Intelligent disaster recovery method for short message channel of Internet of Things equipment

Abstract

The invention relates to an intelligent disaster recovery method for short message channels of an internet of things device, and belongs to the field of IT and software development. The method comprises the following steps: collecting a multi-dimensional short message index of each short message channel, and carrying out preprocessing; performing standardization processing on the short message indexes, determining the weight of each short message index, adjusting the weight of each short message index in real time according to the scene requirement of the Internet of Things, establishing a weighted comprehensive scoring model, and performing quantitative evaluation on the overall performance of each short message channel; according to the scene and the time period of the Internet of Things, dynamically adjusting a triggering threshold value of channel switching, and preferentially selecting a short message channel with the highest score as a standby channel in combination with the sorting of the comprehensive scores of the channels; a pre-verification mechanism is introduced, and a short message is sent to the test equipment before a formal channel is switched so as to perform pre-verification on a standby channel; and executing formal channel switching after the pre-verification is passed. According to the method, the channel state can be automatically monitored, the channel performance can be comprehensively evaluated, and the optimal switching can be quickly triggered.

Description

Technical Field

[0001] This invention belongs to the field of IT and software development, and particularly relates to an intelligent disaster recovery method for SMS channels of Internet of Things (IoT) devices. Background Technology

[0002] Currently, IoT devices (such as smart water meters, gas meters, industrial sensors, and connected vehicles) widely rely on SMS channels to achieve real-time, highly reliable communication. For example, automakers send commands to in-vehicle devices via connected vehicles, industrial sensors send alarm information via SMS in abnormal states, and platforms send parameter configuration commands to devices via SMS. These scenarios place extremely high demands on the real-time performance and stability of the SMS channel. If the channel fails or its success rate decreases, it will lead to data loss and command delivery failures, thus affecting the normal operation of the entire IoT system.

[0003] The existing SMS channel has the following problems:

[0004] 1. Currently, IoT platforms generally manage SMS communication using either a "fixed channel" or a "manually switched channel" approach:

[0005] (1) Fixed Channel Mode: The platform uses a single SMS channel to send SMS messages to devices for an extended period. When the success rate of this channel decreases due to network fluctuations, bandwidth saturation, or equipment failure (e.g., from 99% to below 80%), it cannot automatically adjust, resulting in a large number of SMS sending failures. For example, a smart meter reading platform uses channel A. When the success rate of channel A drops to 75% due to regional network congestion, 25% of water meter data cannot be uploaded. Manual troubleshooting is required before switching channels, resulting in several hours of data interruption.

[0006] (2) Manual Switching Mode: Although some platforms are configured with multiple backup channels, maintenance personnel need to monitor the success rate of the channels in real time. When an anomaly is detected in the target channel, the system configuration is manually modified to switch to the backup channel. This method has the problem of response delay: maintenance personnel cannot monitor in real time 24 hours a day, and the switching operation takes 10-30 minutes. During this period, a large number of SMS messages are still sent through the faulty channel, resulting in a high failure rate. At the same time, manual switching relies on experience judgment, which is prone to problems such as incorrect switching (e.g., switching a channel with a slightly lower success rate but lower cost to a channel with a high success rate but higher cost, increasing unnecessary expenses) or incorrect selection of backup channels (e.g., selecting a backup channel that is itself in a low success rate state).

[0007] 3. Simple polling channel: The platform uses multiple channels in a fixed order (such as channel A → channel B → channel C) without considering the real-time success rate differences of each channel. This may result in a large number of SMS messages being assigned to channels with low success rates. For example, when polling to channel B, its success rate is only 70%, but 1 / 3 of the SMS messages will still be sent through this channel, leading to an increased failure rate.

[0008] 4. Fixed Threshold Switching: Thresholds are set solely based on success rate (e.g., switching occurs when the success rate falls below 90%), without considering other critical factors such as channel latency and cost. For example, channel A has a 91% success rate (above the threshold) but a latency of up to 5 seconds, while channel B has an 89% success rate (below the threshold) but a latency of only 1 second and a lower cost. Under this fixed threshold, channel A should be retained. However, channel B is actually more suitable for latency-sensitive scenarios (such as industrial alarms), and the fixed threshold cannot meet the needs of multiple scenarios.

