A seamless handover method for power dual-mode communication link based on real-time SNR evaluation

Abstract

The application discloses a power dual-mode communication link seamless switching method based on real-time SNR evaluation, and relates to the technical field of power dual-mode communication.The power grid power frequency phase is locked, a statistical mapping table of phase and signal-to-noise ratio is constructed, and a normal distribution model is used to quantitatively predict the link degradation probability in the future millisecond, so that the transition from passive response to active early warning is realized.The application can sharply capture the noise characteristics near the voltage zero-crossing point by deeply mining the inherent periodicity of the power grid, and can change the original passive after-treatment into active early warning, effectively filter out the false alarm caused by instantaneous pulse interference, make the evaluation of the link state no longer limited to the sampling value at a certain isolated moment, but be based on the accurate grasp of the future trend of the channel, and thus greatly improve the intelligent level and anti-interference resilience of the communication decision, and provide more reliable channel situation awareness capability for the upper-layer service.

Description

Technical Field

[0001] This invention relates to the field of power dual-mode communication technology, specifically a method for seamless switching of power dual-mode communication links based on real-time SNR evaluation. Background Technology

[0002] The power line dual-mode communication unit (PLC) possesses two independent physical channels: high-speed power line carrier (PLC) and low-power radio (RF), and is widely used in electricity information collection systems. In existing dual-mode communication systems, after each frame of data is received, the physical layer calculates the subcarrier SNR and average SNR through training sequences and reports them to the MAC sublayer. The data link layer synchronizes the link quality information of each node across the entire network through mechanisms such as the channel quality field of the MPDU selection acknowledgment frame (SACK), the site capability entries of the standard beacon frame, and heartbeat detection. It can also perform link switching through management messages such as proxy change requests / acknowledgments, discovery list messages, and communication success rate reports.

[0003] However, existing link switching methods typically trigger hard switching based on a comparison of a single instantaneous SNR with a fixed threshold. That is, when the current link quality falls below the threshold, the current link is immediately disconnected and switched to another link. The PLC channel noise intensity of this method varies periodically with the zero-crossing point of the power frequency voltage (50Hz / 60Hz). Existing uniformly sampled SNR cannot distinguish between the "good power frequency phase zone" and the "zero-crossing noise zone," leading to frequent misjudgments and ping-pong switching. Traditional solutions often only respond passively after an SNR drop, exhibiting a lag. Furthermore, after the switching decision, the routing, time slots, and keys of the backup link need to be re-established through proxy change requests / confirmations, resulting in interruptions of tens to hundreds of milliseconds, which cannot meet the high real-time requirements of data acquisition services. In addition, existing technologies include schemes that predict link quality based on historical SNR statistics, but these schemes do not integrate the prediction process with the inherent periodicity of the power grid's power frequency phase. This causes their prediction models to fail to accurately capture the noise variation patterns unique to the power line channel and synchronized with the power frequency, thus easily leading to misjudgments near the zero-crossing point. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a seamless switching method for power dual-mode communication links based on real-time SNR evaluation, which solves the problems mentioned in the background.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a seamless handover method for power dual-mode communication links based on real-time SNR assessment, comprising the following steps: Step S1: Utilizing the independent reporting mechanisms of the power line carrier physical layer and the wireless physical layer in the dual-mode communication unit, the subcarrier SNR and average SNR are calculated after each data frame is received and pushed to the data link layer; simultaneously, the current power frequency phase angle is obtained through power frequency zero-crossing detection or preamble signal extraction. Establish the mapping relationship between phase angle and SNR; Step S2, the data link layer maintains a sliding time window, and... Divided into The mean SNR is calculated for each of the three equally spaced phase intervals. and standard deviation Construct and dynamically update phase SNR mapping table; Step S3, let the current time be... The current phase angle is , for The measured instantaneous SNR at time t. and The phases constructed from step S2 are respectively Found in the SNR mapping table Mean and standard deviation of SNR in the phase interval; set the forecast lead. According to the power frequency angular velocity Estimation Future phase angle corresponding to time The table can be consulted to obtain the result. Mean SNR of the phase interval and standard deviation Calculate the lower limit of the predicted SNR:

[0006] in Here is the confidence coefficient; then calculate the initial degradation probability:

[0007] in The standard normal cumulative distribution function is... The switching threshold is used; considering the presence of aperiodic pulse interference in the power line channel, which causes drastic fluctuations in the instantaneous signal-to-noise ratio, a measured value correction factor is introduced here. Used to dynamically adjust the prediction confidence level, measured value correction factor The calculation formula is:

[0008] in This is a numerical stability constant to prevent the denominator from being zero. The value of reflects the degree of deviation between the current measured value of the channel and the historical statistical mean, and a correction coefficient is further introduced. Output the final degradation probability:

