fi 5G band, the hardware isolation is limited and there is a lack of a precise cross-chip time-domain synchronization mechanism. This causes the Wi-Fi system to easily misjudge sudden interference transmitted internally by 5G as external radar signals or channel fading, resulting in invalid channel switching and abnormal decrease in communication rate.
By initializing the physical layer static isolation settings, using a combination of bandpass filters and antenna polarization directions, and combining cross-chip state synchronization and time-domain synchronization, an internal interference fingerprint matrix is generated, and the transmit power and resource scheduling are dynamically adjusted to avoid misjudgment and interference.
It effectively suppressed the interference between the N79 band and the Wi-Fi 5G band, reduced system complexity and power consumption, ensured the stability and efficiency of the communication link, and avoided unexpected channel switching and rate drop.
Smart Images

Figure CN123124099A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of wireless communication technology, specifically to a method for suppressing interference between the N79 band and the Wi-Fi 5G band in 5GCPE. Background Technology
[0002] With the development of 5G communication technology, 5G client terminal equipment typically needs to support the concurrent operation of 5G cellular networks and Wi-Fi local area networks. In one operating configuration described in this application, the N79 RF front-end operates in the 4800MHz to 4960MHz sub-band, while the center frequency of commonly used channels on the Wi-Fi 5GHz low-frequency side can be located in the 5180MHz and above frequency region. Because the two are relatively close in spectral location, and the 5G N79 band operates in time-division duplex mode, when the 5G module performs high-power uplink transmission, its out-of-band RF leakage energy can easily enter the receiving path of the Wi-Fi receiver, causing an increase in the receiver noise floor or receiver blockage.
[0003] Traditional RF coexistence solutions primarily rely on static isolation at the physical layer, such as increasing the physical distance between 5G and Wi-Fi antennas, or adding high-order bandpass filters to the RF front-end. However, in highly integrated client terminal devices, the physical space is limited, and hardware alone cannot provide sufficient isolation. To compensate for the shortcomings of hardware isolation, existing software coexistence technologies often employ static power back-off or coarse-grained frequency avoidance, but this directly reduces the uplink coverage of 5G or wastes Wi-Fi spectrum resources.
[0004] More critically, existing technologies lack a precise time-domain synchronization mechanism across chips at the underlying level. When the N79 band transmits periodic time-division duplex uplink pulses, the Wi-Fi system cannot distinguish whether this burst of energy originates from the external environment or internal to the device. This lack of information leads to two problems: First, the Wi-Fi system's dynamic frequency selection engine may misjudge this internal burst of interference as an external radar signal, triggering the device to perform a forced channel switch, causing an unexplained interruption of the communication link. Second, burst packet loss caused by internal interference can trigger the error recovery logic of the Wi-Fi medium access control layer, causing the rate adaptation algorithm to mistakenly believe that the current spatial channel quality has deteriorated, thus blindly and continuously downgrading the modulation and coding strategy, resulting in a significant decrease in system throughput. Existing technologies cannot intervene in the rate downgrading logic based on the identification of internal interference characteristics, nor can they avoid high-power transmission slots in advance on the time axis, making it difficult to fundamentally guarantee communication efficiency during dual-band concurrent operation. On the other hand, Wi-Fi receive links typically generate receive quality information such as received power, background noise, and automatic gain control status. However, this information primarily serves Wi-Fi's own physical layer receive control and lacks a deterministic temporal correlation with the actual N79 uplink time slots and real-time transmit power of the 5G cellular baseband. Therefore, a key technical problem to be solved is how to identify the causal relationship between existing receive quality measurements and internal N79 transmissions without additionally reconstructing the RF sampling link, and how to further feed the identification results back to DFS decision-making, rate adaptation, and cross-mode resource scheduling. Summary of the Invention
[0005] To address the shortcomings of existing technologies, this invention provides a method for suppressing mutual interference between the N79 band and the Wi-Fi 5G band in 5GCPE. This method solves the problem that when existing 5GCPE devices operate concurrently in the N79 and Wi-Fi 5G bands, the limited hardware isolation and lack of a precise cross-chip time-domain synchronization mechanism cause the Wi-Fi system to easily misjudge sudden interference transmitted internally by 5G as external radar signals or channel fading, resulting in invalid channel switching and abnormally low communication rates.
[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for suppressing interference between the N79 band and the Wi-Fi 5G band in 5GCPE, comprising the following steps: Initialize the physical layer static isolation settings to establish RF out-of-band attenuation suppression and spatial electromagnetic isolation. Configure the N79 antenna to operate in the vertical polarization radiation direction and the Wi-Fi antenna to operate in the horizontal polarization radiation direction. Use the N79 RF front-end link with an internally connected first bandpass filter to perform signal transmission and reception. The passband of the first bandpass filter is set to 4800MHz to 4960MHz. Use the Wi-Fi RF front-end link with an internally connected second bandpass filter to perform signal transmission and reception. The passband of the second bandpass filter is set to 5180MHz to 5825MHz.
[0007] Perform cross-chip state synchronization and risk state determination. The 5G cellular baseband processor sends the current N79 carrier center frequency, N79 carrier channel bandwidth, N79 actual uplink transmit power, and N79 time-division duplex frame structure synchronization information to the Wi-Fi system-on-a-chip via the coexistence management bus; the Wi-Fi system-on-a-chip obtains the lower edge frequency of the current Wi-Fi channel and calculates the dynamic guard distance; the value of the dynamic guard distance is equal to the lower edge frequency of the Wi-Fi channel minus the difference between the sum of the N79 carrier center frequency and half of the N79 carrier channel bandwidth; when the dynamic guard distance is less than the preset safe distance threshold and the N79 actual uplink transmit power is greater than the preset transmit power trigger threshold, it is determined to enter a high mutual interference risk state.
[0008] The system performs receive noise floor monitoring and DFS-assisted event discrimination based on cross-chip time-domain synchronization. Upon determining a high-interference-risk state, the Wi-Fi system-on-a-chip (SoC) generates a hardware-level time-domain gating signal corresponding to the N79 uplink transmit time slot based on the N79 time-division duplex frame structure information transmitted via the coexistence management bus. It then uses the resident receive quality monitoring unit in the Wi-Fi receive link to obtain background noise estimates within the N79 uplink transmit time slot and adjacent non-uplink reference periods. The difference between these estimates is used to calculate the noise floor rise, and an internal interference fingerprint matrix is generated by combining this with the actual uplink transmit power of the N79. The receive quality monitoring unit reuses at least one of the following information generated during normal Wi-Fi system reception: receive power estimate, background noise estimate, signal-to-noise ratio estimate, and automatic gain control status. It does not require the use of an RF front-end, mixer, local oscillator, or analog-to-digital converter independent of the Wi-Fi RF front-end link.
[0009] When Wi-Fi operates in the forced DFS channel, the DFS scanning engine maintains the standard DFS settings to perform radar detection. When a burst energy pulse is captured, the timestamp of the burst energy pulse is extracted and compared with the synchronously acquired N79 uplink transmission time slot boundary in the time domain. If the timestamp falls completely within the time slot, the peak values of multiple consecutive burst energy pulses are cached in memory to form a measured peak sequence, and the actual uplink transmission power of N79 during the same period is extracted to form a reference power sequence. Then, the Pearson correlation coefficient between the two sequences is calculated. When the calculated Pearson correlation coefficient is greater than the set correlation threshold, it is determined that the burst energy is an artifact caused by internal out-of-band blocking. The underlying driver triggers a hardware interrupt rejection signal to intercept the channel switching command, and the background noise estimate output by the reception quality monitoring unit in the corresponding time period is extracted as the measured noise floor.