[0009] In summary, current IoT devices rely heavily on SMS channels, making it impossible to achieve "real-time, intelligent, and multi-dimensional" disaster recovery switching of SMS channels. There is an urgent need for a technical solution that can automatically monitor channel status, comprehensively evaluate channel performance, and quickly trigger optimal switching. Summary of the Invention

[0010] In view of the shortcomings of the prior art, the purpose of the invention is to provide an intelligent disaster recovery method for SMS channels of Internet of Things devices, which can automatically monitor the channel status, comprehensively evaluate the channel performance, and quickly trigger the optimal switchover.

[0011] This invention proposes an intelligent disaster recovery method for SMS channels of IoT devices, comprising:

[0012] S1. Collect multi-dimensional SMS metrics for each SMS channel and preprocess the SMS metrics to remove outliers and make corrections.

[0013] S2, the SMS metrics are standardized, and the weights of each SMS metric are determined using the analytic hierarchy process (AHP). The weights of each SMS metric are adjusted in real time according to the needs of the IoT scenario to establish a weighted comprehensive scoring model. This model is used to quantitatively evaluate the overall performance of each SMS channel, resulting in a comprehensive channel score for each SMS channel. The weighted comprehensive scoring model is as follows:

[0014] S=(S_R×W_R)+(S_T×W_T)+(S_C×W_C)+(S_B×W_B)

[0015] Where: S is the channel comprehensive score, S_R is the standardized score of channel success rate, S_T is the standardized score of channel latency, S_C is the standardized score of channel cost, S_B is the standardized score of channel bandwidth, W_R is the weight of channel success rate, W_T is the weight of channel latency, W_C is the weight of channel cost, and W_B is the weight of channel bandwidth.

[0016] S3. Based on the IoT scenario and time period, dynamically adjust the trigger threshold for channel switching, and combine the ranking of the comprehensive channel scores to prioritize the SMS channel with the highest score as the backup channel; wherein, the trigger threshold for channel switching is the threshold of the SMS index.

[0017] S4 introduces a pre-verification mechanism during the channel switching process. When the trigger threshold for channel switching is detected, an SMS is sent to the test device to pre-verify the backup channel before the formal channel switching. After the pre-verification is passed, the formal channel switching is performed, and the entire system is switched to the backup channel to achieve disaster recovery switching.

[0018] Furthermore, in S1, the multi-dimensional SMS metrics include: channel success rate, channel latency, channel cost, and channel bandwidth.

[0019] Furthermore, in S2, the channel success rate is preprocessed, including: when the channel success rate is detected to be greater than 100%, the channel success rate is corrected to 100%; when the channel success rate is detected to be less than 0%, the channel success rate is corrected to 0%; when the number of SMS messages sent in one minute is less than 10, the average success rate of the channel in the first 5 minutes is used as the channel success rate.

[0020] The channel latency is preprocessed, including: removing latency anomalies that are greater than the sum of the historical average latency of the SMS channel and 3 times the standard deviation, and then recalculating the average latency after the removal.

[0021] The channel cost and channel bandwidth are preprocessed, including: when the channel cost or channel bandwidth data obtained from the API interface is detected to be empty or exceeds a preset reasonable range, the historical default value of the channel is used as the channel cost or channel bandwidth.

[0022] Furthermore, in S2, the SMS indicators are standardized, including: using a linear normalization method to convert SMS indicators of different dimensions into a unified score on a percentage scale, and dividing the SMS indicators into two categories: positive indicators and negative indicators, and processing them using corresponding standardization formulas.

[0023] Furthermore, the standardized formulas for positive indicators of channel success rate and channel bandwidth are as follows:

[0024]

[0025] Among them, the larger the positive value of S_, the better the corresponding SMS indicator;

[0026] The standardized formula for using negative indices for channel latency and channel cost is as follows:

[0027]

[0028] Among them, the smaller the negative value of S_, the better the corresponding SMS indicator.

[0029] Furthermore, in S3, based on real-time monitoring of SMS channel operation data, the channel status is automatically determined. When the success rate of the target channel is lower than the trigger threshold, the channel switching mechanism is triggered to transfer the SMS sending task to the backup channel with a higher success rate.

[0030] Furthermore, in S3, the weight allocation of various SMS indicators and the trigger threshold for channel switching are based on historical data.

