Self-adaptive link selection and switching method in power line and wireless dual-mode communication

Abstract

The invention relates to the technical field of network communication, and discloses a self-adaptive link selection and switching method in power line and wireless dual-mode communication, which comprises the following steps of: 1, acquiring a service quality physical quantity of a wireless link, forming a service quality state vector by the service quality physical quantity, the service quality physical quantity comprises effective throughput, end-to-end delay, delay jitter and packet loss rate. By adopting the technical scheme of aggregating the physical quantities of the quality of service of a wireless link into a single comprehensive utility value through non-additive fuzzy measurement and Choquet integral operation, the technical effect of quantifying the collaboration and redundancy effects among different physical quantities of the quality of service is achieved; compared with the technical scheme of independently evaluating the physical quantities by adopting a linear weighting model in the prior art, the link quality evaluation method has the advantage that the defects that the link quality evaluation result is one-sided and the comprehensive experience cannot be truly reflected due to neglecting the internal relevance among the physical quantities are overcome.

Description

Technical Field

[0001] This invention relates to the field of network communication technology, specifically to an adaptive link selection and switching method in power line and wireless dual-mode communication. Background Technology

[0002] The widespread adoption of the Internet of Things (IoT), smart homes, and multimedia applications has placed high demands on users for stable network connectivity and a high-quality user experience. Power line communication (PLC) utilizes existing power lines for data transmission, offering advantages such as convenient deployment and strong wall penetration; while wireless communication provides high bandwidth and flexible mobile access. Combining these two technologies, the PLC / wireless dual-mode communication effectively integrates their strengths, acting as a backup for each other, thus providing users with a theoretically reliable and comprehensive network coverage solution. The key to achieving dual-mode communication performance lies in an intelligent and efficient adaptive link selection and switching method. This method must dynamically select the optimal data transmission path between links and smoothly switch when link quality changes, ensuring the service quality of upper-layer applications is always guaranteed.

[0003] However, existing dual-mode communication link switching technologies still have the following shortcomings that urgently need to be addressed in terms of achieving intelligence and optimal user experience: Current link quality assessment models suffer from biased decision-making. Traditional handover methods typically rely on key physical performance indicators, such as signal strength, signal-to-noise ratio, or single throughput, for judgment. Even though some solutions attempt to integrate multiple indicators, often using linear weighted summation to arrive at a comprehensive score, the core flaw of such models lies in the default assumption that the contributions of each service quality indicator to user experience are independent. However, in real-world scenarios, complex collaborative or redundant relationships exist between different QoS indicators. For example, in real-time video conferencing, the synergistic gain from extremely low latency and extremely low jitter occurring simultaneously far exceeds the sum of their individual effects; conversely, when latency is extremely high, the marginal benefit of simply increasing throughput decreases significantly. Linear models struggle to quantify such non-additive interactions, resulting in incomplete assessments that fail to accurately reflect the link's overall carrying capacity for a specific application, leading to flawed handover decisions.

[0004] Secondly, existing switching strategies generally lack quantitative awareness of the upper-layer business experience. Most switching methods use a fixed set of switching thresholds or rules that are independent of application type. This means that whether it's large file transfers, real-time voice calls, or online competitive games, the system uses the same standard to judge the quality of the link. This approach ignores the fact that different applications have drastically different sensitivities to network service quality. For example, file download applications have high throughput requirements but can tolerate greater latency and jitter; while real-time games are the opposite, extremely sensitive to latency and jitter, but have relatively lower throughput requirements. Because existing technologies cannot dynamically adjust evaluation models and decision preferences based on the currently running applications, link selection is often blind, frequently choosing links that seem to have superior physical parameters but actually provide a poor experience for the current application.

[0005] Finally, existing handover execution mechanisms suffer from an uneven handover process. At the decision-trigger level, many solutions rely on instantaneous network parameter values, triggering a handover immediately once a certain indicator exceeds a preset threshold. However, due to inherent network fluctuations and sudden interference, this mechanism easily leads to frequent and ineffective back-and-forth handovers, wasting system resources and introducing additional network jitter, further degrading the user experience. At the handover execution level, the widely adopted "disconnect first, reconnect later" mechanism results in a clear service interruption window during the handover process. For real-time applications requiring continuous connectivity, such as video calls or online payments, this interruption is unacceptable. Therefore, existing technologies are insufficient in terms of decision stability and execution continuity, making it difficult to achieve a smooth, seamless handover for the user.

[0006] To address these issues, this invention proposes an adaptive link selection and switching method for power line and wireless dual-mode communication. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides an adaptive link selection and switching method in power line and wireless dual-mode communication, thereby solving the problems mentioned in the background section.