[0010] The system performs rate adaptation sampling and masking based on interference fingerprints. The Wi-Fi system-on-a-chip (SoC) monitors MAC layer packet acknowledgment timeout events and detects packet loss timestamps. When a packet loss timestamp falls within any N79 time-division duplex uplink transmit time slot closed interval recorded in the internal interference fingerprint matrix, and the noise floor rise exceeds a preset tolerance threshold, the SoC generates a sampling and masking command with an active masking flag and sends it to the rate adaptation microcontroller. Upon receiving the sampling and masking command, the rate adaptation microcontroller masks the corresponding abnormal sample from the modulation and coding scheme downsampling statistics and prevents the modulation and coding scheme from being triggered by the abnormal sample alone. The modulation and coding strategy is downgraded, and the valid samples that are not blocked continue to participate in the normal update of the rate adaptation state machine. Weighted filtering is performed on the channel state information sampling sequence within the sliding sampling window containing the N79 time-division duplex uplink transmit time slot. The smoothed channel quality assessment value is calculated by using the sum of the products of the instantaneous signal-to-noise ratio corresponding to the unblocked sample and the sample weighting coefficient as the numerator, and the sum of the sample weighting coefficients corresponding to the unblocked sample and the denominator correction factor as the denominator. The denominator correction factor is a preset, very small positive constant greater than zero, used to prevent division-by-zero overflow anomalies. When all samples within the sliding sampling window are blocked due to interference, the downgrade action of the modulation and coding strategy is frozen.
[0011] Closed-loop dynamic resource scheduling and state recovery are implemented. When the noise floor rise exceeds the tolerance threshold, the Wi-Fi system-on-a-chip sends a power back-off request to the 5G cellular baseband processor. The 5G cellular baseband processor performs back-off based on the target uplink transmit power, which is obtained by subtracting the product of the closed-loop adjustment coefficient and the out-of-range noise floor rise from the current N79's actual uplink transmit power. When the obtained target uplink transmit power is lower than the safety critical transmit power required to maintain minimum modulation and demodulation, the safety critical transmit power is forcibly used as the actual output. Meanwhile, the Wi-Fi system-on-a-chip (SoC) reads the time-division duplex frame structure and predicts the arrival time boundary of the next high-power uplink transmission slot. When the data aggregation transmission opportunity to be initiated overlaps with the predicted next high-power uplink transmission slot on the time axis, the Wi-Fi SoC splits the aggregation frame at the logic layer, so that the split current data transmission segment and its corresponding necessary acknowledgment interaction end before the arrival of the predicted high-power uplink transmission slot. During the high-power uplink transmission slot, the initiation of new data aggregation transmission opportunities is suspended, and the link is put into a low-power reception mode during the interference period to achieve active peak shifting. After the high-power uplink transmission slot ends, normal transmission scheduling is restored.
[0012] The system state recovery determination is performed based on hysteresis parameters. The arithmetic mean of the noise floor rise from each acquisition within the sliding time window is calculated to obtain a smoothed noise floor evaluation value. When there are no valid samples due to radio frequency sleep within the sliding time window, the computation stream is actively intercepted and the recovery process is suspended, maintaining the current high protection state. When the smoothed noise floor evaluation value is lower than a preset recovery release threshold for two consecutive monitoring cycles, the Wi-Fi system-on-a-chip cancels the sent sampling masking command, sends a power-limited state release information to the 5G cellular baseband processor, restores normal communication settings, and clears internal interference fingerprint data exceeding the effective time window. The recovery release threshold is lower than a preset tolerance threshold.
[0013] This invention employs a combination of cross-chip time-domain synchronization and differential measurement of receiver quality from the same source to establish an internal interference fingerprint. The Wi-Fi system-on-a-chip (SoC) utilizes the coexistence management bus to obtain the N79 time-division duplex uplink time slot boundary. Without altering the Wi-Fi RF receiver center frequency and DFS RF detection settings, it performs time-domain selection on existing receiver quality monitoring results to obtain the noise floor variation of the N79 uplink transmission period relative to adjacent non-uplink reference periods. Since the background noise measurements used for comparison originate from the same Wi-Fi receiver quality measurement path, this scheme uses the relative noise floor variation, rather than the absolute leakage power of N79, as the internal mutual interference characterization quantity, thereby reducing the impact of gain deviations between different chips and absolute power conversion on interference identification.
[0014] Under the forced DFS channel, the DFS scanning engine always maintains standard radar detection functionality. This scheme extracts the burst pulse peak sequence and calculates the Pearson correlation coefficient with the N79 transmit power. For interference identified as internal artifacts, the underlying driver directly triggers a hardware interrupt rejection signal to intercept the channel switching command. This avoids the huge overhead of reconfiguring proprietary RF hardware and reduces the interference of candidate pulses formed by N79 crosstalk within the device on the DFS decision process.
[0015] By combining the packet loss timestamp with the internal interference fingerprint matrix, channel state information samples affected by interference are masked during the rate adaptation stage. A denominator correction factor is introduced to maintain the numerical stability of the filtering calculation. Furthermore, the downsampling action is frozen during full masking to prevent unexpected degradation of the modulation and coding strategy due to periodic time-division duplex interference. Through closed-loop power adjustment and predictive frame splitting of transmission slots, out-of-band leakage energy is limited at the transmitter, and a low-power receive mode is used at the Wi-Fi protocol scheduling end to avoid periods of high mutual interference overlap, thus maintaining the communication link throughput.
[0016] This invention provides a method for suppressing interference between the N79 band and the Wi-Fi 5G band in 5GCPE. It has the following beneficial effects: 1. This invention utilizes the N79 time-division duplex frame structure information provided by the coexistence management bus to establish a hardware time correlation between the reception quality monitoring results generated by the Wi-Fi receiving link and the N79 uplink transmission time slot. It calculates the relative noise floor rise by using the noise floor measurements obtained from the same receiving measurement path during the N79 uplink transmission period and adjacent non-uplink reference periods. Since it directly reuses the continuously operating automatic gain control and noise floor monitoring circuitry within the Wi-Fi network to obtain the instantaneous difference, it eliminates the need to wake up and reset the dedicated hardware chain containing the mixer, local oscillator, and analog-to-digital converter. It also eliminates the need to reconfigure the DFS scanning engine as an out-of-band leakage measurement device or establish a complex end-to-end power conversion calibration model. This mechanism not only significantly reduces system complexity, power consumption, and startup latency, but also ensures that the DFS scanning engine maintains its standard radar detection function under forced DFS channel conditions. Furthermore, it reduces misjudgments caused by internal N79 crosstalk by leveraging cross-chip timing and internal interference fingerprint information, while ensuring that the standard DFS radar decision results have processing priority.
[0017] 2. This invention utilizes an internal interference fingerprint matrix generated from the noise floor rise and timestamps to perform time-domain comparison of packet loss during packet acknowledgment timeout detection. When packet loss is confirmed to be caused by internal interference, a sampling masking command is proactively sent to the rate adaptation microcontroller, preventing the corresponding abnormal samples from participating in the modulation and coding strategy downsizing statistics, rather than unconditionally locking the entire rate control state machine. Furthermore, weighted filtering is applied to the channel state information sampling sequence. This cross-layer collaborative mechanism eliminates abnormal channel quality assessment values caused by periodic interference, while allowing unaffected samples to continue reflecting real spatial channel changes, preventing unexpected degradation of Wi-Fi transmission rates.
[0018] 3. This invention adopts a closed-loop dynamic resource scheduling mechanism. When interference exceeds the limit, it not only triggers the backoff of the transmission power on the 5G side to limit out-of-band leakage energy, but also predicts the arrival time of the next high-power uplink transmission slot on the Wi-Fi side based on the time-division duplex frame structure. By actively splitting the data aggregation frame before the time axis overlaps and pausing the initiation of new aggregation transmission opportunities during the high-power uplink transmission slot, it achieves active avoidance of transmission timing, reduces the probability of data packets and their acknowledgment interactions being subject to internal interference during dual-frequency concurrency, and improves the overall coexistence efficiency of the system. Attached Figure Description
[0019] Figure 1 This is a schematic diagram of the system architecture of the present invention; Figure 2 This is a flowchart of the method of the present invention; Figure 3 This is a comparison chart of the throughput of the media access control layer application of the present invention; Figure 4 This is a comparison diagram of the modulation and coding strategies of the present invention.