[0009] Step S4, according to And the redundancy level of the predicted SNR decision: The sending end according to And predict SNR, select one from multiple preset redundancy levels as the current communication mode, the multiple redundancy levels include at least level 0, level 1 and level 2, which correspond to different link usage strategies and data protection strengths respectively; The sending end notifies the receiving end of the selected redundancy level via a management frame; After receiving a management frame, the receiving end determines whether it supports the redundancy level. If it does, it prepares to receive at that level; if it does not, it sends a downgrade request to the sending end. When the link quality meets the first preset condition, level 0 redundancy is adopted, and transmission is only carried out through the power line carrier link, while the wireless link goes into sleep mode. When the link quality meets the second preset condition, Level 1 redundancy is adopted. The same MAC frame is copied and transmitted simultaneously on both the power line carrier and wireless links. The receiving end removes duplicates according to the frame sequence number. When the link quality meets the third preset condition, a level 2 redundancy is adopted, and K original data packets are generated by Reed-Solomon encoding. Each fragment is sent interleaved on two links according to its parity index. Fragments with odd indices are sent via the PLC link, and fragments with even indices are sent via the RF link. The two links are triggered alternately in ascending order of fragment index; that is, the PLC link sends the first fragment while the RF link sends the second fragment, then the PLC link sends the third fragment while the RF link sends the fourth fragment, and so on. The receiving end collects any K fragments and then decodes and recovers the data. Step S5, when If the redundancy level remains below the degradation threshold corresponding to the current redundancy level, the threshold for downgrading from level 1 to level 0 is 0.2, and the threshold for downgrading from level 2 to level 1 is 0.5. After meeting the corresponding hysteresis time, the redundancy level is gradually reduced and eventually reverts to level 0 single-link mode. Steps S1 to S3 constitute a link quality prediction method based on power frequency phase synchronization, and steps S4 to S5 constitute a gradual soft handover method based on dual-mode parallel redundancy; both are achieved through... and Tightly coupled.

[0010] Furthermore, in step S1, after each frame of data is received, the power line carrier physical layer performs channel estimation using a preamble and a training sequence, calculates the subcarrier SNR and the 1-byte average SNR, and the average SNR value ranges from [value missing]. The generated reporting primitives are pushed to the MAC sublayer; the wireless physical layer calculates the average SNR of a single carrier and the entire frame based on the PPDU data packet LTF long training domain and reports it to the data link layer in real time.

[0011] Furthermore, in step S1, the power frequency phase angle The extraction method is as follows: Hardware zero-crossing detection circuits extract voltage zero-crossing points or parse the phase information built into the carrier preamble frame synchronization header; a local 16-bit phase counter provides high resolution. Each SNR sample is bound to a phase. Timestamp, sampling per wave no less than the preset number of times.

[0012] Furthermore, in step S2, the sliding window size Number of phase intervals A dynamic adaptive adjustment mechanism based on channel state is adopted; specifically: During the communication link initialization phase or the period of stable signal-to-noise ratio, reduce the number of phase interval divisions. The value is adjusted to reduce computational overhead and maintain a small sliding window size; when drastic fluctuations in the signal-to-noise ratio or sudden impulse noise interference are detected, the system automatically increases the number of phase interval divisions. To improve phase resolution, while increasing the sliding window size Expand to a preset maximum value to smooth the impact of extreme noise samples on statistical results; the sliding window stores the most recent... A complete power frequency cycle sampling data; number of phase interval divisions ; use The criteria remove outlier noise samples, and the Welford online recursive algorithm is used to update the interval mean and standard deviation in real time.

[0013] Furthermore, in step S3, the threshold is switched. Set the minimum demodulation SNR plus a 3dB margin based on the selected MCS modulation method.

[0014] Furthermore, in step S4, the first preset condition, the second preset condition, and the third preset condition are as follows: First presupposition condition: when and At that time, level 0 redundancy is used; Second presupposition condition: When or At that time, Level 1 redundancy is adopted; Third pre-set condition: When or At that time, two levels of redundancy are adopted; In Level 1 redundancy, the MAC layer assigns a unique frame sequence number (SN) to each service frame and adds a 1-byte redundancy flag field to the frame header: 0x01 indicates a single-link frame, 0x02 indicates a duplicate frame, and 0x03 indicates a fragmented frame. Two identical MPDUs are constructed and sent to the power line carrier and radio transmission queue respectively. The two physical layers transmit independently with a transmission time difference of no more than one OFDM symbol. The receiver maintains a receive buffer indexed by SN and discards duplicate frames directly.

[0015] Furthermore, in step S4, the level 2 redundancy employs a systematic Reed architecture. Solomon code ,Will Each fragment is marked as Power line carrier links transmit odd-index fragments, while wireless links transmit even-index fragments; the receiver collects arbitrary... After each different fragment is processed, RS decoding is performed immediately to recover the original data, with a decoding delay of no more than one OFDM symbol.