[0008] To achieve the above objectives, the present invention provides the following technical solution: an adaptive link selection and switching method in power line and wireless dual-mode communication, comprising the following steps: Step 1: Obtain the physical quantities of service quality for the wireless link. The physical quantities of service quality constitute the service quality state vector. The physical quantities of service quality include effective throughput, end-to-end latency, latency jitter, and packet loss rate. Step 2: Based on the current application type, call the preset single-attribute utility function to transform the physical quantity of service quality in the service quality state vector into a single-attribute experience utility value, thus forming an experience utility vector; Step 3: Invoke the preset fuzzy measure corresponding to the current application type, and calculate the comprehensive utility value for the wireless link by performing Choquet integration on the experience utility vector; Step 4: Compare the overall utility value of the wireless link, and decide whether to perform link switching based on the comparison result and the preset switching stability conditions.

[0009] Preferably, the single-attribute utility function is a nonlinear function, wherein the function used to transform the effective throughput is a logarithmic function, the function used to transform the end-to-end delay and delay jitter is an exponential decay function, and the function used to transform the packet loss rate is an S-shaped function.

[0010] Preferably, the preset single-attribute utility function and the preset fuzzy measure together constitute a collaborative utility profile bound to a specific application type. The adaptive link selection and switching method includes identifying the current application type through traffic analysis before making a decision and loading the corresponding collaborative utility profile.

[0011] Preferably, the process of performing Choquet integration on the experience utility vector includes: First, sort the single-attribute experience utility values ​​in the experience utility vector in ascending order to obtain an ordered sequence of utility values ​​and the corresponding ordered attribute indexes; Construct a nested subset of attributes based on the ordered attribute index; The difference between adjacent utility values ​​in the ordered utility value sequence is multiplied by the fuzzy measure value of the attribute subset corresponding to the higher of the differences, and all products are summed to obtain the comprehensive utility value.

[0012] Preferably, the switching stability condition includes a hysteresis margin condition and a trigger time condition; the hysteresis margin condition is a preset margin value in which the overall utility value of the backup link exceeds the preset overall utility value of the currently active link; the trigger time condition is a state duration in which the hysteresis margin condition is satisfied exceeds a preset time threshold.

[0013] Preferably, the adaptive link selection and switching method adopts a first-connect-then-disconnect execution mechanism when making a decision to switch links. The execution mechanism includes first establishing a network connection to the target link, then redirecting the data stream to the target link, and finally disconnecting the original link. The adaptive link selection and switching method includes an emergency switching mechanism. When the packet loss rate or end-to-end latency in the physical quantity of the quality of service of the currently active link exceeds a preset extreme deterioration threshold, the switching stability condition is bypassed and the link switching is directly decided and executed.

[0014] Preferably, step 1 further includes: Sub-step 1.1: Initialize active probe and record the departure sending timestamp; The system generates a set of quantities. The active probe data packets are the data packets in this group. Assign a unique serial number And in the data packet Record the departure timestamp when leaving the sending interface. ; Sub-step 1.2: Collect probe responses and calculate individual round-trip times; Based on the actively probed data packets sent, the system listens for and receives response data packets returned by the network target; for each successfully received response data packet with a sequence number of... The response data packet records the time of arrival at the receiving interface as the arrival timestamp. And using the arrival timestamp Sending timestamps after departure Calculate the individual round-trip time of the data packet. The calculation formula is: , At the same time, the total number of successfully received response data packets is counted and denoted as . ; Sub-step 1.3: Calculate the packet loss rate and average round-trip time; Using the defined total number of sends Compared with the total number of received statistics Calculate the packet loss rate The calculation formula is: , in, The preset total number of probe data packets to send. The total number of response packets successfully received; At the same time, using the calculated round-trip times of all individuals Calculate the average round trip time The calculation formula is: , in, This is the set of sequence numbers for all packets that successfully received a response; Sub-step 1.4: Calculate delay jitter and end-to-end delay; Using the calculated average round trip time The calculated individual round-trip time of each data packet Calculate the delay jitter The delay jitter Defined as the standard deviation of round-trip time for all individuals, the formula is: , Then, the calculated average round-trip time was used. Calculate the end-to-end delay ; Sub-step 1.5: Measure the effective throughput and construct the quality of service state vector; The system operates within a preset time window. Within the monitoring network interface controller, the byte transfer volume is counted, and the cumulative byte count at the start of the time window is obtained. Cumulative bytes at the end Calculate the effective throughput The calculation formula is: , in, This represents the cumulative number of bytes transmitted by the interface at the start of the time window. This represents the cumulative number of bytes transmitted by the interface at the end of the time window. The duration of the time window; Finally, the calculated effective throughput End-to-end delay Delay jitter With packet loss rate The combination of these elements constitutes the service quality state vector for this link.