[0020] Among them, 110 is the 5G cellular baseband processor; 120 is the Wi-Fi system-on-a-chip; 121 is the DFS scanning engine; 122 is the rate adaptation microcontroller; 123 is the quality monitoring unit; 130 is the coexistence management bus; 140 is the N79 RF front-end link; 141 is the first bandpass filter; 142 is the N79 antenna; 150 is the Wi-Fi RF front-end link; 151 is the second bandpass filter; and 152 is the Wi-Fi antenna. Detailed Implementation
[0021] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0022] Please see the appendix Figure 1 The present invention provides a system for suppressing the mutual interference between the N79 band and the Wi-Fi 5G band in 5G CPE. The mutual interference suppression system includes a 5G cellular baseband processor 110 and a Wi-Fi system-on-a-chip 120.
[0023] The 5G cellular baseband processor 110 and the Wi-Fi system-on-a-chip 120 are connected via a coexistence management bus 130. The coexistence management bus 130 is a packet flow arbitration interface bus or a PCIe serial bus, or other inter-chip coexistence interface that can transmit frame structure timing information, carrier status and transmit power information. The coexistence management bus 130 is used to transmit time-division duplex frame structure synchronization information and carrier scheduling signaling.
[0024] The 5G cellular baseband processor 110 is connected to an N79 RF front-end link 140. A first bandpass filter 141 is connected in series on the N79 RF front-end link 140. The first bandpass filter 141 is an FBAR bandpass filter dedicated to the N79 frequency band. The back end of the first bandpass filter 141 is connected to the N79 antenna 142.
[0025] The Wi-Fi system-on-a-chip 120 is connected to a Wi-Fi radio frequency front-end link 150, and a second bandpass filter 151 is connected in series on the Wi-Fi radio frequency front-end link 150. The rear end of the second bandpass filter 151 is connected to the Wi-Fi antenna 152.
[0026] The Wi-Fi system-on-a-chip 120 integrates a DFS scanning engine 121 and a rate adaptation microcontroller 122, and includes a reception quality monitoring unit 123. The reception quality monitoring unit 123 is a functional unit associated with the Wi-Fi physical layer reception quality measurement path. It can read at least one of the following information generated during normal Wi-Fi reception: received power estimation, background noise estimation, signal-to-noise ratio estimation, and automatic gain control status. The reception quality monitoring unit 123 can be implemented using existing physical layer measurement logic and register read logic within the Wi-Fi system-on-a-chip 120, without requiring a separate RF front-end, mixer, local oscillator, or analog-to-digital converter independent of the Wi-Fi RF front-end link 150.
[0027] Please see the appendix Figure 2 This invention provides a method for suppressing mutual interference between the N79 band and the Wi-Fi 5G band in a 5GCPE. This method is based on the aforementioned mutual interference suppression system and specifically includes the following steps: Step S100: Initialize the physical layer static isolation configuration; after the mutual interference suppression system is powered on, load the setting parameters of N79 RF front-end link 140 and Wi-Fi RF front-end link 150. The mutual interference suppression system sets N79 antenna 142 to work in the vertical polarization radiation direction and sets Wi-Fi antenna 152 to work in the horizontal polarization radiation direction. The mutual interference suppression system 100 establishes RF out-of-band attenuation suppression through the first bandpass filter 141 and the second bandpass filter 151, and establishes spatial electromagnetic isolation through the antenna polarization direction.
[0028] Step S200: Perform cross-chip state synchronization and risk state determination. The 5G cellular baseband processor 110 sends the current N79 carrier's center frequency, allocated bandwidth, actual uplink transmit power, and time-division duplex frame structure synchronization information to the Wi-Fi system-on-a-chip 120 via the coexistence management bus 130. The Wi-Fi system-on-a-chip 120 obtains the current Wi-Fi channel's boundary frequency. The Wi-Fi system-on-a-chip 120 calculates the dynamic guard distance between the N79 carrier and the Wi-Fi channel. When the guard distance is less than a preset safe distance threshold and the actual uplink transmit power is greater than a preset transmit power trigger threshold, the mutual interference suppression system is determined to enter a high mutual interference risk state.
[0029] Step S300 involves performing receive noise monitoring, internal interference fingerprint generation, and DFS-assisted event discrimination based on cross-chip time-domain synchronization. Upon determining that a high mutual interference risk state has been entered, the Wi-Fi system-on-a-chip 120 establishes the local time boundary of the N79 uplink transmission slot based on the N79 time-division duplex frame structure synchronization information transmitted by the coexistence management bus 130, and controls the receive quality monitoring unit 123 to read the background noise estimation results of the N79 uplink transmission slot and adjacent non-uplink reference periods to obtain the relative noise floor rise. The system correlates the noise floor rise with the corresponding N79 uplink transmission slot timestamp, slot duration, and actual uplink transmission power to form an internal interference fingerprint matrix. For non-DFS channels, the DFS scanning engine 121 remains in its original non-operating state and is not additionally woken up due to internal interference monitoring; for forced DFS channels, the DFS scanning engine 121 maintains the standard DFS settings to perform radar detection. The system only uses the internal interference fingerprint as auxiliary correlation information for radar candidate events and does not cover the standard DFS radar decision logic's judgment of valid radar events.
[0030] Step S400: Perform rate adaptation sampling masking based on interference fingerprints. The Wi-Fi system-on-a-chip 120 monitors data packet acknowledgment timeout events at the MAC layer. When it detects that the packet loss timestamp coincides with the N79 uplink transmission time slot recorded in the internal interference fingerprint matrix, and the noise floor rise is greater than a preset tolerance threshold, the Wi-Fi system-on-a-chip 120 generates a sampling masking command. This sampling masking command is sent to the rate adaptation microcontroller 122. The rate adaptation microcontroller 122, according to the sampling masking command, prohibits the corresponding abnormal samples from participating in the modulation and coding strategy downsampling statistics, while continuing to use the unmasked valid channel state information samples to perform normal rate adaptation.
[0031] Step S500: Perform closed-loop dynamic resource scheduling and hysteresis state recovery. When the noise floor rise exceeds a preset tolerance threshold, the Wi-Fi system-on-a-chip 120 sends a power back-off request to the 5G cellular baseband processor 110 via the coexistence management bus 130; the 5G cellular baseband processor 110 adjusts the N79 uplink transmit power without falling below the safety critical transmit power. If necessary, the Wi-Fi system-on-a-chip 120 predicts the high-power uplink transmit time slot in advance based on the N79 time-division duplex frame structure and performs aggregated frame splitting and peak-shifting scheduling. When the smoothed noise floor evaluation value is lower than the recovery release threshold for two consecutive monitoring cycles, the Wi-Fi system-on-a-chip 120 cancels the sampling shielding state and sends a power-limited state release information, then clears internal interference fingerprint data exceeding the effective time window and restores normal communication settings.
[0032] Please see the appendix Figure 2The present invention provides implementation details of a method for suppressing mutual interference. During the initialization process of the mutual interference suppression system 100, the implementation of the physical layer static isolation settings mainly relies on the spectrum-selective filtering of the radio frequency link and the polar decoupling of the antenna spatial domain.
[0033] In this embodiment, the mutual interference suppression system establishes basic electromagnetic compatibility boundaries through hardware criteria. The specific execution logic and technical principles of step S100 are as follows: S101, establish the frequency domain filtering characteristics of the RF front-end link. The signal output by the 5G cellular baseband processor 110 enters the first bandpass filter 141 through the N79 RF front-end link 140. Considering the high transmit power of the N79 band and the possibility that its harmonics or spurious emissions may cover the Wi-Fi channel, the first bandpass filter 141 is used as a thin-film bulk acoustic resonant filter with high suppression characteristics. Based on the bulk acoustic waves generated by the piezoelectric effect oscillating within the crystal, the first bandpass filter 141 can obtain a steep out-of-band roll-off characteristic within a small frequency interval. In this embodiment, the passband of the first bandpass filter 141 is set to 4800MHz to 4960MHz, and its out-of-band attenuation for the Wi-Fi 5G band is set to be no less than 40dB. The value of this out-of-band attenuation is determined based on the difference between the maximum 5G transmit power and the Wi-Fi receiver blocking level, ensuring that strong signals do not cause the Wi-Fi RF front-end to enter the nonlinear region.