[0031] Furthermore, in S4, during the pre-verification phase, representative devices covering different regions and models are selected as test devices, and test SMS messages are sent to the test devices through a backup channel to verify various indicators of the backup channel.

[0032] Furthermore, in S4, when the test results of the test equipment meet the requirements of the scenario, the pre-verification is deemed successful, and a full switch of the backup channel is executed.

[0033] Furthermore, in S4, the automatic switching of the backup channel is achieved by automatically modifying the SMS sending route configuration.

[0034] The beneficial effects of this invention are as follows:

[0035] The method described in this invention can automatically monitor channel status, comprehensively evaluate channel performance, and quickly trigger optimal switching, achieving efficiency in improving switching response time without manual intervention and realizing a dynamic balance between prioritizing SMS performance and optimizing costs.

[0036] 1. This invention constructs a multi-dimensional channel comprehensive evaluation model, breaking through the limitation of existing technologies that only use "success rate" as a single indicator. This invention adopts a multi-dimensional indicator real-time collection mechanism, comprehensively considering the four core indicators of SMS channel success rate, latency, cost and bandwidth, forming a multi-dimensional data collection closed loop.

[0037] Establish a weighted comprehensive scoring model to quantitatively evaluate the overall performance of each channel, avoiding one-sided selections that "only consider success rate and ignore cost" or "only consider latency and ignore stability." This ensures that the switched channels not only meet success rate requirements but also adapt to the personalized needs of different scenarios (e.g., prioritizing low-cost channels for cost-sensitive scenarios and low-latency channels for latency-sensitive scenarios). For example, in industrial alarm scenarios, a channel with "1-second latency and 89% success rate" (higher overall score) can be prioritized over a channel with "5-second latency and 91% success rate" to meet the needs of latency-sensitive scenarios; cost-sensitive scenarios (such as ordinary equipment status feedback) can prioritize low-cost channels, reducing operating costs by 15%-20%.

[0038] Multi-dimensional metric synergistic optimization (such as success rate, cost, and latency balance) can reduce overall operating costs while ensuring communication reliability. For example, after the application of a certain IoT platform, the average annual SMS communication cost decreased by 18%, and the system stability (service interruption caused by channel failure) increased to 99.9%.

[0039] 2. This invention employs standardized processing and quantitative scoring to avoid errors from human experience-based judgment, eliminating erroneous switching (such as the abuse of high-cost channels), and improving the rationality of channel selection to over 90%. This invention distinguishes between positive indicators (success rate, bandwidth) and negative indicators (latency, cost), using a linear normalization formula to convert indicators of different dimensions into a unified score of 0-100, thus resolving the problem of incomparable indicators.

[0040] 3. This invention employs intelligent preprocessing rules for abnormal data, and formulates differentiated preprocessing strategies for different indicator characteristics to ensure data accuracy.

[0041] 4. This invention achieves dynamic optimization of the switching strategy: Based on historical communication data and real-time channel status, this invention dynamically adjusts the index weights of the weighted comprehensive scoring model and the trigger threshold for channel switching through an algorithm, thereby ensuring data integrity and reducing operating costs.

[0042] By employing a dynamic threshold and historical data feedback optimization mechanism, the evaluation model and switching strategy can adapt to the needs of different scenarios and time periods. For example, data integrity is ensured during peak hours, while costs are controlled during off-peak hours, achieving a dynamic balance between performance priority and cost optimization.

[0043] This invention employs a scenario-based dynamic weight allocation model, constructing a weight system based on the Analytic Hierarchy Process (AHP). It can adjust the weights of each indicator in real time according to the needs of IoT scenarios, achieving a precise match between the evaluation model and scenario requirements. This invention dynamically adjusts the switching trigger threshold based on the scenario and time period, and, combined with comprehensive score ranking, prioritizes the backup channel with the highest score, avoiding the limitations of a single threshold.

[0044] 5. This invention reduces communication risks after handover by incorporating a pre-verification mechanism during the handover process. Before handover, SMS messages are sent to a small number of test devices to verify the actual success rate and performance of the backup channel. This avoids handover to channels that are "highly rated on the surface but actually unusable" (e.g., channel B has a high overall score, but due to regional network limitations, its success rate for a certain batch of devices is extremely low), further ensuring communication stability after handover.