[0015] Preferably, step 2 further includes: Sub-step 2.1: Application type identification and collaborative utility profile loading; The system identifies the current application type through traffic analysis and loads a unique collaborative utility profile corresponding to the current application type from a pre-set collaborative utility profile library. This collaborative utility profile provides subsequent transformation steps with parameters including throughput sensitivity. Maximum throughput reference value Delay sensitivity parameter jitter sensitivity parameters Packet loss rate curve steepness parameter and the critical center point of packet loss rate A set of utility function parameters, including; Sub-step 2.2, logarithmic transformation from effective throughput to experienced utility value; Utilizing the throughput sensitivity parameter of the load Compared with the maximum throughput reference value Effective throughput in the quality of service state vector Perform a logarithmic utility transformation to calculate the throughput experience utility value. The calculation formula is: , in, The effective throughput physical quantity is the input. The normalized experience utility value is the output. This is a preset positive parameter related to throughput sensitivity, which is relevant to the current application type. This is a preset maximum throughput reference value for normalization, which is related to the current application type. Sub-step 2.3, delay and jitter to exponential decay transformation of the experience utility value; Using the loading latency sensitivity parameter With jitter sensitivity parameter The end-to-end delay in the quality of service state vector is then processed sequentially. With delay jitter Perform an exponentially decaying utility transformation to calculate the delayed experience utility value. With shaking experience utility value The calculation formula is as follows: , , in, The input is the end-to-end delay physical quantity. For the input delay jitter physical quantity, and The normalized experience utility value is the output. and A preset positive parameter for sensitivity to latency and jitter, related to the current application type; Sub-step 2.4, S-shaped function transformation from packet loss rate to experience utility value; Using the kurtosis parameter of the loaded packet loss rate curve Critical center point of packet loss rate Packet loss rate in the quality of service state vector Perform an S-shaped utility transformation to calculate the experience utility value of packet loss rate. The calculation formula is: , in, The input is the physical quantity of packet loss rate. The normalized experience utility value is the output. A positive parameter representing the degree of drastic change in the preset control utility curve near the critical point, which is related to the current application type. The pre-defined packet loss rate midpoint where the user experience quality begins to deteriorate sharply, which is relevant to the current application type. Sub-step 2.5: Construct the experience utility vector; The calculated throughput experience utility value Delayed experience utility value With shaking experience utility value And packet loss rate experience utility value The combination ultimately constitutes the experience utility vector of this link. The experience utility vector The mathematical expression is as follows: , in, For the experience utility vector.

[0016] Preferably, step 3 further includes: Sub-step 3.1: Load the fuzzy measure corresponding to the application type; Based on the determined current application type, a preset fuzzy measure is loaded from the collaborative utility profile corresponding to the application type. The fuzzy measure It is defined in the service quality attribute set Set functions on the power set, which are subsets of each attribute. Assign a fuzzy measure value to represent the importance of this attribute combination. ; Sub-step 3.2: Experience utility vector sorting and ordered sequence generation; For the generated experience utility vector Sort the single-attribute experience utility values ​​in ascending order to obtain an ordered utility value sequence. ≤ ≤ ≤ And obtain an ordered attribute index sequence that corresponds one-to-one with the ordered utility value sequence. ; Simultaneously define the initial utility value. Set the initial utility value The value is ; Sub-step 3.3: Construct nested attribute subsets; Based on the generated ordered attribute index sequence, each index in the sequence is... Construct the corresponding attribute subset The attribute subset By index And all sorts in The attribute indexes are then constructed, and the mathematical expression is: , This generates a set of nested attribute subsets. ; Sub-step 3.4: Perform Choquet integration to aggregate the overall utility value; Using the generated ordered sequence of utility values Compared with the initial utility value and nested attribute subsets And call the loaded fuzzy measure Get each attribute subset Corresponding fuzzy measure value The final overall utility value is calculated by performing discrete Choquet integration. The calculation formula is: , in, The output is the overall utility value. For the summation index, For the first in the ordered utility value sequence value, For the th in the ordered utility value sequence value, Subset of attributes Importance metric.

[0017] Preferably, step 4 further includes: Sub-step 4.1: Define the link roles and obtain the overall utility value; Based on the current active link status, the system defines one of the power line links and the wireless links as the current link and the other as the backup link. Furthermore, the current comprehensive utility value is obtained for the current link based on the calculation results of step 3. To obtain the backup comprehensive utility value for the backup link. ; Sub-step 4.2: Perform emergency switchover condition determination; The system obtains the packet loss rate of the current link measured in step 1. With end-to-end delay Furthermore, it is determined whether the current link meets the emergency handover condition, and the mathematical expression for the emergency handover condition is: , in, The preset threshold for extreme packet loss rate, The preset extreme degradation delay threshold, For logical OR operator; If the emergency switching conditions are met, a switching instruction is generated directly and this step ends. Sub-step 4.3: Perform lag margin condition determination; If the emergency switching condition in sub-step 4.2 is not met, the system uses the current comprehensive utility value obtained in sub-step 4.1. Combined utility value with reserves To determine whether the lag margin condition is met, the mathematical expression for the lag margin condition is as follows: , in, This is a preset positive margin value used to prevent ping-pong switching; If the aforementioned hysteresis margin condition is met, the system starts or maintains a dominant state timer. ; If not satisfied, reset the dominant state timer. It is zero; Sub-step 4.4: Execute the trigger time condition determination and generate the final decision; The system determines the dominant state timer Whether the count value meets the trigger time condition, the mathematical expression for the trigger time condition is: ≥ , in, This is a preset trigger time threshold; If the triggering time condition is met, the system generates a switching instruction; If the conditions are not met, the system will not generate a switching instruction; ultimately, the result of either generating a switching instruction or not generating a switching instruction will be the final output of this decision.