[0034] Meanwhile, the signal received by the Wi-Fi system-on-a-chip 120 enters the second bandpass filter 151 through the Wi-Fi RF front-end link 150. The second bandpass filter 151 acts as a pre-selection filter, with a passband for 5180MHz to 5825MHz. To prevent the main wave energy of the N79 band from directly entering the Wi-Fi receiving path, the out-of-band attenuation of the second bandpass filter 151 for the 4800MHz to 4960MHz band is set between 30dB and 40dB. Through the suppression at the transmitting end by the first bandpass filter 141 and the blocking at the receiving end by the second bandpass filter 151, the system establishes the first level of frequency domain isolation.
[0035] S102, implement polarization orthogonal decoupling in the antenna space. In physical space deployment, the polarization directions of the N79 antenna 142 and the Wi-Fi antenna 152 meet the orthogonality requirement. In this embodiment, the N79 antenna 142 adopts a vertical polarization mode, and the Wi-Fi antenna 152 adopts a horizontal polarization mode. In a practical engineering environment, considering the limitations of multipath reflection, structural coupling, and cross-polarization discrimination rate, the orthogonal arrangement can provide additional polarization isolation.
[0036] Under fixed antenna spacing, overall structure, and test attitude conditions, filter out-of-band attenuation, antenna polarization isolation, and structural coupling loss can be jointly evaluated as components of the overall isolation budget, without requiring different propagation mechanisms to be simply regarded as a linear superposition under ideal free space conditions.
[0037] In this embodiment, the typical range of polarization mismatch loss is 15dB to 20dB, which is determined based on the measured values of cross-polarization components in antenna isolation testing. By accumulating the physical indicators for different interference mechanisms, the system ensures that the mutual interference at the hardware level has been reduced to within a preset linear dynamic range before entering the software coordination process. Regarding the specific impedance matching logic and electromagnetic shielding packaging of the first bandpass filter 141 and the second bandpass filter 151, those skilled in the art can lay them out according to the specific board space constraints. Such impedance control and wiring shielding methods are well-known technologies in the field and will not be elaborated further here.
[0038] Based on the established physical isolation boundary, the mutual interference suppression system will further suppress residual interference through cross-chip signaling coordination. Based on the hardware environment established by the physical layer static isolation settings, the system achieves real-time awareness of communication conditions through cross-chip state synchronization and dynamic risk assessment logic. In this embodiment, the 5G cellular baseband processor 110 and the Wi-Fi system-on-a-chip 120 establish close signaling interaction through the coexistence management bus 130. The specific implementation of this part is as follows: S201, a cross-chip state synchronization mechanism based on hardware contacts is established. During the operation of the mutual interference suppression system, the 5G cellular baseband processor 110 and the Wi-Fi system-on-a-chip 120 maintain hardware-level time synchronization accuracy capable of distinguishing the boundaries of the N79 uplink transmission time slot, to ensure accurate subsequent interference localization for the time-division duplex uplink time slot. The 5G cellular baseband processor 110 extracts the system frame information of the current cellular network and periodically pushes the scheduling parameters of the N79 carrier to the Wi-Fi system-on-a-chip 120 through the coexistence management bus 130. In this embodiment, these scheduling parameters include the N79 carrier center frequency, N79 carrier channel bandwidth, N79 time-division duplex frame structure, corresponding uplink time slot boundaries, and the actual uplink transmission power for the corresponding time period.
[0039] After receiving the above parameters, the Wi-Fi system-on-a-chip 120 associates them with a local hardware timing reference, thereby determining whether a specific reception quality measurement sample corresponds to the uplink, downlink, or guard period of the N79. This synchronization is used to establish cross-chip event timing correlations and does not require the two heterogeneous processors to maintain identical system clocks across all internal logic cycles.
[0040] S202, perform real-time calculation of dynamic guard distance. The Wi-Fi system-on-a-chip 120 reads the operating parameters in the current RF driver to determine the lower edge frequency of the Wi-Fi channel. To determine the degree of spectral isolation between two frequency bands, the system needs to calculate the dynamic guard distance between them. In this embodiment, the calculation process is based on the physical boundary of the spectral envelope, which is physically equal to the lower edge frequency of the Wi-Fi channel minus the sum of the center frequency of the N79 carrier and half of the bandwidth of the N79 carrier channel. Before performing the subtraction operation, the mutual interference suppression system processes the calculation of half of the N79 carrier channel bandwidth through a preset fixed-point right shift operation logic to avoid kernel latency caused by floating-point operations. If the calculated dynamic guard distance is less than or equal to zero, it indicates that the N79 transmission spectrum and the Wi-Fi operating frequency band have physically overlapped. As a security protection mechanism, the mutual interference suppression system will directly determine that it has entered a high mutual interference risk state without performing subsequent power criterion comparison. This determination based on frequency domain causality can complete the algorithm preparation before physical interference occurs.
[0041] When the dynamic protection distance is negative or zero, it indicates that the two operating bandwidths have overlapped or touched according to the above spectrum boundary model; when the dynamic protection distance is positive, the smaller the value, the closer the edges of the two operating spectrums are. The system uses this distance indicator in conjunction with the actual N79 uplink transmit power for risk assessment, rather than directly activating all interference suppression measures based solely on the distance.
[0042] S203 executes multi-dimensional risk state determination logic. The determination process depends not only on the physical distance within the frequency band but also on the radiation intensity at the transmitting end. The mutual interference suppression system pre-stores safe frequency distance thresholds and transmit power trigger thresholds.
[0043] The safe frequency range threshold is not limited to a fixed range of 20MHz to 100MHz, but is calibrated based on the transition band attenuation of the first bandpass filter 141 and the second bandpass filter 151, the anti-blocking performance of the Wi-Fi receiver, antenna isolation, and the overall coexistence test results; the transmit power trigger threshold is set based on the output capability of the N79 RF front end and the tolerable blocking level of the Wi-Fi receiver.
[0044] The Wi-Fi system-on-a-chip 120 compares real-time parameters with the threshold. The mutual interference risk flag is set to active only when the dynamic protection distance is less than the safe distance threshold and the actual uplink transmit power of N79 is greater than the transmit power trigger threshold; otherwise, it remains inactive. This determination is a dual-condition joint determination logic.
[0045] By introducing the actual uplink transmission power of N79 as a decision factor, the system can avoid unconditionally executing the high-protection-level interference suppression process when N79 is in a low-power transmission state.
[0046] Wi-Fi receive links typically obtain receive power, automatic gain control status, or background noise estimation results at the physical layer. However, in normal operating mode, this receive quality information primarily serves local Wi-Fi receive control and lacks a definite temporal correlation with the N79 time-division duplex transmission timing of the 5G cellular baseband. Therefore, relying solely on regularly periodically read receive quality results cannot determine whether a particular noise floor anomaly is causally related to the device's internal N79 uplink transmission. This embodiment does not change the center frequency or sampling settings of the Wi-Fi receive RF link. Instead, it utilizes the N79 frame structure synchronization information provided by the coexistence management bus to perform time-domain marking and windowed reading of existing receive quality monitoring results, thereby establishing a temporal correlation between N79 uplink transmission behavior and changes in Wi-Fi receive noise floor.
[0047] The specific implementation method of this mechanism is as follows: S301, Establish a time-domain gating mechanism for receiving quality monitoring.
[0048] The Wi-Fi system-on-a-chip 120 reads the channel state table maintained by the current media access control layer and obtains whether the current working channel belongs to the channel that requires DFS to be performed.
[0049] Meanwhile, the Wi-Fi system-on-a-chip 120 utilizes the N79 time-division duplex frame structure synchronization information transmitted via the coexistence management bus 130 to generate a time-domain gating signal or time stamp corresponding to the N79 uplink transmit time slot in a local hardware timer or register. The time-domain gating signal is used for selecting and timing the receive quality measurement results, but not for switching the RF center frequency of the DFS scanning engine 121, nor does it require resetting the analog-to-digital converter, filter, or radar detection parameters of the DFS scanning engine 121.
[0050] The reception quality monitoring unit 123 acquires the background noise estimate within the measurement window corresponding to the uplink transmit time slot of N79, and acquires the reference background noise estimate within the adjacent non-uplink reference window. To reduce the boundary error caused by clock skew and transmit / receive state switching, the system can set a guard interval near the boundary of the uplink transmit time slot of N79 and avoid selecting measurement results located in the transient region of transmit / receive switching.