[0016] Furthermore, in step S5, the hysteresis protection rule specifically states that a downgrade from level 1 to level 0 requires meeting certain conditions. And it must remain stable for more than 200ms hysteresis; to degrade from level 2 back to level 1, the following must be met. And it remains stable for more than 200ms hysteresis; when Stable below 0.2 and When necessary, the wireless transmission channel is shut down, and the wireless physical layer is put into a low-power sleep mode; the threshold and hysteresis duration are determined based on the on-site channel environment.

[0017] Furthermore, in step S4, the negotiation of redundancy levels is achieved through a newly added management frame. This management frame reuses the existing MME frame type and extends the subtype fields, including: STA→PCO / CCO, carrying Value, current phase, and recommended redundancy level; PCO / CCO → STA, confirm the level change and allocate temporary shard ID space; It periodically reports the current redundancy level, link quality statistics, and phase information in both directions.

[0018] Furthermore, in step S4, the data link layer uses a message encapsulation format common to both power line carrier links and wireless links for transmission, and there is no need to reassemble packets during the switching process; the dual links use a unified sequence number and retransmission mechanism, and there is no need to retransmit the entire packet during switching, so that the upper layer services are unaware of the process.

[0019] This invention provides a seamless handover method for power dual-mode communication links based on real-time SNR evaluation, which has the following advantages: 1. This seamless handover method for dual-mode power communication links based on real-time SNR assessment fundamentally overcomes the persistent problems of handover lag and misjudgment caused by blindly relying on instantaneous signal quality in traditional power line communication. By deeply exploring the inherent periodic patterns of the power grid, it can keenly capture noise characteristics near the voltage zero-crossing point, transforming the original passive post-event remediation into proactive pre-event warning. This predictive logic based on statistical laws effectively filters out false alarms caused by instantaneous pulse interference, making the assessment of link status no longer limited to the sampled value at a certain isolated moment, but based on a precise grasp of the future trend of the channel. This greatly improves the intelligence level and anti-interference resilience of communication decisions, providing more reliable channel situational awareness capabilities for upper-layer services.

[0020] 2. This seamless handover method for dual-mode power communication links based on real-time SNR assessment fully utilizes the prediction results and completely eliminates the service interruption window in the traditional hard handover process. By converting the degradation probability of the prediction output into a hierarchical and progressive redundancy strategy, it realizes on-demand scheduling and dynamic matching of physical layer dual-channel resources, ensuring that the data protection strength always accurately matches the channel severity. More importantly, because the prediction module provides ample reaction time, the redundancy mechanism can easily establish a hot standby link in the background, avoiding the cumbersome signaling interaction and route reconstruction process in traditional handover. This ensures that even when the channel quality fluctuates drastically, upper-layer services can still enjoy seamless continuous data stream services. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the steps of a method for seamless switching of a power dual-mode communication link based on real-time SNR assessment according to the present invention. Detailed Implementation

[0022] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and should not be construed as limiting the scope of the invention.

[0023] like Figure 1 As shown, the present invention provides a technical solution: a seamless handover method for power dual-mode communication links based on real-time SNR evaluation, comprising the following steps: Step S1, Power frequency phase locking and SNR synchronous acquisition: Each station (STA, PCO, CCO) has a built-in zero-crossing detection circuit that extracts the power frequency voltage zero-crossing signal from the power line and generates a square wave pulse synchronized with the grid frequency. If dedicated hardware is unavailable, the power frequency phase information implicit in the frame synchronization header of the PLC preamble can be used. Each station maintains a 16-bit phase counter with a resolution of 0.1° (i.e., counting 3600 times per power frequency cycle). Each zero-crossing pulse resets the counter to 0, and then increments it by 1 every 1 / (50×3600)≈5.56μs. Current phase angle. All SNR sampling data include the current phase angle. and timestamp The SNR sampling frequency is preferably no less than 16 times per cycle to ensure phase stability. The real-time performance of the SNR mapping table. If the power grid frequency is 60Hz, then... The rest of the calculations are the same.

[0024] Simultaneously, after receiving each frame of data, the power line carrier physical layer performs channel estimation using preambles and training sequences, calculates the SNR and average SNR of all subcarriers, and generates primitives containing SNR arrays and average SNR values, which are then pushed to the MAC sublayer. The average SNR is represented by 1 byte, and its value ranges from [missing value]. 31 +31; The wireless physical layer calculates the single-carrier SNR and the average SNR of the entire frame based on the LTF long training domain of each PPDU data packet, and reports them to the data link layer in real time. The MAC sublayer stores the received SNR value in association with the current phase angle.

[0025] Step S2, Phase SNR dynamic mapping table construction: Each site maintains a circular buffer with a capacity equal to the nearest [number of sites]. The system requires sampling data for a complete power frequency cycle. Considering the time-varying characteristics of the power line channel, the system needs adaptive adjustment capabilities: during periods of channel stability, to reduce computational overhead, the sampling rate can be appropriately reduced. The value (e.g., reduced to 36, i.e., in 10° intervals); while during periods of channel abrupt change (e.g., when a continuous increase in packet loss rate is detected), it needs to be increased. To obtain more statistically significant samples, the initial configuration should be set to 40 (corresponding to 800ms); considering both real-time performance and accuracy, the recommended initial configuration is... (Corresponding to a 400ms window) =72, meaning each interval is 5°.