[0045] The pre-verification mechanism further ensures communication stability after handover, prevents handover to faulty backup channels, and increases the actual success rate of the handover channel to over 99%.

[0046] 6. Addressing the pain points of existing fixed channels failing to automatically respond to declining success rates and the delays in manual switching, this invention combines multi-dimensional real-time monitoring with second-level automatic switching. It monitors the success rate of each SMS channel in real time. When the success rate of the target channel falls below a preset threshold, an automatic switching mechanism is triggered, transferring the SMS sending task to a backup channel with a higher success rate without manual intervention. This achieves second-level response, reduces the SMS sending failure rate, and solves the problems of untimely manual monitoring and time-consuming switching. For example, in smart metering scenarios, it can avoid hours of data gaps caused by channel failures, reducing the SMS sending failure rate to below 5%.

[0047] This invention enables automatic switching of execution processes within seconds. After pre-verification, the system automatically modifies the SMS sending route configuration without manual intervention. The switching response time is controlled within seconds, significantly reducing the impact duration of faulty channels.

[0048] 7. A weighted comprehensive scoring algorithm is used to sort channels based on their performance, avoiding the problem of simply round-robin assigning SMS messages to channels with low success rates. For example, in a scenario where one of the three channels has a success rate of 70%, sorting by score can allocate more than 90% of SMS messages to channels with a success rate of over 95%, significantly reducing the overall failure rate. The weighted comprehensive scoring algorithm quantifies channel performance and sorts channels according to their scores, providing a quantitative basis for switching decisions.

[0049] 8. Adaptive decision-making is employed, continuously optimizing based on historical data. As usage time increases, the weights and threshold parameters become increasingly aligned with actual needs, reducing subsequent maintenance and adjustment costs and decreasing the workload of maintenance personnel in channel management by over 60%. This invention is based on historical data feedback optimization logic, continuously optimizing weight allocation and threshold parameters to enhance the adaptability of the decision-making model. Attached Figure Description

[0050] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts. It is obvious that the drawings described below are merely some embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings.

[0051] Figure 1 This is a flowchart of an intelligent disaster recovery method for SMS channels of Internet of Things (IoT) devices, according to an embodiment of the present invention. Detailed Implementation

[0052] To enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. It should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0053] Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concepts disclosed in this invention.

[0054] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicating orientation or positional relationships based on the orientation or positional relationships shown in the accompanying drawings, are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance. The terms "installed," "connected," and "linked" should be interpreted broadly; for example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0055] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of methods and systems consistent with some aspects of the invention as detailed in the appended claims.

[0056] like Figure 1 As shown in the figure, the intelligent disaster recovery method for SMS channels of IoT devices provided in this embodiment of the invention includes the following steps:

[0057] S1. Collect multi-dimensional SMS metrics for each SMS channel and preprocess the SMS metrics to remove outliers and make corrections.

[0058] Specifically, this invention adopts a multi-dimensional indicator real-time acquisition mechanism, with a basic acquisition frequency of 1 minute (which can be dynamically adjusted according to the scenario), to synchronously collect multi-dimensional SMS indicators from the SMS channel, forming a multi-dimensional data acquisition closed loop.

[0059] The multi-dimensional SMS metrics include: channel success rate (R), channel latency (T), channel cost (C), and channel bandwidth (B). The specific metrics and collection methods are shown in Table 1.

[0060] Table 1. Definition and Data Collection Method of SMS Metrics

[0061]

[0062]

[0063] Because the raw data collected may contain outliers, such as a delay of up to 100 seconds for a text message due to occasional device offlineness, or a success rate of 101% due to statistical errors, it is necessary to preprocess the text message metrics to remove outliers and make corrections.

[0064] In this invention, differentiated preprocessing strategies are formulated for different indicator characteristics, and intelligent preprocessing rules for abnormal data are adopted to ensure data accuracy.

[0065] (1) Preprocess the channel success rate, including: when the channel success rate is detected to be greater than 100%, the channel success rate is corrected to 100%; when the channel success rate is detected to be less than 0% (i.e., negative value), the channel success rate is corrected to 0%; when the number of SMS messages sent in a certain minute is less than 10, it is determined that the sample size is too small, and the average success rate of the first 5 minutes of the channel is used as the channel success rate.