[0051] Only background noise measurement results that meet preset valid conditions are written into the internal interference fingerprint calculation process. For example, measurement results obtained during Wi-Fi local transmission, during demodulation of valid Wi-Fi physical layer protocol data units, or when the receive gain is saturated can be marked as invalid samples, thereby reducing the impact of normal Wi-Fi service signals on background noise estimation.
[0052] S302 executes branch detection logic based on channel regulatory attributes. Considering the compliance requirements of communication equipment, the mutual interference suppression system performs differentiated energy extraction and interception control for different types of channels. In non-DFS channels, radio regulations do not require the equipment to perform weather radar wave avoidance, and the DFS scanning engine 121 remains in its original non-operating state, not being additionally awakened due to internal interference monitoring. At this time, the Wi-Fi system-on-a-chip 120 directly reuses the receive quality monitoring unit 123 (which internally includes a received signal strength indicator and noise floor monitoring circuit for automatic gain control) to obtain the transient noise floor evaluation value in real time within the time window indicated by the time-domain gating signal. This process directly reads the underlying hardware measurement registers without waking up the DFS engine or resetting dedicated RF hardware links such as mixers, local oscillators, or analog-to-digital converters. In forced DFS channels, the equipment must maintain a compliant radar wave scanning thread. Since the high-power transmission of the N79 RF front-end link 140 is very likely to trigger radar false alarms in the DFS scanning engine 121, the mutual interference suppression system introduces a radar judgment rejection mechanism based on multi-dimensional comparison in this condition.
[0053] When the DFS scanning engine 121 captures a burst energy pulse, the Wi-Fi system-on-a-chip 120 extracts the timestamp of the burst energy pulse and compares it with the synchronously acquired uplink transmission time slot boundary in the time domain. If the timestamp falls entirely within the uplink transmission time slot, the mutual interference suppression system caches the peak values of multiple consecutive such burst energy pulses in memory to form a measured peak sequence, and extracts the actual uplink transmission power of N79 transmitted by the coexistence management bus 130 at the same time to form a control power sequence, and then calculates the Pearson correlation coefficient between the two sequences. When the calculated Pearson correlation coefficient is greater than the set correlation threshold, the system determines that the burst energy is an artifact caused by internal out-of-band blocking, rather than an external radar signal. In this embodiment, the correlation threshold is usually set between 0.8 and 0.9, and its specific value is determined based on the statistical distribution of a large number of measured interference samples, so as to avoid the risk of misjudgment caused by relying solely on a single energy extreme value. Based on this determination, the underlying driver of the Wi-Fi system-on-a-chip 120 triggers a hardware interrupt rejection signal. The rejection signal forcibly intercepts the channel switching command that is about to be executed, maintains the current communication link from being interrupted, and simultaneously extracts the background noise estimate value output by the receiving quality monitoring unit 123 in the corresponding time period as the measured background noise.
[0054] S303 quantifies the transient noise floor rise and generates an internal interference fingerprint matrix. After obtaining the measured noise floor, the mutual interference suppression system performs calculations to eliminate ambient noise. In the receiver noise model, the transient noise floor rise is the direct physical cause of signal-to-noise ratio degradation. Based on the principle of energy conservation, the system calculates the noise floor rise caused by the high-power transmission of N79 using the following formula. : ; In the formula, Indicates the amount of noise floor rise; Indicates the first The Wi-Fi reception noise floor output by the reception quality monitoring unit 123 during each N79 uplink transmission time slot; This represents the reference noise floor obtained from one or more non-N79 uplink time slots adjacent to this uplink time slot. To ensure the purity of the reference data, in this embodiment, The value is the background energy value that is pre-measured by the reception quality monitoring unit 123 at the current operating frequency during the idle time slot or downlink reception time slot of N79 and then processed by the moving average filtering.
[0055] The floor noise increase was calculated. Subsequently, the Wi-Fi system-on-a-chip 120 structurally binds this value to the corresponding time-division duplex uplink transmit time slot sequence. The system constructs a data structure in memory consisting of timestamps, time slot durations, and noise floor rise. An internal interference fingerprint matrix is constructed using multiple dimensions. Considering that continuous data accumulation may lead to matrix dimensionality expansion and memory overflow, the mutual interference suppression system uses a circular buffer structure to store this matrix and performs first-in-first-out overlay logic based on the timestamp's age. This internal interference fingerprint matrix transforms discrete radio frequency interference into structured data with temporal causality, providing deterministic data input for subsequent direct intervention in Wi-Fi rate adaptation protocols. Regarding the Fast Fourier Transform spectrum analysis and radar waveform feature matching algorithms involved in the DFS scanning engine 121 when performing standard radar scanning, those skilled in the art can implement them conventionally based on relevant mandatory radio communication standards. The underlying radar identification feature library is well-known technology in the field and will not be elaborated upon here.
[0056] Based on the transient noise floor characteristics recorded by the internal interference fingerprint matrix, a rate adaptation sampling masking mechanism is implemented in the mutual interference suppression system. In the underlying wireless communication protocol, conventional rate control algorithms typically attribute packet loss directly to channel fading or collisions, thus triggering blind rate reduction logic. To address the unnecessary rate reduction problem caused by misjudging internal pulse interference in Wi-Fi links, in this embodiment, the Wi-Fi system-on-a-chip 120 achieves algorithm immunity against N79 band interference by intervening in the input samples of the underlying rate control algorithm. The specific implementation of this mechanism is as follows: S401, perform temporal correlation matching between packet loss events and interference fingerprints. During communication, the Wi-Fi system-on-a-chip 120 continuously monitors the transmission status of data packets. When a transmission failure or reception acknowledgment timeout event is detected, the system's underlying driver captures the packet loss timestamp corresponding to the event. Based on the internal interference fingerprint matrix constructed above, the Wi-Fi system-on-a-chip 120 searches all uplink transmission time slots recorded in the matrix. In this embodiment, the system determines whether the packet loss has a physical causal relationship caused by high-power transmission by comparing whether the packet loss timestamp falls within the closed interval of any uplink transmission time slot. If the timestamp match is successful, it indicates that the communication failure occurred during the high-power radiation period of the N79 RF front-end link 140, meeting the preliminary conditions for interference.
[0057] S402 generates a sampling masking command based on multi-dimensional interference intensity. Simple time-domain overlap does not necessarily lead to demodulation failure. To avoid excessive interference with normal channel fluctuations, the Wi-Fi system-on-a-chip 120 needs to further incorporate interference intensity for weighted logic adjudication. The system extracts the noise floor rise corresponding to the matched uplink transmit time slot. The result is compared with a preset tolerance threshold, which is set between 6dB and 15dB. This threshold is determined based on the signal-to-noise ratio margin required by the current modulation and coding strategy to maintain the target bit error rate.
[0058] Based on the above time-domain and energy-domain determination results, the Wi-Fi system-on-a-chip 120 sends a sampling masking command to the rate-adaptive microcontroller 122 via its internal high-speed bus. This command includes a Boolean masking flag. In terms of logical generation rules, the masking is only applied when the packet loss timestamp falls within the uplink transmission time slot interval and simultaneously satisfies the noise floor rise requirement. When the interference exceeds the tolerance threshold, the shielding flag is set to the active state (assigned 1); otherwise, the shielding flag remains in the inactive state (assigned 0). The above judgment logic combines the temporal correlation of the interference with the energy destructiveness to prevent one-sided judgments caused by relying solely on a single time match.
[0059] S403, freeze the rate control state machine and perform sample weighted filtering. After receiving the instruction that the masking flag is 1, the rate adaptation microcontroller 122 immediately intervenes in the current rate adjustment logic. In this embodiment, the rate adaptation microcontroller 122 forcibly freezes the down-adjustment action of the modulation and coding strategy at the state machine level, thereby preventing the algorithm from reducing the current modulation order due to the detection of instantaneous packet loss. In addition, the system further performs weighted filtering on the sampling sequence of channel state information.