[0026] When initializing the mapping table, set 0° 360° divided into Each interval has an equally spaced phase interval. The statistical window contains the number of SNR samples falling within the specified interval and their arithmetic mean. and standard deviation .

[0027] Outliers with burst impulse noise are removed using the 3σ principle: if a sample deviates from the interval mean by more than 3 times the standard deviation, it is considered an outlier and is not included in the count; template updates use the Welford online recursive algorithm to avoid full recalculation and ensure real-time operation on embedded devices.

[0028] Other parameters The selection of phase resolution directly affects system power consumption: smaller phase resolution... It will reduce time synchronization accuracy, leading to increased phase jitter; larger While it can provide a finer phase-SNR mapping, it will increase by approximately The increased division operations result in higher power consumption; therefore, in practical deployments, the power consumption should be dynamically adjusted within the range of [36, 144] based on the crystal oscillator accuracy and battery capacity of the device. The value of .

[0029] Step S3, Short-term link quality prediction: Let the current time be The current phase angle is The measured instantaneous SNR is Query the mapping table to get Expectation of the interval and Set the lead time for forecasting. (Recommended 5s), based on power frequency angular velocity Estimation Phase angle at time 1 (like Then reduce ); query Expectation of the interval and standard deviation ; Take confidence coefficient Calculate the lower limit of the predicted SNR:

[0030] This value represents the future. The lowest possible SNR level at any given time is used for subsequent redundancy level decisions; Define the link quality degradation event as "predicted SNR is lower than the handover threshold". Then the degradation probability is:

[0031] in Let be the standard normal cumulative distribution function, if Then when hour Take 0 or 1, otherwise This formula transforms the statistical distribution of predicted SNR into a degradation probability, quantifying the risk of impending link failure. To further utilize the deviation between the current measured value and the expected value to correct the prediction reliability, a correction factor is introduced:

[0032] in This is a numerical stability constant to prevent the denominator from being zero. This reflects the degree of deviation between the current measured SNR and the historical statistical expectation. The greater the deviation, the higher the probability of non-periodic disturbances in the channel, and the lower the prediction reliability should be. Will Introducing corrections, the final degradation probability is output:

[0033] in, For the measured deviation correction factor, a value of 0.2 is recommended (this value of 0.2 is a specific constant determined through extensive field experiments on non-periodic disturbances to power lines, used to adjust the correction amplitude). When the deviation is large (the measured value deviates significantly from the expected value), The amplification makes the prediction more conservative when abnormal disturbances occur, avoiding false switching caused by non-periodic disturbances.

[0034] Steps S4-S5 specifically include the following processes: enter and A three-level redundancy strategy is defined, as shown in Table 1: Table 1. Schematic diagram of Level 3 redundancy strategy definition

[0035] A hysteresis protection mechanism is introduced to avoid frequent handovers caused by link quality fluctuations near a critical threshold. The hysteresis timer starts from the moment a link state transition occurs. Specifically: When the device is in a Level 1 redundancy state, the degradation probability is calculated in real time. If the value is less than 0.2, the system does not immediately perform a degradation operation. Instead, it starts a preferred hysteresis timer (with a duration set to 200ms). During the 200ms period, the system needs to continuously monitor... If the value is consistently below 0.2 (meaning the sample mean for each monitoring period must meet this threshold), it is permissible to downgrade from level 1 back to level 0. Similarly, when the equipment is in a level 2 redundancy state, if <0.5 System startup preferred hysteresis timer (duration set to 100ms); if within 100ms, If the value remains below 0.5, a downgrade to level 1 redundancy will be triggered. If during the hysteresis period If the rebound exceeds the aforementioned threshold, the current fluctuation is determined to be a transient disturbance, the timer is immediately reset, and the current redundancy level is maintained. The aforementioned hysteresis threshold and time can be determined based on on-site debugging.

[0036] The MAC layer assigns a unique sequence number (SN) to each service frame to be transmitted and adds a 1-byte redundancy flag field to the frame header: 0x01 indicates a single-link frame, 0x02 indicates a duplicated frame, and 0x03 indicates a fragmented frame. In Level 1 redundancy, the MAC layer constructs two identical MPDUs (containing the same SN and payload), which are sent to the transmission queues of the PLC and RF respectively. The two physical layers transmit independently, with a transmission time difference not exceeding one OFDM symbol (approximately 200 μs). The specific implementation of the SN management protocol is as follows: The value range of SN is defined as 1 to 255 (0x00 is reserved as an invalid identifier). Whenever the MAC layer receives data to be sent from the upper layer, it assigns the smallest SN value that is not currently in use. To prevent sequence number wrap-around conflicts in extreme cases, the high 4 bits and low 4 bits adopt a dynamic obfuscation algorithm based on timestamps to ensure the uniqueness of SN in distributed multi-node concurrent scenarios. When in Level 1 redundancy, the MAC layer constructs two identical MPDUs (containing the same SN and payload), which are sent to the transmit queues of the PLC and RF respectively; the two physical layers transmit independently, with the transmission time difference strictly controlled within one OFDM symbol (approximately 200μs), to ensure that the receiver can complete signal diversity combining using the short frame header gap. The MPDU in the sending queue only marks the SN as recyclable after receiving an explicit ACK confirmation from the receiver for that SN, and releases it back to the available pool after a preset aging time (e.g., 1 second). If no ACK is received within the aging time, the MAC layer triggers a timeout retransmission mechanism to keep the SN unchanged and re-enqueue it for transmission, avoiding ambiguity in the receiver's sequence number caused by rapid reuse of the SN.