[0018] This invention provides an adaptive link selection and switching method for power line and wireless dual-mode communication. It has the following advantages: 1. This invention adopts a technical solution that aggregates the physical quantities of service quality of power lines and wireless links into a single comprehensive utility value through non-additive fuzzy measurement and Choquet integral operation. This achieves the technical effect of quantifying the synergistic and redundancy effects between different physical quantities of service quality. Compared with the existing technical solution that uses a linear weighted model to independently evaluate physical quantities, this invention solves the shortcomings of the link quality evaluation results, which are one-sided and cannot truly reflect the overall experience, due to ignoring the inherent correlation between physical quantities.

[0019] 2. This invention employs a pre-defined collaborative utility profile bound to a specific application type. By dynamically identifying the current application type through traffic analysis and loading corresponding profile parameters, it achieves the technical effect of enabling link selection decisions to match the service quality requirements of the current application. Compared with the existing technology that uses fixed, application-independent switching thresholds for decision-making, this invention solves the shortcomings of blind link selection and failure to provide optimal services for specific applications due to a lack of quantitative perception of business experience.

[0020] 3. This invention adopts a stabilization mechanism that combines hysteresis margin conditions with trigger time conditions for handover decisions. It employs a "connect first, then disconnect" mechanism to execute handover operations, thereby avoiding invalid ping-pong handovers caused by instantaneous network fluctuations and ensuring service continuity during handover. Compared to existing technologies that rely on a single instantaneous threshold to trigger handover and employ a "disconnect first, then connect" mechanism, this invention addresses the shortcomings of frequent jitter during handover, which can easily cause data interruptions and result in an unsmooth handover experience. Attached Figure Description

[0021] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0022] To enable those skilled in the art to understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort should fall within the scope of protection of the present invention.

[0023] The present invention will now be described in detail with reference to the accompanying drawings: Example: Please see the appendix Figure 1 This invention provides an adaptive link selection and switching method in power line and wireless dual-mode communication, comprising the following steps: Step 1: Obtain the physical quantities of service quality for the wireless link. The physical quantities of service quality constitute the service quality state vector. The physical quantities of service quality include effective throughput, end-to-end latency, latency jitter, and packet loss rate. Step 2: Based on the current application type, call the preset single-attribute utility function to transform the physical quantity of service quality in the service quality state vector into a single-attribute experience utility value, thus forming an experience utility vector; Step 3: Invoke the preset fuzzy measure corresponding to the current application type, and calculate the comprehensive utility value for the wireless link by performing Choquet integration on the experience utility vector; Step 4: Compare the overall utility value of the wireless links, and decide whether to perform link switching based on the comparison results and the preset handover stability conditions.

[0024] Step 1 further includes: Sub-step 1.1: Initialize active probe and record the departure sending timestamp; The system generates a set of quantities. The active probe data packets are the data packets in this group. Assign a unique serial number And in the data packet Record the departure timestamp when leaving the sending interface. ; Sub-step 1.2: Collect probe responses and calculate individual round-trip times; Based on the actively probed data packets sent, the system listens for and receives response data packets returned by the network target; for each successfully received response data packet with a sequence number of... The response data packet records the time of arrival at the receiving interface as the arrival timestamp. And using arrival timestamps Sending timestamps after departure Calculate the individual round-trip time of the data packet. The calculation formula is: , At the same time, the total number of successfully received response data packets is counted and denoted as . ; Sub-step 1.3: Calculate the packet loss rate and average round-trip time; Using the defined total number of sends Compared with the total number of received statistics Calculate the packet loss rate The calculation formula is: , in, The preset total number of probe data packets to send. The total number of response packets successfully received; At the same time, using the calculated round-trip times of all individuals Calculate the average round trip time The calculation formula is: , in, This is the set of sequence numbers for all packets that successfully received a response; Sub-step 1.4: Calculate delay jitter and end-to-end delay; Using the calculated average round trip time The calculated individual round-trip time of each data packet Calculate the delay jitter Delay jitter Defined as the standard deviation of round-trip time for all individuals, the formula is: , Then, the calculated average round-trip time was used. Calculate the end-to-end delay ; Sub-step 1.5: Measure the effective throughput and construct the quality of service state vector; The system operates within a preset time window. Internally, monitor the byte transfer volume count of the network interface controller by obtaining the cumulative byte count at the start of the time window. Cumulative bytes at the end Calculate the effective throughput The calculation formula is: , in, This represents the cumulative number of bytes transmitted by the interface at the start of the time window. This represents the cumulative number of bytes transmitted by the interface at the end of the time window. The duration of the time window; Finally, the calculated effective throughput End-to-end delay Delay jitter With packet loss rate The combination of these elements constitutes the service quality state vector for this link.