[0060] For a sliding sampling window that includes interference time slots, the rate-adaptive microcontroller 122 corrects the sampled samples, and its smoothed channel quality assessment value... The calculation logic follows as follows: ; In the above formula, This represents the smoothed channel quality assessment value; This indicates the total number of samples within the current sampling window; Indicates the first The masking flag corresponding to each sample; This represents the preset sample weighting coefficients, which are usually assigned different weights according to the order in which the samples were collected. Indicates the first The instantaneous signal-to-noise ratio corresponding to each sample; A pre-defined positive denominator correction factor, whose value is greater than zero and much smaller than the sum of the effective sample weights under normal circumstances, is introduced; this tiny constant is used. The technical purpose is to serve as a denominator correction factor. When all samples in the sampling window are disturbed and shielded under extreme conditions (i.e., the summation denominator was originally zero), it prevents the underlying processor from causing a division-to-zero overflow exception when performing division operations, thereby avoiding a division-to-zero exception when the sum of the effective sample weights is zero and improving the stability of numerical calculation.
[0061] Through the logical processing of this formula, the contribution of anomalous samples affected by N79 out-of-band leakage interference is forcibly set to zero. Since these failed samples are eliminated, the rate adaptation microcontroller 122 can maintain a stable assessment of the Wi-Fi link rate based on clean, non-interference time slot samples. As for the automatic repeat request mechanism and modulation and coding scheme mapping table lookup involved, the underlying rate switching state machine is a well-known technology in the field and will not be described in detail here.
[0062] Based on the rate-adaptive sampling shielding mechanism established at the receiving end, the system suppresses mutual interference and further performs closed-loop dynamic resource allocation at the transmitting end and the medium access control layer. In traditional heterogeneous communication coexistence schemes, absolute time-division multiplexing or physical frequency band isolation is usually adopted. This rigid cutoff mechanism is prone to serious waste of spectrum resources. To solve the above problems, in this embodiment, the system achieves active optimization of the space electromagnetic environment while ensuring service continuity through cross-chip multi-dimensional parameter feedback and hysteresis recovery logic. The specific implementation of this stage is as follows: S501 performs adaptive backoff of physical layer transmit power. During communication link operation, the Wi-Fi system-on-a-chip 120 continuously assesses the current channel impairment level and sends real-time low-noise status feedback to the 5G cellular baseband processor 110 via the coexistence management bus 130. Based on the physical radiation characteristics of radio waves, the out-of-band leakage power of the RF link exhibits a highly linear positive correlation with the main wave transmit power. To actively reduce spurious radiation, when the noise floor rises... Greater than the tolerance threshold At that time, the power control module inside the 5G cellular baseband processor 110 dynamically calculates the target uplink transmit power and sends it down to the underlying radio frequency driver. Its calculation logic follows the mathematical formula below: ; In the above formula, Indicates the target's uplink transmission power; This indicates the actual uplink transmit power of the N79; This represents the closed-loop control coefficient; Indicates the amount of noise floor rise; Indicates the tolerance threshold; closed-loop adjustment coefficient. The value is set between 0.5 and 1.0, and its specific value is determined based on the step resolution of the power amplifier inside the N79 RF front-end link 140. The physical purpose of introducing this parameter is to convert the noise floor degradation measured at the receiver into an equivalent amount of power reduction at the transmitter, thus achieving closed-loop feedback. In addition, to ensure the basic connectivity of the cellular network, the system calculates the target uplink transmit power. A lower limit clamping mechanism is implemented. Specifically, when the calculated value is lower than the safety critical transmit power required to maintain the minimum modulation and demodulation, the system forces the use of this critical power as the actual output to avoid the base station from determining that the equipment is offline due to excessive power reduction.
[0063] S502 executes time-domain truncation and peak-shifting scheduling at the media access control layer. When the aforementioned power back-off action is constrained by the lower limit clamp and still cannot suppress the noise floor within a safe range, the system needs to implement avoidance in the time dimension. The Wi-Fi system-on-a-chip 120 reads the transmitted time-division duplex frame structure and predicts in advance the arrival time boundary of the next high-power uplink transmission slot. Based on the causality of time overlap, if the data aggregation transmission to be initiated overlaps with the predicted uplink slot on the time axis, the Wi-Fi system-on-a-chip 120 will actively implement transmission truncation. At the logic layer, the system splits the originally continuous ultra-long aggregation frame, forcibly terminating the current transmission action before the transmission slot arrives, and puts the link into a low-power reception mode during the interference period. Through active peak shifting, the system effectively avoids the damage of strong interference to long packet data.
[0064] S503 performs system state recovery determination based on hysteresis parameters. As the physical location of the equipment moves or the external communication environment changes, internal interference sources may naturally weaken. To promptly release suppressed system resources and suppress mutual interference, the system performs background state recovery monitoring. To avoid single-measurement errors and frequent system state oscillations, the state determination logic uses the noise floor rise amount collected in each iteration within a sliding time window. The arithmetic mean is calculated to obtain the smooth noise floor evaluation value.
[0065] To ensure the integrity of the underlying algorithm, when there are no valid samples within the decision window due to radio frequency sleep, the underlying processor actively intercepts the computation flow and directly suspends the recovery process to maintain the current high protection state and prevent the system from crashing due to an empty window.
[0066] After successfully obtaining the smooth noise floor evaluation value, the system compares it with a preset recovery release threshold. In this embodiment, the recovery release threshold is set slightly lower than the aforementioned tolerance threshold. The two typically maintain a fixed difference of 3dB to 5dB; this difference range logically constitutes a hysteresis protection band against Gaussian white noise fluctuations. Only when the smoothed noise floor assessment value is consistently below this recovery release threshold for two consecutive monitoring cycles does the Wi-Fi system-on-a-chip 120 finally determine that the mutual interference has been substantially resolved. Subsequently, the system cancels the sent sampling masking command and allows the 5G cellular baseband processor 110 to exit restricted mode, thereby safely and smoothly completing the closed-loop control workflow.
[0067] Specific application example: 5G CPE home high-load concurrency scenario Application scenario settings: Users use the 5GCPE device provided in this embodiment of the invention at home. The CPE connects to the operator's base station via the 5GN79 frequency band (center frequency 4900MHz, bandwidth 100MHz), and simultaneously provides 4K video streaming services to the user's smart TV within the local area network via the Wi-Fi 5G frequency band (operating on channel 36, center frequency 5180MHz, bandwidth 80MHz).
[0068] At this time, the user initiated an upload task to the cloud for a very large file (such as local NAS backup data) on another computer on the intranet, causing the 5GN79 band to enter a continuous uplink transmission state with high duty cycle and full power (23dBm).
[0069] System collaborative processing flow: Step S100 (Construction of Physical Isolation Boundary): The first bandpass filter (FBAR) of the RF front end provides 40dB of out-of-band attenuation for the Wi-Fi band, and the N79 vertically polarized antenna and the Wi-Fi horizontally polarized antenna provide 18dB of spatial isolation. However, due to the N79 transmitting at full power (23dBm), the noise floor leaking into the Wi-Fi receiver still exceeds the linear tolerance range.
[0070] Step S200 (Risk State Synchronization Determination): The Wi-Fi system-on-a-chip obtains the current center frequency of N79 (4900MHz) through the coexistence management bus and calculates the dynamic protection distance as: 5140MHz (Wi-Fi lower edge) - [4900MHz + 50MHz] = 90MHz. This distance is less than the safe distance threshold (100MHz), and the transmit power (23dBm) is greater than the trigger threshold (15dBm), so the system immediately activates the high mutual interference risk state.
[0071] Step S300 (Interference Fingerprint Matrix Generation): The Wi-Fi system-on-a-chip (SoC) obtains the N79 uplink time slot boundary based on the coexistence management bus and performs time-domain labeling on the noise floor measurement results output by the internal receive quality monitoring unit. The Wi-Fi receive noise floor measured within the N79 uplink transmit time slot is... 78dBm, with a reference noise floor measured during the adjacent non-uplink reference period. 90dBm, therefore the noise floor increase is obtained. =12dB. The system writes the start time, duration, 12dB noise floor rise, and actual uplink transmit power of N79 in the internal interference fingerprint matrix for the corresponding uplink time slot of N79.