[0037] The receiver maintains a receive buffer indexed by serial number (SN). When an MPDU is received, it checks whether the SN has already been received: if not, it stores the received frame in the buffer and submits it to the upper layer; if it has been received (from another link), it discards duplicate frames and records the difference in arrival delay between the two links for subsequent optimization.

[0038] Adopting a systematic Reed Solomon code ,recommend This means that every four original data packets are encoded into six fragments, allowing recovery even if any two fragments are lost. Encoding is performed at the MAC layer, without altering the upper-layer data. The six fragments are then marked as... ,in For the original data packet, For verification packets. Allocation rule: PLC link sends odd-indexed fragments. RF link transmits even-numbered index fragments This means that even if one link is completely interrupted, the other link can still provide at least three fragments (containing two original packets and one check packet), which is sufficient to recover all four original packets. The encoding and decoding latency is no more than one OFDM symbol, which is negligible compared to the power frequency cycle (10ms / 16.6ms), and the 50% redundancy overhead is far lower than the 200% overhead of the traditional three-replica mechanism. The receiving end collects fragments from both links and reassembles them according to their sequence numbers. When any fragment is received... After dividing the data into different segments, RS decoding is immediately performed to restore the original data. The first data packet is sent to the upper layer. Any remaining fragments are discarded upon arrival.

[0039] When step S3 outputs If the redundancy level remains below the degradation threshold corresponding to the current redundancy level (0.2 for downgrading from level 1 to level 0, and 0.5 for downgrading from level 2 to level 1) and the corresponding hysteresis time is met, a redundancy level degradation is triggered. For example, when downgrading from level 2 to level 1: sending check fragments stops. Only the original data packets are copied and transmitted on both links (equivalent to returning to level 1); the receiving end stops RS decoding and switches to simple deduplication. Stable below 0.2 and When this happens, the RF transmission channel is shut down, only the PLC single link is retained, and the RF physical layer is put into a low-power sleep mode; To negotiate redundancy levels, three types of management frames are defined (reusing existing MME frame types and extending subtype fields), as shown in Table 2. Table 2 Definitions of Three Types of Management Frames

[0040] The MMeRedundancyRequest frame structure is defined as shown in Table 3: Table 3 MMeRedundancyRequest Frame Structure Definition Table

[0041] The MMeRedundancyConfirm frame structure is defined as shown in Table 4: Table 4 MMeRedundancyConfirm Frame Structure Definition Table

[0042] The MMeRedundancyStatus frame structure is defined as shown in Table 5: Table 5 MMeRedundancyStatus Frame Structure Definition Table

[0043] Example of interaction flow: STA calculates through steps S1-S3 This triggers a Level 1 redundancy requirement; the STA sends an MMeRedundancyRequest (Level = 1). , Upon receiving the data, the PCO checks its own resources (whether the RF channel is idle) and replies with MMeRedundancyConfirm (level = 1, starting SN = 1000); both the STA and PCO simultaneously enable data replication mode, and subsequent frames will start with SN from 1000; when After the value drops to 0.15 and stabilizes for 200ms, the STA sends an MMeRedundancyRequest (level = 0); after confirmation by the PCO, both parties exit redundancy mode.

[0044] The redundancy level changes in steps S4-S5 indirectly affect the prediction model updates in steps S1-S3 through hysteresis protection. For example, when the system enters level 2 redundancy, the RF link is activated, but the SNR sampling of the PLC link continues. The prediction model operates independently of the redundancy level and is calculated solely based on the physical layer SNR data. When the redundancy level drops back to level 0, the RF link goes into hibernation, and the prediction model continues to run independently. This one-way dependency ensures the objectivity of the prediction, while hysteresis protection prevents positive feedback oscillations between prediction and decision.