[0025] Step 2 further includes: Sub-step 2.1: Application type identification and collaborative utility profile loading; The system identifies the current application type through traffic analysis and loads a unique collaborative utility profile corresponding to the current application type from a pre-defined collaborative utility profile library. This collaborative utility profile provides subsequent transformation steps with parameters including throughput sensitivity. Maximum throughput reference value Delay sensitivity parameter jitter sensitivity parameters Packet loss rate curve steepness parameter and the critical center point of packet loss rate A set of utility function parameters, including; Sub-step 2.2, logarithmic transformation from effective throughput to experienced utility value; Utilizing the throughput sensitivity parameter of the load Compared with the maximum throughput reference value Effective throughput in the quality of service state vector Perform a logarithmic utility transformation to calculate the throughput experience utility value. The calculation formula is: , in, The effective throughput physical quantity is the input. The normalized experience utility value is the output. This is a preset positive parameter related to throughput sensitivity, which is relevant to the current application type. This is a preset maximum throughput reference value for normalization, which is related to the current application type. Sub-step 2.3, delay and jitter to exponential decay transformation of the experience utility value; Using the loading latency sensitivity parameter With jitter sensitivity parameter The end-to-end delay in the quality of service state vector is then processed sequentially. With delay jitter Perform an exponentially decaying utility transformation to calculate the delayed experience utility value. With shaking experience utility value The calculation formula is: , , in, The input is the end-to-end delay physical quantity. For the input delay jitter physical quantity, and The normalized experience utility value is the output. and A preset positive parameter for sensitivity to latency and jitter, related to the current application type; Sub-step 2.4, S-shaped function transformation from packet loss rate to experience utility value; Using the kurtosis parameter of the loaded packet loss rate curve Critical center point of packet loss rate Packet loss rate in the service quality state vector Perform an S-shaped utility transformation to calculate the experience utility value of packet loss rate. The calculation formula is: , in, The input is the physical quantity of packet loss rate. The normalized experience utility value is the output. A positive parameter representing the degree of drastic change in the preset control utility curve near the critical point, which is related to the current application type. The pre-defined center point of packet loss rate where the experience quality begins to deteriorate sharply, which is related to the current application type; Sub-step 2.5: Construct the experience utility vector; The calculated throughput experience utility value Delayed experience utility value With shaking experience utility value And packet loss rate experience utility value The combination ultimately constitutes the experience utility vector of this link. Experiencing utility vector The mathematical expression is as follows: , in, For the experience utility vector.

[0026] Step 3 further includes: Sub-step 3.1: Load the fuzzy measure corresponding to the application type; Based on the determined current application type, a preset fuzzy measure is loaded from the collaborative utility profile corresponding to the application type. fuzzy measure It is defined in the service quality attribute set Set functions on the power set, which are subsets of each attribute. Assign a fuzzy measure value to represent the importance of this attribute combination. ; Sub-step 3.2: Experience utility vector sorting and ordered sequence generation; For the generated experience utility vector Sort the single-attribute experience utility values ​​in ascending order to obtain an ordered utility value sequence. ≤ ≤ ≤ And obtain an ordered attribute index sequence that corresponds one-to-one with the ordered utility value sequence. ; Simultaneously define the initial utility value. Set initial utility value The value is ; Sub-step 3.3: Construct nested attribute subsets; Based on the generated ordered attribute index sequence, assign each index in the sequence to... Construct the corresponding attribute subset Attribute subset By index And all sorts in The attribute indexes are then constructed, and the mathematical expression is: , This generates a set of nested attribute subsets. ; Sub-step 3.4: Perform Choquet integration to aggregate the overall utility value; Using the generated ordered sequence of utility values Compared with the initial utility value and nested attribute subsets And call the loaded fuzzy measure Get each attribute subset Corresponding fuzzy measure value The final overall utility value is calculated by performing discrete Choquet integration. The calculation formula is: , in, The output is the overall utility value. For the summation index, For the first in the ordered utility value sequence value, For the th in the ordered utility value sequence value, Subset of attributes Importance metric.