[0072] Step S400 (Rate Sampling Masking): During the strong transmission of N79, the video data packets sent by Wi-Fi to the smart TV experienced acknowledgment timeouts (packet loss). The underlying driver compared the timestamps and found that they perfectly overlapped with the interference fingerprint matrix, and the 12dB rise was greater than the tolerance threshold (8dB). At this time, the system intercepted the blind rate reduction logic of the traditional Wi-Fi protocol, froze the MCS (Modulation Coding Strategy) downsampling, maintained a high rate of 256-QAM, and only discarded the CSI quality assessment samples of these interfered time slots.
[0073] Step S500 (Dynamic Scheduling and Recovery): To further reduce the retransmission rate, adaptive power backoff is triggered internally within the CPE. The baseband processor actively reduces the transmit power of the N79 edge RB by 3dB. Simultaneously, Wi-Fi implements off-peak scheduling at the MAC layer to avoid sending excessively long aggregated frames in the N79 uplink time slot. A few minutes later, the user's upload task ends, the N79 power drops, the noise floor smoothing evaluation value falls below the recovery release threshold, the system cancels all blocking commands, and seamlessly returns to normal operation. No buffering or stuttering occurred during the entire 4K video playback.
[0074] Experimental verification and effect comparison: To verify the effectiveness of the present invention, comparative experiments were conducted in a microwave anechoic chamber and a test environment simulating real flow load.
[0075] Test conditions: N79 band full power (23dBm) continuous TDD uplink transmission; Wi-Fi 5G operates at Ch36, carrying downlink UDP packet injection test.
[0076] Comparison objects: Control group (traditional solution): Only physical layer filtering isolation is provided, and cross-chip signaling coordination and MCS sampling shielding mechanism are not enabled.
[0077] Experimental group (the present invention scheme): The S100-S500 collaborative suppression of mutual interference logic in the fully enabled embodiment is used.
[0078] Comparison of results and analysis of accompanying figures: Comparison of Modulation-Coding Scheme (MCS) Stability (with Appendix) Figure 4 analyze): Please see the appendix Figure 4 (Modulation and Coding Strategy Comparison Chart) This chart visually reflects the dynamic changes in the underlying communication rate level. The horizontal axis represents the test time (unit: seconds), indicating the total time of the experiment, from zero to ten seconds; the vertical axis represents the strategy order (range: zero to nine), reflecting the chip's underlying noise immunity and coding complexity. A higher order (e.g., nine) indicates a larger amount of data that can be carried in a single transmission; a lower order indicates that the system considers the environment extremely harsh and must use the most basic coding to send packets to ensure no connection drops. The light gray shaded area in the chart represents the period of strong interference injection in the cellular network, i.e., the period of high-power transmission within the device (second to eighth second).
[0079] For the control group (traditional protocol, i.e., the dark gray dashed line with square markings in the figure): When encountering high-power pulses that cause packet loss, traditional chips rely on rigid algorithms to determine that the distance is too far or the signal is fading. They forcibly pull the high-order (ninth-order) signal down to an extremely low tensile strength (first-order or second-order), resulting in unnecessary speed reduction. In actual testing, the MCS drops precipitously from the highest 9th order (256-QAM) to the 2nd order or even the 1st order (QPSK), and cannot climb back up during interference.
[0080] For the experimental group (the scheme of this invention, i.e., the solid black line with diamond-shaped markings in the figure): This patent freezes the degradation logic (blocking false positives) by identifying the interference fingerprint matrix in step S400. The system accurately identifies that packet loss is caused by internal interference rather than poor signal strength, thus forcibly maintaining the strategy order at a high level of eighth or ninth order. In actual testing, the MCS remained stable at 8th-9th order throughout, completely avoiding the pseudo-speed reduction effect and ensuring high-quality transmission of the underlying link.
[0081] Comparison of application throughput at Media Access Control (MAC) layer (with appendix) Figure 3 analyze): Please see the appendix Figure 3 (Media Access Control Layer Application Throughput Comparison Chart) This chart visually reflects the actual data transmission rate of a wireless LAN. The horizontal axis represents the test time (in seconds), from zero seconds to ten seconds; the vertical axis represents the throughput (in megabits per second), with higher values indicating smoother network performance. The light gray shaded area in the chart also represents the period of strong interference injection in the cellular network (from the second to the eighth second), during which the cellular communication module in the device transmits data at full power, causing severe electromagnetic noise floor increase interference to the LAN.
[0082] For the control group (traditional protocol, marked with a dark gray dashed line and a cross in the figure): This represents the performance of devices that do not employ the collaborative mechanism of this invention. When interference occurs, the system misjudges it as channel fading, causing the throughput to plummet from the baseline of 850 megabits per second to below approximately 150 megabits per second. This results in severe stuttering during video playback or downloads, significantly impacting the user experience.
[0083] For the experimental group (the scheme of this invention, i.e., the solid black line with a circle in the figure): This demonstrates the performance of devices utilizing the shielding and scheduling technology described in this patent. Even during periods of strong interference, the system maintains a high and stable throughput level, consistently above approximately 650 megabits per second, through staggered scheduling in steps S400 and S500 and underlying fingerprint shielding (staggering only during interfered time slots). Compared to traditional solutions, the overall system capacity is increased by approximately 333%, fully guaranteeing the user experience.
Claims
1. A method for suppressing interference between the N79 band and the Wi-Fi 5G band in 5GCPE, characterized in that, Includes the following steps: Initialize the physical layer static isolation settings; Synchronize the N79 carrier parameters, the actual uplink transmit power of N79, and the synchronization information of the N79 time-division duplex frame structure, calculate the dynamic guard distance, and determine the high mutual interference risk state by combining the actual uplink transmit power of N79. Under the high mutual interference risk state, a time-domain gating signal corresponding to the uplink transmission time slot of N79 is generated according to the synchronization information of the time-division duplex frame structure. The background noise estimation value within the uplink transmission time slot of N79 and the adjacent non-uplink reference time period is obtained by the reception quality monitoring unit. The noise floor rise is calculated and an internal interference fingerprint matrix is generated. Compare the packet loss timestamp with the internal interference fingerprint matrix, block abnormal channel quality samples caused by internal mutual interference and prevent them from triggering modulation and coding strategy downgrading, while valid channel quality samples that are not blocked continue to participate in rate adaptation. Based on the noise floor rise, a power back-off request is sent to the 5G cellular baseband processor, and based on the time division duplex frame structure synchronization information, the arrival time boundary of the next high-power uplink transmission time slot is predicted. For Wi-Fi data aggregation transmissions with overlapping time sequences, frame splitting and peak-shifting scheduling are performed.
2. The method for suppressing interference between the N79 band and the Wi-Fi 5G band in 5GCPE according to claim 1, characterized in that, Initializing physical layer static isolation settings includes: The N79 antenna (142) is configured to operate in the vertical polarization radiation direction, and the Wi-Fi antenna (152) is configured to operate in the horizontal polarization radiation direction. Signal transmission and reception are performed using an N79 RF front-end link (140) with a first bandpass filter (141) connected in series internally. The passband of the first bandpass filter (141) is set to 4800MHz to 4960MHz. Signal transmission and reception are performed using a Wi-Fi RF front-end link (150) with a second bandpass filter (151) connected in series internally, the passband of which is set to 5180MHz to 5825MHz.
3. The method for suppressing mutual interference between the N79 band and the Wi-Fi 5G band in 5GCPE according to claim 1, characterized in that, The synchronization of N79 carrier parameters, actual N79 uplink transmit power, and N79 time-division duplex frame structure synchronization information are used to calculate the dynamic guard distance. The determination of entering a high mutual interference risk state, based on the N79 actual uplink transmit power, includes: The 5G cellular baseband processor (110) sends the current N79 carrier center frequency, N79 carrier channel bandwidth, N79 actual uplink transmit power and N79 time division duplex frame structure synchronization information to the Wi-Fi system-on-a-chip (120) through the coexistence management bus (130). The Wi-Fi system-on-a-chip (120) acquires the lower edge frequency of the current Wi-Fi channel; The dynamic protection frequency distance is the difference between the lower edge frequency of the Wi-Fi channel and the sum of the center frequency of the N79 carrier and half the bandwidth of the N79 carrier channel. When the dynamic protection frequency distance is less than the preset safe frequency distance threshold and the actual uplink transmission power of N79 is greater than the preset transmission power trigger threshold, it is determined to enter a high mutual interference risk state.