[0045] This example demonstrates how predictive and redundancy mechanisms achieve seamless switching in a scenario where a STA encounters power frequency zero-crossing noise in the PLC link, leading to quality degradation. The power grid frequency is 50Hz, the zero-crossing interference period is 10ms, and the system's switching threshold is set accordingly. Lag time ; Steady-state monitoring phase (normal state): The STA operates in single-link mode by default, enabling only the PLC physical layer to transmit data while the RF physical layer remains in low-power sleep mode. Steps S1-S3 (prediction engine) run continuously in the background, monitoring the current grid phase in real time. The SNR was then calculated. At this point, the measured SNR stabilized above 25 dB, and the link degradation probability was output by the prediction model. If the redundancy level is maintained at a low level (0.1), steps S4-S5 (decision engine) determine that no redundancy is needed at present, and the system maintains a redundancy level of 0.

[0046] Risk warning and redundancy establishment phase (t=0ms~t=1ms): t=0ms (prediction triggered): Current phase (Close to zero crossing). The measured SNR is 25dB, but based on the historical phase... SNR mapping table, predicting that in 5ms ( When power frequency noise is superimposed, the SNR is expected to drop to 12dB (below the switching threshold). ); Apply formulas 1-4 in sequence: First, calculate the lower limit of the predicted SNR. Then calculate the initial degradation probability. Next, the correction factor for the measured value is calculated. Final output .

[0047] t=0.5ms (Decision and Signaling): The prediction engine receives... If the risk exceeds the preset risk threshold (0.5), it is determined that the system must enter Level 2 redundancy mode. The STA sends an MMeRedundancyRequest management frame to the PCO via the MAC layer. This frame carries the current... Value, phase And request level (Level 2).

[0048] t=1ms (Resource Allocation and Switchover): The PCO resolves the request, confirms sufficient channel resources, and replies with an MMeRedundancyConfirm frame, allocating a new temporary fragment sequence number (Starting SN=1000) within the frame. Upon receiving the confirmation, the STA activates the RF physical layer, and both parties synchronously enter Level 2 redundancy state. The MAC layer starts the RS(4,6) coding module.

[0049] During the interference-resistant transmission phase (t=5ms), the PLC link enters the zero-noise zone precisely on time, and the actual SNR drops sharply to 8dB, causing errors in some OFDM symbols transmitted by the PLC side. PLC link sends odd number of fragments The newly activated RF link sends even-numbered fragments. .

[0050] The receiving end (STA or PCO) receives 3 PLC fragments and 3 RF fragments. Due to the use of RS(4,6) erasure coding, the receiving end only needs to successfully decode any 4 fragments to recover the original data. In this example, all 4 original packets were successfully recovered and submitted to the upper-layer application, thus achieving the effect that despite the PLC link failure, the upper-layer business data flow remained completely uninterrupted.

[0051] Redundancy exit and single-link regression phase (t=15ms~t=35ms): t=15ms (state reversal): The PLC link leaves the zero-crossing zone, the SNR recovers to 28dB, and recalculation begins. The value quickly dropped to 0.1; t=15ms~t=35ms (Hysteresis protection determination): even though When the value drops to 0.1, the system will not immediately shut down the RF link; the prediction engine will detect this. It needs to remain consistently below the downgrade threshold (0.2) for a duration exceeding the hysteresis time. ).

[0052] t=35ms (Execute downgrade): Once the lag time has elapsed, it is determined that the link has been fully restored, triggering the exit process. First downgrade (Level 2 → Level 1): The MAC layer stops RS encoding, releases RF transmission resources, and switches to copying and transmitting the original data packet simultaneously on both the PLC and RF (although the RF is still transmitting, it is no longer logically included in the redundancy calculation).

[0053] Second downgrade (Level 1 → Level 0): After a short period of observation (to prevent vibration), confirmation. It remains at a low level; the STA sends an MMeRedundancyRequest (level=0), and after the PCO confirms, the STA closes the RF physical layer transmission channel, causing it to re-enter deep sleep mode; the system eventually returns to a steady state of single PLC link.

[0054] Throughout the entire process described above, the upper-layer services remain continuous and uninterrupted, and RF resources are consumed only when necessary.

[0055] In summary, this seamless switching method for dual-mode power communication links based on real-time SNR assessment fundamentally overcomes the persistent problems of switching lag and misjudgment caused by blindly relying on instantaneous signal quality in traditional power line communication. By deeply exploring the inherent periodic patterns of the power grid, it can keenly capture noise characteristics near the voltage zero-crossing point, transforming the original passive post-event remediation into proactive pre-event warning. This predictive logic based on statistical laws effectively filters out false alarms caused by instantaneous pulse interference, making the assessment of link status no longer limited to the sampled value at a certain isolated moment, but based on a precise grasp of the future trend of the channel. This greatly improves the intelligence level and anti-interference resilience of communication decisions, providing more reliable channel situational awareness capabilities for upper-layer services.

[0056] By fully utilizing the prediction results, the service interruption window in the traditional hard handover process is completely eliminated. By transforming the degradation probability of the prediction output into a hierarchical and progressive redundancy strategy, the on-demand scheduling and dynamic matching of physical layer dual-channel resources are realized. This ensures that the data protection strength is always accurately adapted to the channel severity. More importantly, because the prediction module provides ample reaction time, the redundancy mechanism can easily establish a hot standby link in the background, avoiding the cumbersome signaling interaction and route reconstruction process in traditional handover. This ensures that even when the channel quality fluctuates drastically, upper-layer services can still enjoy a seamless and continuous data stream service.