[0027] Step 4 further includes: Sub-step 4.1: Define the link roles and obtain the overall utility value; Based on the current active link status, the system defines one of the power line links and the wireless links as the current link and the other as the backup link. Furthermore, the current comprehensive utility value is obtained from the calculation results of step 3 for the current link. To obtain the backup comprehensive utility value for backup links ; Sub-step 4.2: Perform emergency switchover condition determination; The system obtains the packet loss rate of the current link measured in step 1. With end-to-end delay Furthermore, it determines whether the current link meets the emergency handover conditions. The mathematical expression for the emergency handover conditions is: , in, The preset threshold for extreme packet loss rate, The preset extreme degradation delay threshold, For logical OR operator; If the emergency switchover conditions are met, a switchover instruction is generated directly and this step ends. Sub-step 4.3: Perform lag margin condition determination; If the emergency switching condition in sub-step 4.2 is not met, the system uses the current comprehensive utility value obtained in sub-step 4.1. Combined utility value with reserves To determine whether the lag margin condition is met, the mathematical expression for the lag margin condition is: , in, This is a preset positive margin value used to prevent ping-pong switching; If the lag margin condition is met, the system starts or maintains the dominant state timer. ; If not satisfied, reset the dominant state timer. It is zero; Sub-step 4.4: Execute the trigger time condition determination and generate the final decision; System dominance state timer Whether the count value meets the trigger time condition, the mathematical expression for the trigger time condition is: ≥ , in, This is a preset trigger time threshold; If the trigger time condition is met, the system generates a switching instruction; If the conditions are not met, the system will not generate a switching instruction; ultimately, the result of either generating a switching instruction or not generating a switching instruction will be the final output of this decision.

[0028] By quantifying and collecting physical quantities related to the quality of service (QoS) of power lines and wireless links through a series of well-defined sub-steps and calculation formulas, the advantage lies in providing a comprehensive, accurate, and real-time raw data foundation for subsequent decision-making. This includes throughput, latency, jitter, and packet loss rate—quantities required for evaluating network performance. Specific methods, such as calculating jitter using standard deviation and monitoring throughput through time windows, ensure the objectivity and reliability of the data, avoiding the pitfalls of relying on a single or inaccurate physical indicator for biased judgments.

[0029] By introducing a nonlinear single-attribute utility function bound to specific application types, the objective physical quantities collected in step 1 are transformed into subjective experience utility values. The advantage is that the evaluation criteria for link quality are elevated from a purely technical dimension to a user experience-oriented dimension. This step deeply understands the impact of different physical quantities on user experience, and that different applications have varying sensitivities to each indicator. It uses logarithmic, exponential decay, and sigmoid functions to accurately simulate the real-world patterns of experience changes caused by physical quantities, resolving the core contradiction in traditional methods that physical optimality does not equate to optimal experience. This ensures that subsequent link selection decisions truly aim to maximize user satisfaction for specific applications.

[0030] By employing Choquet integral computation based on fuzzy measures to aggregate the individual attribute experience utility values, the advantage lies in its ability to scientifically and precisely evaluate the synergistic and redundant effects among different service quality indicators, resulting in a highly accurate total utility score that reflects the overall service capability of the link. Compared to the traditional linear weighted model, this step can identify, for example, the synergistic gains in real-time communication when "low latency" and "low jitter" are simultaneously satisfied, or the redundant compensation effect of "high throughput" on a small amount of "packet loss," completely resolving the fundamental deficiency of linear models in handling non-additive relationships between indicators.

[0031] By introducing three conditional decision-making mechanisms—lag margin, trigger time, and emergency handover—based on a comparison of comprehensive utility values, the advantage lies in ensuring that the final handover decision possesses stability, timeliness, and robustness. The combination of lag margin and trigger time conditions constitutes an effective "decision filter," which can avoid unnecessary and experience-damaging "ping-pong handovers" caused by instantaneous and minor network fluctuations. Meanwhile, the emergency handover mechanism ensures that when the current link experiences severe degradation, the system can bypass the regular stabilization process and restore service as quickly as possible, achieving a perfect balance between ensuring stable normal operation and responding to sudden network failures.

[0032] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An adaptive link selection and handover method in power line and wireless dual-mode communication, characterized in that, Includes the following steps: Step 1: Obtain the physical quantities of service quality for the wireless link. The physical quantities of service quality constitute the service quality state vector. The physical quantities of service quality include effective throughput, end-to-end latency, latency jitter, and packet loss rate. Step 2: Based on the current application type, call the preset single-attribute utility function to transform the physical quantity of service quality in the service quality state vector into a single-attribute experience utility value, thus forming an experience utility vector; Step 3: Invoke the preset fuzzy measure corresponding to the current application type, and calculate the comprehensive utility value for the wireless link by performing Choquet integration on the experience utility vector; Step 4: Compare the overall utility value of the wireless link, and decide whether to perform link switching based on the comparison result and the preset switching stability conditions.

2. The adaptive link selection and switching method in power line and wireless dual-mode communication according to claim 1, characterized in that, The single-attribute utility function is a nonlinear function, wherein the function used to transform the effective throughput is a logarithmic function, the function used to transform the end-to-end delay and delay jitter is an exponential decay function, and the function used to transform the packet loss rate is an sigmoid function.

3. The adaptive link selection and switching method in power line and wireless dual-mode communication according to claim 1, characterized in that, The preset single-attribute utility function and the preset fuzzy measure together constitute a collaborative utility profile bound to a specific application type. The adaptive link selection and switching method includes identifying the current application type through traffic analysis before making a decision, and loading the corresponding collaborative utility profile.