[0066] (2) Preprocess the channel latency, including: removing latency anomalies that are greater than the sum of the historical average latency of the SMS channel and 3 times the standard deviation, i.e. removing latency data that are greater than "the historical average latency of the channel + 3 times the standard deviation", and then recalculating the average latency after removal.

[0067] (3) Preprocessing the channel cost and channel bandwidth, including: when the channel cost or channel bandwidth data obtained from the API interface is detected to be empty or exceeds the preset reasonable range (e.g., cost > 0.1 yuan / line, bandwidth < 100 lines / minute), the historical default value of the channel is used as the channel cost or channel bandwidth.

[0068] The following section uses a smart meter reading platform based on the Internet of Things (IoT) as an example to illustrate the process of channel data acquisition, monitoring, and preprocessing.

[0069] A certain IoT smart meter reading platform is configured with three SMS channels (channel 1, channel 2, and channel 3). The raw monitoring data for 00:05 on [Date] is shown in Table 2.

[0070] Table 2. Raw monitoring data of a certain IoT smart meter reading platform

[0071]

[0072] The preprocessing process for SMS metrics is as follows:

[0073] Channel 1: The latency of 105 seconds is an outlier (historical average latency is 3 seconds, 3 times the standard deviation is 6 seconds, 105>3+6=9). After removing it, the average latency is (2+3+4+2) / 4=2.75 seconds; the success rate is 185 / 200=92.5%.

[0074] Channel 2: No outliers, success rate = 148 / 150 ≈ 98.67%, average latency = (1+2+3+2+1) / 5 = 1.8 seconds.

[0075] Channel 3: Number of SMS messages sent: 8 < 10. The average success rate of 90% in the first 5 minutes is used as a substitute. There is no abnormal delay. The average delay is (3+4+5+3+4) / 5 = 3.8 seconds.

[0076] Table 3 shows the preprocessed data for the three channels.

[0077] Table 3. Data after preprocessing

[0078]

[0079]

[0080] S2 standardizes SMS metrics, determines the weight of each SMS metric using the analytic hierarchy process, adjusts the weight of each SMS metric in real time according to the needs of the IoT scenario, establishes a weighted comprehensive scoring model, and quantitatively evaluates the overall performance of each SMS channel to obtain a comprehensive channel score for each SMS channel.

[0081] Specifically, due to the different units of measurement for each indicator, such as channel success rate as a percentage, latency in seconds, and cost in yuan, the various SMS indicators cannot be directly compared. They need to be standardized first, converting each indicator into a unified score of 0-100 to eliminate the influence of the units of measurement.

[0082] In this invention, SMS metrics are standardized by: using a linear normalization method to convert SMS metrics of different dimensions into a unified percentage score, and dividing SMS metrics into two categories: positive metrics and negative metrics, and processing them with corresponding standardization formulas to solve the problem of incomparability of metrics.

[0083] Positive metrics include: channel success rate and channel bandwidth. The standardized formulas for positive metrics are as follows:

[0084]

[0085] The larger the value of S_positive, the better the corresponding SMS metrics. That is, the larger the value of S_positive, the better the corresponding channel success rate and channel bandwidth. If the maximum value equals the minimum value, then S_positive = 100.

[0086] Negative metrics include channel latency and channel cost. The standardized formulas for negative metrics are as follows:

[0087]

[0088] In this context, a smaller S_negative value indicates better SMS metrics. Conversely, a larger S_negative value indicates better channel latency and channel cost metrics. If the maximum value equals the minimum value, then S_negative = 100.

[0089] In this invention, after standardizing various SMS indicators, a weighted comprehensive scoring model is constructed by comprehensively considering multiple dimensions such as SMS channel success rate, channel latency, channel cost, and channel bandwidth. This model calculates the comprehensive channel score by assigning different weights to each indicator, and is used to quantitatively evaluate the overall performance of the channel.

[0090] Specifically, the weights of each indicator are determined by the Analytic Hierarchy Process (AHP), and the weights of each SMS indicator are adjusted in real time according to the needs of the IoT scenario. A weighted comprehensive scoring model is established to quantitatively evaluate the overall performance of each SMS channel and obtain a comprehensive channel score for each SMS channel.