[0057] The embodiments of the present invention are given for illustrative and descriptive purposes only, and are not intended to be exhaustive or to limit the invention to the forms disclosed. Many modifications and variations will be apparent to those skilled in the art. The embodiments were chosen and described in order to better illustrate the principles and practical application of the invention, and to enable those skilled in the art to understand the invention and to design various embodiments with various modifications suitable for a particular purpose.

Claims

1. A seamless handover method for power dual-mode communication links based on real-time SNR assessment, characterized in that: Includes the following steps: Step S1: Utilizing the independent reporting mechanisms of the power line carrier physical layer and the wireless physical layer in the dual-mode communication unit, the subcarrier SNR and average SNR are calculated after each frame of data is received and pushed to the data link layer; at the same time, the current power frequency phase angle is obtained through power frequency zero-crossing detection or preamble signal extraction, and a mapping relationship between the phase angle and SNR is established. Step S2: The data link layer maintains a sliding time window, dividing 0°~360° into M equally spaced phase intervals, calculating the mean and standard deviation of SNR in each interval, and constructing and dynamically updating the phase-SNR mapping table. Step S3: Based on the measured instantaneous SNR, current phase angle and the phase-SNR mapping table at the current moment, calculate the future phase angle corresponding to the future prediction moment, look up the table to obtain the mean and standard deviation of SNR in the phase interval where the future phase angle is located, and calculate the lower limit of predicted SNR and the final degradation probability. Step S4: Based on the final degradation probability and the predicted SNR lower limit, select one from a set of multiple redundancy levels as the current communication mode. The multiple redundancy levels include at least level 0, level 1, and level 2, which correspond to different link usage strategies and data protection strengths, respectively. The sending end notifies the receiving end of the selected redundancy level through a management frame. After receiving the management frame, the receiving end determines whether it supports the redundancy level. If it supports it, it prepares to receive according to the redundancy level; if it does not support it, it sends a downgrade request to the sending end. Step S5: When the final degradation probability continues to be lower than the degradation threshold corresponding to the current redundancy level and the corresponding hysteresis time is met, the redundancy level is gradually reduced and eventually reverts to level 0 single-link mode. Steps S1 to S3 constitute a link quality prediction method based on power frequency phase synchronization, and steps S4 to S5 constitute a gradual soft handover method based on dual-mode parallel redundancy; the two are closely coupled through the final degradation probability and the predicted SNR lower limit.

2. The seamless handover method for power dual-mode communication links based on real-time SNR evaluation according to claim 1, characterized in that: In step S1, after each frame of data is received, the power line carrier physical layer completes channel estimation using a preamble and training sequence, calculates the subcarrier SNR and the 1-byte average SNR, and the average SNR value ranges from [value missing]. Generate reporting primitives and push them to the MAC sub-layer; The wireless physical layer calculates the average SNR of a single carrier and the entire frame based on the PPDU data packet LTF long training domain and reports it to the data link layer in real time. Simultaneously with the push to the data link layer, the current power frequency phase angle is obtained through power frequency zero-crossing detection or preamble signal extraction. Establish the mapping relationship between phase angle and SNR.

3. The seamless handover method for power dual-mode communication links based on real-time SNR evaluation according to claim 1, characterized in that: In step S1, the power frequency phase angle The extraction method is as follows: Hardware zero-crossing detection circuits extract voltage zero-crossing points or parse the phase information built into the carrier preamble frame synchronization header; a local 16-bit phase counter provides high resolution. Each SNR sample is bound to a phase. Timestamp, with no less than a preset number of samples per wave.

4. The seamless handover method for power dual-mode communication links based on real-time SNR evaluation according to claim 1, characterized in that: In step S2, the sliding window size Number of phase intervals A dynamic adaptive adjustment mechanism based on channel state is adopted; specifically: During the communication link initialization phase or the period of stable signal-to-noise ratio, reduce the number of phase interval divisions. The value is adjusted to reduce computational overhead and maintain a small sliding window size; when drastic fluctuations in the signal-to-noise ratio or sudden impulse noise interference are detected, the system automatically increases the number of phase interval divisions. To improve phase resolution, while increasing the sliding window size Expand to a preset maximum value to smooth the impact of extreme noise samples on statistical results; the sliding window stores the most recent... A complete power frequency cycle sampling data; number of phase interval divisions ; use The criteria remove outlier noise samples, and the Welford online recursive algorithm is used to update the interval mean and standard deviation in real time.