4. The adaptive link selection and switching method in power line and wireless dual-mode communication according to claim 1, characterized in that, The process of performing Choquet integration on the experience utility vector includes: First, sort the single-attribute experience utility values ​​in the experience utility vector in ascending order to obtain an ordered sequence of utility values ​​and the corresponding ordered attribute indexes; Construct a nested subset of attributes based on the ordered attribute index; The difference between adjacent utility values ​​in the ordered utility value sequence is multiplied by the fuzzy measure value of the attribute subset corresponding to the higher of the differences, and all products are summed to obtain the comprehensive utility value.

5. The adaptive link selection and switching method in power line and wireless dual-mode communication according to claim 1, characterized in that, The switching stability conditions include a hysteresis margin condition and a trigger time condition; the hysteresis margin condition is a margin value by which the overall utility value of the backup link exceeds the preset overall utility value of the currently active link; the trigger time condition is a state duration that satisfies the hysteresis margin condition that exceeds a preset time threshold.

6. The adaptive link selection and switching method in power line and wireless dual-mode communication according to claim 1, characterized in that, The adaptive link selection and switching method adopts a connection-then-disconnection execution mechanism when making a decision to switch links. The execution mechanism includes first establishing a network connection to the target link, then redirecting the data flow to the target link, and finally disconnecting the original link. The adaptive link selection and switching method includes an emergency switching mechanism. When the packet loss rate or end-to-end latency in the physical quantity of the quality of service of the currently active link exceeds a preset extreme deterioration threshold, the switching stability condition is bypassed and the link switching is directly decided and executed.

7. The adaptive link selection and switching method in power line and wireless dual-mode communication according to claim 1, characterized in that, Step 1 further includes: Sub-step 1.1: Initialize active probe and record the departure sending timestamp; The system generates a set of quantities. The active probe data packets are the data packets in this group. Assign a unique serial number And in the data packet Record the departure timestamp when leaving the sending interface. ; Sub-step 1.2: Collect probe responses and calculate individual round-trip times; Based on the actively probed data packets sent, the system listens for and receives response data packets returned by the network target; for each successfully received response data packet with a sequence number of... The response data packet records the time of arrival at the receiving interface as the arrival timestamp. And using the arrival timestamp Sending timestamps after departure Calculate the individual round-trip time of the data packet. The calculation formula is: ， At the same time, the total number of successfully received response data packets is counted and denoted as . ; Sub-step 1.3: Calculate the packet loss rate and average round-trip time; Using the defined total number of sends Compared with the total number of received statistics Calculate the packet loss rate The calculation formula is: ， in, The preset total number of probe data packets to send. The total number of response packets successfully received; At the same time, using the calculated round-trip times of all individuals Calculate the average round trip time The calculation formula is: ， in, This is the set of sequence numbers for all packets that successfully received a response; Sub-step 1.4: Calculate delay jitter and end-to-end delay; Using the calculated average round trip time The calculated round-trip time of each data packet Calculate the delay jitter The delay jitter Defined as the standard deviation of round-trip time for all individuals, the formula is: ， Then, the calculated average round-trip time was used. Calculate the end-to-end delay ; Sub-step 1.5: Measure the effective throughput and construct the quality of service state vector; The system operates within a preset time window. Within the monitoring network interface controller, the byte transfer volume is counted, and the cumulative byte count at the start of the time window is obtained. Cumulative bytes at the end Calculate the effective throughput The calculation formula is: ， in, This represents the cumulative number of bytes transmitted by the interface at the start of the time window. This represents the cumulative number of bytes transmitted by the interface at the end of the time window. The duration of the time window; Finally, the calculated effective throughput End-to-end delay Delay jitter With packet loss rate The combination of these elements constitutes the service quality state vector for this link.