[0091] The comprehensive scoring formula for the weighted comprehensive scoring model is as follows:

[0092] S=(S_R×W_R)+(S_T×W_T)+(S_C×W_C)+(S_B×W_B) (3)

[0093] Wherein, S represents the overall channel score, using a 0-100 point scale, with higher scores indicating better channel performance; S_R represents the standardized score for channel success rate; S_T represents the standardized score for channel latency; S_C represents the standardized score for channel cost; S_B represents the standardized score for channel bandwidth; W_R represents the weight of channel success rate; W_T represents the weight of channel latency; W_C represents the weight of channel cost; and W_B represents the weight of channel bandwidth. The sum of W_R, W_T, W_C, and W_B is 100%.

[0094] The following examples illustrate the weighting of metrics for different IoT scenarios:

[0095] Scenario 1: Smart Meter Reading

[0096] The key feature of smart meter reading lies in data uploading, requiring a high success rate, with cost being a secondary consideration. Based on these characteristics, the weights are allocated using the Analytic Hierarchy Process (AHP): channel success rate (40%), channel cost (30%), channel latency (20%), and channel bandwidth (10%).

[0097] Scenario 2: Industrial Alarms

[0098] Industrial alarms are characterized by high latency requirements, with success rate being a secondary factor. Based on these characteristics, the weights are allocated using the analytic hierarchy process as follows: channel latency (40%), channel success rate (30%), channel bandwidth (20%), and channel cost (10%).

[0099] Scenario 3: Status feedback of ordinary devices

[0100] The characteristic of status feedback for ordinary equipment is that it is sensitive to cost, and the success rate only needs to meet basic requirements. Based on the above characteristics, the weights are allocated as follows using the analytic hierarchy process: cost (40%), success rate (30%), latency (15%), and bandwidth (15%).

[0101] As can be seen from the above, this invention adopts a scenario-based dynamic weight allocation model: a weight system is constructed based on the analytic hierarchy process (AHP), which can adjust the weights of each indicator in real time according to the needs of IoT scenarios. For example, in scenarios 1 to 3 above, the weight of the same indicator is dynamically adjusted in different scenarios, achieving accurate matching between the evaluation model and scenario requirements.

[0102] Furthermore, this invention employs a weighted comprehensive scoring algorithm to sort channels based on performance, avoiding the problem of allocating SMS messages to low-success-rate channels caused by simple round-robin scheduling. For example:

[0103] 1. In a certain scenario, if one of the three channels has a success rate of 70%, more than 90% of SMS messages can be assigned to the channel with a success rate of more than 95% by sorting them by rating, which will greatly reduce the overall failure rate.

[0104] 2. In industrial alarm scenarios, the channel with "1 second latency and 89% success rate" (higher overall score) should be selected first, rather than the channel with "5 seconds latency and 91% success rate", to meet the needs of latency-sensitive scenarios.

[0105] 3. In cost-sensitive scenarios (such as status feedback for ordinary equipment), low-cost channels can be prioritized to reduce operating costs by 15%-20%.

[0106] The following example uses a smart meter reading scenario, with the following weight allocation: W_R = 40%, W_W_T = 20%, W_C = 30%, W_B = 10%. A comprehensive score is calculated based on the preprocessed channel data.

[0107] Step 1: Calculate the maximum and minimum values ​​of each SMS indicator (based on data from 3 channels), as shown in Table 4.

[0108] Table 4 shows the maximum and minimum values ​​of each SMS indicator.

[0109] SMS metrics Maximum value Minimum value R(％) 98.67 90 T (seconds) 3.8 1.8 C (yuan / item) 0.06 0.04 B (items / minute) 300 180

[0110] Step 2: Standardized score calculation:

[0111] Channel 1:

[0112] S_R=(92.5-90) / (98.67-90)×100≈2.5 / 8.67×100≈28.84;

[0113] S_T=(3.8-2.75) / (3.8-1.8)×100≈1.05 / 2×100=52.5;

[0114] S_C=(0.06-0.05) / (0.06-0.04)×100=0.01 / 0.02×100=50;

[0115] S_B=(250-180) / (300-180)×100=70 / 120×100≈58.33.