5. The seamless handover method for power dual-mode communication links based on real-time SNR evaluation according to claim 1, characterized in that: In step S3, the threshold is switched. Set the minimum demodulation SNR plus a 3dB margin based on the selected MCS modulation method; Calculate the lower bound of the predicted SNR: in Here is the confidence coefficient; then calculate the initial degradation probability: in The standard normal cumulative distribution function is... The switching threshold is used; considering the presence of aperiodic pulse interference in the power line channel, which causes drastic fluctuations in the instantaneous signal-to-noise ratio, a measured value correction factor is introduced here. Used to dynamically adjust the prediction confidence level, measured value correction factor The calculation formula is: in This is a numerical stability constant to prevent the denominator from being zero. The value of reflects the degree of deviation between the current measured value of the channel and the historical statistical mean, and a correction coefficient is further introduced. Output the final degradation probability: 。 6. The seamless handover method for power dual-mode communication links based on real-time SNR evaluation according to claim 1, characterized in that: In step S4, the multiple redundancy levels include at least level 0, level 1 and level 2, which correspond to different link usage strategies and data protection strengths, respectively. When the link quality meets the first preset condition, level 0 redundancy is adopted, and transmission is only carried out through the power line carrier link, while the wireless link goes into sleep mode. When the link quality meets the second preset condition, Level 1 redundancy is adopted. The same MAC frame is copied and transmitted simultaneously on both the power line carrier and wireless links. The receiving end removes duplicates according to the frame sequence number. When the link quality meets the third preset condition, a level 2 redundancy is adopted, and K original data packets are generated by Reed-Solomon encoding. Each fragment is sent interleaved on two links according to its parity index. Fragments with odd indices are sent via the PLC link, and fragments with even indices are sent via the RF link. The two links are triggered alternately in ascending order of fragment index; that is, the PLC link sends the first fragment while the RF link sends the second fragment, then the PLC link sends the third fragment while the RF link sends the fourth fragment, and so on. The receiving end collects any K fragments and then decodes and recovers the data. The first, second, and third preset conditions are as follows: First presupposition condition: when and At that time, level 0 redundancy is adopted; Second presupposition condition: When or At that time, Level 1 redundancy is adopted; Third pre-set condition: When or At that time, two levels of redundancy are adopted; In Level 1 redundancy, the MAC layer assigns a unique frame sequence number (SN) to each service frame and adds a 1-byte redundancy flag field to the frame header: 0x01 indicates a single-link frame, 0x02 indicates a duplicate frame, and 0x03 indicates a fragmented frame. Two identical MPDUs are constructed and sent to the power line carrier and radio transmission queue respectively. The two physical layers transmit independently with a transmission time difference of no more than one OFDM symbol. The receiver maintains a receive buffer indexed by SN and discards duplicate frames directly.

7. The seamless handover method for power dual-mode communication links based on real-time SNR evaluation according to claim 1, characterized in that: In step S4, level 2 redundancy employs a systematic Reed architecture. Solomon code ,Will Each fragment is marked as Power line carrier links transmit odd-index fragments, while wireless links transmit even-index fragments; the receiver collects arbitrary... After each different fragment is processed, RS decoding is performed immediately to recover the original data, with a decoding delay of no more than one OFDM symbol.

8. The seamless handover method for power dual-mode communication links based on real-time SNR evaluation according to claim 1, characterized in that: Also includes: Step S5, when If the redundancy level remains below the degradation threshold corresponding to the current redundancy level, the threshold for downgrading from level 1 to level 0 is 0.2, and the threshold for downgrading from level 2 to level 1 is 0.

5. After meeting the corresponding hysteresis time, the redundancy level is gradually reduced and eventually reverts to level 0 single-link mode. In step S5, the hysteresis protection rule is as follows: To downgrade from level 1 to level 0, the following conditions must be met. And it must remain stable for more than 200ms hysteresis; to degrade from level 2 back to level 1, the following must be met. And it remains stable for more than 200ms hysteresis; when Stable below 0.2 and When necessary, the wireless transmission channel is shut down, and the wireless physical layer is put into low-power sleep mode; the threshold and hysteresis duration are determined by debugging according to the on-site channel environment. Steps S1 to S3 constitute a link quality prediction method based on power frequency phase synchronization, and steps S4 to S5 constitute a gradual soft handover method based on dual-mode parallel redundancy; both are achieved through... and Tightly coupled.

9. A seamless handover method for power dual-mode communication links based on real-time SNR assessment according to claim 1, characterized in that: In step S4, the redundancy level negotiation is achieved through a newly added management frame. This management frame reuses the existing MME frame type and extends the subtype fields, including: STA→PCO / CCO, carrying Value, current phase, and recommended redundancy level; PCO / CCO → STA, confirm the level change and allocate temporary shard ID space; It periodically reports the current redundancy level, link quality statistics, and phase information in both directions.

10. A seamless handover method for power dual-mode communication links based on real-time SNR assessment according to claim 1, characterized in that: In step S4, the data link layer uses a message encapsulation format common to both power line carrier links and wireless links for transmission, and there is no need to reassemble packets during the switching process; the dual links use a unified sequence number and retransmission mechanism, and there is no need to retransmit the entire packet during switching, so that the upper layer services are unaware of the process.

Images