4. The method for suppressing mutual interference between the N79 band and the Wi-Fi 5G band in 5GCPE according to claim 1, characterized in that, Under the high mutual interference risk state, the process of generating a time-domain gating signal corresponding to the N79 uplink transmission time slot based on the synchronization information of the time-division duplex frame structure, obtaining background noise estimates within the N79 uplink transmission time slot and adjacent non-uplink reference periods using the reception quality monitoring unit, calculating the noise floor rise, and generating an internal interference fingerprint matrix includes: The Wi-Fi system-on-a-chip (120) uses the coexistence management bus (130) to obtain the N79 time-division duplex frame structure synchronization information and generates a time-domain gating signal corresponding to the N79 uplink transmission time slot in the local hardware timer or register. The Wi-Fi system-on-a-chip (120) uses a reception quality monitoring unit (123) associated with the Wi-Fi physical layer reception quality measurement path to acquire a background noise estimate within the measurement window indicated by the time-domain gating signal; The receiving quality monitoring unit (123) obtains a reference background noise estimate in a non-uplink reference window adjacent to the uplink transmission time slot of the N79, and uses it as a reference noise floor. The difference between the estimated background noise value and the reference noise floor is determined as the noise floor rise amount corresponding to the N79 uplink transmission time slot; The timestamp of the corresponding uplink transmission time slot of N79, the duration of the time slot, the noise floor rise and the actual uplink transmission power of N79 are associated and stored to generate an internal interference fingerprint matrix. The receiving quality monitoring unit (123) reuses at least one of the information generated by the Wi-Fi system-on-a-chip (120) during normal reception, including the received power estimation, background noise estimation, signal-to-noise ratio estimation, and automatic gain control status. It does not require the setting of a radio frequency front-end, mixer, local oscillator, or analog-to-digital converter that is independent of the Wi-Fi radio frequency front-end link (150).
5. The method for suppressing mutual interference between the N79 band and the Wi-Fi 5G band in 5GCPE according to claim 4, characterized in that, Also includes: When the Wi-Fi operating channel is a forced DFS channel: The DFS scanning engine (121) maintains the standard DFS settings to perform radar detection. When a burst energy pulse is captured, the Wi-Fi system-on-a-chip (120) extracts the timestamp of the burst energy pulse and compares it with the N79 uplink transmission time slot boundary obtained synchronously in the time domain. If the occurrence timestamp falls entirely within the uplink transmission time slot of the N79, the peak values of multiple consecutive burst energy pulses are cached in memory to form a measured peak value sequence, and the actual uplink transmission power of the N79 transmitted by the coexistence management bus (130) at the same time is extracted to form a reference power sequence. Calculate the Pearson correlation coefficient between the measured peak sequence and the control power sequence; When the calculated Pearson correlation coefficient is greater than the set correlation threshold, the burst energy is determined to be an artifact caused by internal out-of-band blocking. The underlying driver triggers a hardware interrupt rejection signal to intercept the channel switching instruction, and simultaneously extracts the background noise estimate output by the receiving quality monitoring unit (123) in the corresponding time period as the measured background noise.
6. The method for suppressing mutual interference between the N79 band and the Wi-Fi 5G band in 5GCPE according to claim 1, characterized in that, The process of comparing packet loss timestamps with the internal interference fingerprint matrix, shielding abnormal channel quality samples caused by internal interference and preventing them from triggering modulation and coding strategy downgrading includes: Wi-Fi system-on-a-chip (120) monitors data packet acknowledgment timeout events at the MAC layer and detects packet loss timestamps; When the packet loss timestamp is detected to fall within any N79 time division duplex uplink transmission time slot closed interval recorded by the internal interference fingerprint matrix and the noise floor rise is greater than the preset tolerance threshold, the Wi-Fi system-on-a-chip (120) generates a sampling shielding command with an active state shielding flag and sends it to the rate adaptation microcontroller (122). After receiving the sampling masking command, the rate adaptation microcontroller (122) masks the abnormal channel quality samples in the corresponding N79 uplink transmission time slot from the modulation and coding strategy downsizing statistics, and prohibits the modulation and coding strategy downsizing from being triggered by the abnormal channel quality samples alone. The valid samples that are not masked continue to participate in the normal update of the rate adaptation state machine.
7. A method for suppressing interference between the N79 band and the Wi-Fi 5G band in a 5GCPE according to claim 6, characterized in that, The rate-adaptive microcontroller (122) performs weighted filtering on the channel state information sampling sequence within the sliding sampling window containing the N79 time-division duplex uplink transmit time slot to obtain a smoothed channel quality assessment value; The smoothed channel quality assessment value is calculated by using the sum of the products of the instantaneous signal-to-noise ratio corresponding to the unmasked sample and the sample weighting coefficient as the numerator, and the sum of the sample weighting coefficients corresponding to the unmasked sample and the denominator correction factor as the denominator. The denominator correction factor is a preset, very small positive constant greater than zero, used to prevent division by zero overflow anomalies. When all samples within the sliding sampling window are interfered with and blocked, the down-adjustment action of the modulation and coding strategy is frozen.
8. The method for suppressing mutual interference between the N79 band and the Wi-Fi 5G band in 5GCPE according to claim 1, characterized in that, Sending a power back-off request to the 5G cellular baseband processor based on the noise floor rise includes: When the noise floor rise exceeds the preset tolerance threshold, the Wi-Fi system-on-a-chip sends a power back-off request to the 5G cellular baseband processor (110); The 5G cellular baseband processor calculates the target uplink transmit power based on the noise floor rise exceeding the range value; The value for noise floor rise exceeding the range is the difference between the noise floor rise and the aforementioned tolerance threshold. The target uplink transmit power is obtained by subtracting the closed-loop adjustment coefficient from the product of the out-of-range noise floor rise value from the current actual uplink transmit power of N79; When the obtained target uplink transmission power is lower than the safety critical transmission power, the safety critical transmission power is used as the target output value.
9. A method for suppressing mutual interference between the N79 band and the Wi-Fi 5G band in a 5GCPE according to claim 1, characterized in that, The step of predicting the arrival time boundary of the next high-power uplink transmission time slot based on the synchronization information of the time-division duplex frame structure, and performing frame splitting and peak-shifting scheduling for Wi-Fi data aggregation transmissions with overlapping time sequences, includes: The Wi-Fi system-on-a-chip (120) reads the transmitted time-division duplex frame structure to predict the arrival time boundary of the next high-power uplink transmit time slot; When a data aggregation transmission opportunity to be initiated overlaps with the predicted next high-power uplink transmission slot on the timeline, the Wi-Fi system-on-a-chip (120) splits the aggregation frame at the logic layer, so that the split current data transmission segment and its corresponding necessary acknowledgment interaction end before the predicted high-power uplink transmission slot arrives; during the high-power uplink transmission slot, the initiation of new data aggregation transmission opportunities is suspended, and normal transmission scheduling is resumed after the high-power uplink transmission slot ends.
10. A method for suppressing mutual interference between the N79 band and the Wi-Fi 5G band in a 5GCPE according to claim 1, characterized in that, It also includes performing system state recovery determination based on hysteresis parameters: The smoothed noise floor assessment value is obtained by calculating the arithmetic mean of the noise floor rise during each acquisition within the sliding time window. When there are no valid samples due to radio frequency sleep within the sliding time window, the calculation stream is actively intercepted and the recovery process is suspended to maintain the current high protection state. When the smooth noise floor evaluation value is lower than the preset recovery release threshold for two consecutive monitoring cycles, the Wi-Fi system-on-a-chip (120) cancels the sent sampling shielding command and sends power-limited state release information to the 5G cellular baseband processor (110), restores normal communication settings, and clears internal interference fingerprint data that exceeds the effective time window; The recovery release threshold is lower than a preset tolerance threshold.