8. The adaptive link selection and switching method in power line and wireless dual-mode communication according to claim 1, characterized in that, Step 2 further includes: Sub-step 2.1: Application type identification and collaborative utility profile loading; The system identifies the current application type through traffic analysis and loads a unique collaborative utility profile corresponding to the current application type from a pre-set collaborative utility profile library. This collaborative utility profile provides subsequent transformation steps with parameters including throughput sensitivity. Maximum throughput reference value Delay sensitivity parameter jitter sensitivity parameters Packet loss rate curve steepness parameter and the critical center point of packet loss rate A set of utility function parameters, including; Sub-step 2.2, logarithmic transformation from effective throughput to experienced utility value; Utilizing the throughput sensitivity parameter of the load Compared with the maximum throughput reference value Effective throughput in the service quality state vector Perform a logarithmic utility transformation to calculate the throughput experience utility value. The calculation formula is: ， in, The effective throughput physical quantity is the input. The normalized experience utility value is the output. This is a preset positive parameter related to throughput sensitivity, which is relevant to the current application type. This is a preset maximum throughput reference value for normalization, which is related to the current application type. Sub-step 2.3, delay and jitter to exponential decay transformation of the experience utility value; Using the loading latency sensitivity parameter With jitter sensitivity parameter The end-to-end delay in the quality of service state vector is then processed sequentially. With delay jitter Perform an exponentially decaying utility transformation to calculate the delayed experience utility value. With shaking experience utility value The calculation formula is as follows: ， ， in, The input is the end-to-end delay physical quantity. For the input delay jitter physical quantity, and The normalized experience utility value is the output. and A preset positive parameter for sensitivity to latency and jitter, related to the current application type; Sub-step 2.4, S-shaped function transformation from packet loss rate to experience utility value; Using the kurtosis parameter of the loaded packet loss rate curve Critical center point of packet loss rate Packet loss rate in the quality of service state vector Perform an S-shaped utility transformation to calculate the experience utility value of packet loss rate. The calculation formula is: ， in, The input is the physical quantity of packet loss rate. The normalized experience utility value is the output. A positive parameter representing the degree of drastic change in the preset control utility curve near the critical point, which is related to the current application type. The pre-defined packet loss rate midpoint where the user experience quality begins to deteriorate sharply, which is relevant to the current application type. Sub-step 2.5: Construct the experience utility vector; The calculated throughput experience utility value Delayed experience utility value With shaking experience utility value And packet loss rate experience utility value The combination ultimately constitutes the experience utility vector of this link. The experience utility vector The mathematical expression is as follows: ， in, For the experience utility vector.

9. The adaptive link selection and switching method in power line and wireless dual-mode communication according to claim 1, characterized in that, Step 3 further includes: Sub-step 3.1: Load the fuzzy measure corresponding to the application type; Based on the determined current application type, a preset fuzzy measure is loaded from the collaborative utility profile corresponding to the application type. The fuzzy measure It is defined in the service quality attribute set Set functions on the power set, which are subsets of each attribute. Assign a fuzzy measure value to represent the importance of this attribute combination. ; Sub-step 3.2: Experience utility vector sorting and ordered sequence generation; For the generated experience utility vector Sort the single-attribute experience utility values ​​in ascending order to obtain an ordered utility value sequence. ≤ ≤ ≤ And obtain an ordered attribute index sequence that corresponds one-to-one with the ordered utility value sequence. ; Simultaneously define the initial utility value. Set the initial utility value The value is ; Sub-step 3.3: Construct nested attribute subsets; Based on the generated ordered attribute index sequence, each index in the sequence is... Construct the corresponding attribute subset The attribute subset By index And all sorts in The attribute indexes are then constructed, and the mathematical expression is: ， This generates a set of nested attribute subsets. ; Sub-step 3.4: Perform Choquet integration to aggregate the overall utility value; Using the generated ordered sequence of utility values Compared with the initial utility value and nested attribute subsets And call the loaded fuzzy measure Get each attribute subset Corresponding fuzzy measure value The final overall utility value is calculated by performing discrete Choquet integration. The calculation formula is: ， in, The output is the overall utility value. For the summation index, For the first in the ordered utility value sequence value, For the th in the ordered utility value sequence value, Subset of attributes Importance metric.

10. The adaptive link selection and switching method in power line and wireless dual-mode communication according to claim 1, characterized in that, Step 4 further includes: Sub-step 4.1: Define the link roles and obtain the overall utility value; Based on the current active link status, the system defines one of the power line links and the wireless links as the current link and the other as the backup link. Furthermore, the current comprehensive utility value is obtained for the current link based on the calculation results of step 3. To obtain the backup comprehensive utility value for the backup link. ; Sub-step 4.2: Perform emergency switchover condition determination; The system obtains the packet loss rate of the current link measured in step 1. With end-to-end delay Furthermore, it is determined whether the current link meets the emergency handover condition, and the mathematical expression for the emergency handover condition is: ， in, The preset threshold for extreme packet loss rate, The preset extreme degradation delay threshold, For logical OR operator; If the emergency switching conditions are met, a switching instruction is generated directly and this step ends. Sub-step 4.3: Perform lag margin condition determination; If the emergency switching condition in sub-step 4.2 is not met, the system uses the current comprehensive utility value obtained in sub-step 4.

1. Combined utility value with reserves To determine whether the lag margin condition is met, the mathematical expression for the lag margin condition is as follows: ， in, This is a preset positive margin value used to prevent ping-pong switching; If the aforementioned hysteresis margin condition is met, the system starts or maintains a dominant state timer. ; If not satisfied, reset the dominant state timer. It is zero; Sub-step 4.4: Execute the trigger time condition determination and generate the final decision; The system determines the dominant state timer Whether the count value meets the trigger time condition, the mathematical expression for the trigger time condition is: ≥ ， in, This is a preset trigger time threshold; If the triggering time condition is met, the system generates a switching instruction; If the conditions are not met, the system will not generate a switching instruction; ultimately, the result of either generating a switching instruction or not generating a switching instruction will be the final output of this decision.

Images