[0116] Channel 2:

[0117] S_R=(98.67-90) / (98.67-90)×100=100;

[0118] S_T=(3.8-1.8) / (3.8-1.8)×100=100;

[0119] S_C=(0.06-0.06) / (0.06-0.04)×100=0;

[0120] S_B=(300-180) / (300-180)×100=100.

[0121] Channel 3:

[0122] S_R=(90-90) / (98.67-90)×100=0;

[0123] S_T=(3.8-3.8) / (3.8-1.8)×100=0;

[0124] S_C=(0.06-0.04) / (0.06-0.04)×100=10.

[0125] S3 dynamically adjusts the trigger threshold for channel switching based on the IoT scenario and time period, and prioritizes the SMS channel with the highest score as the backup channel, based on the ranking of the comprehensive channel scores; where the trigger threshold for channel switching is the threshold of the SMS indicator.

[0126] This invention employs a dynamic threshold triggering mechanism: dynamically adjusting the switching trigger threshold according to the scene and time period.

[0127] For example, when the IoT scenario is detected as smart meter reading, the trigger threshold for channel switching is set to the threshold of channel success rate, and the trigger threshold for channel switching during peak hours is set higher than the trigger threshold for channel switching during off-peak hours.

[0128] For example, during peak data upload times for smart meter reading (e.g., 0-2 AM), the weight of channel success rate is increased, setting the success rate threshold to 95% (while lowering the switching threshold (e.g., switching when the success rate falls below 95%)) to ensure data integrity. During off-peak times (e.g., 10-12 AM), the weight of channel cost is increased, and the switching threshold is appropriately raised (e.g., switching only when the success rate falls below 85%) to reduce operating costs. Simultaneously, a comprehensive scoring ranking is used to prioritize the highest-scoring backup channel, avoiding the limitations of a single threshold.

[0129] This invention can perform intelligent disaster recovery switching: the trigger threshold for channel switching is set to the threshold of channel success rate.

[0130] Based on real-time monitoring of SMS channel operation data, the algorithm automatically judges the channel status. When the success rate of the target channel is lower than the channel success rate threshold, the channel switching mechanism is triggered to transfer the SMS sending task to the backup channel with a higher success rate, thus ensuring communication continuity.

[0131] Furthermore, this invention assigns weights to various SMS metrics and sets trigger thresholds for channel switching based on historical data.

[0132] Specifically, based on historical communication data (such as success rate and cost consumption after switching) and real-time channel status, the weights of the indicators in the weighted comprehensive scoring model and the trigger threshold for channel switching are dynamically adjusted through algorithms to continuously optimize the weight allocation and threshold parameters.

[0133] Historical data feedback optimization logic: Based on historical communication data, such as if the cost exceeds expectations after multiple switching of a certain channel, the weight of the channel cost indicator in this scenario is reduced to improve the adaptability of the decision model.

[0134] S4 introduces a pre-verification mechanism during the channel switching process. When the trigger threshold for channel switching is detected, an SMS is sent to the test device to pre-verify the backup channel before the formal channel switching. After the pre-verification is passed, the formal channel switching is performed, and the entire system is switched to the backup channel to achieve disaster recovery switching.

[0135] In this invention, a pre-verification mechanism is adopted: before the switchover, test SMS messages are sent to a small number of test devices to verify the actual success rate, latency and other indicators of the backup channel. The full switchover is only performed when the test results meet the scenario requirements, thereby avoiding the risk of "high surface score but actual unusable".

[0136] Specifically, in the pre-verification phase, a small number of representative devices covering different regions and models are selected as test devices. Test SMS messages are sent to the test devices through a backup channel to verify various indicators of the backup channel.

[0137] When the test results of the test equipment meet the requirements of the scenario (such as channel success rate ≥ threshold, channel latency ≤ scenario upper limit), the pre-verification is considered successful and a full switch of the backup channel is executed.

[0138] This invention achieves a second-level automatic switching execution process: after pre-verification, the backup channel is automatically switched over by automatically modifying the SMS sending route configuration, without manual intervention. The switching response time is controlled within seconds, which greatly shortens the impact time of the faulty channel.

[0139] In summary, the disaster recovery switchover process of this invention is as follows:

[0140] 1. First, determine whether the trigger threshold is met;

[0141] 2. If the trigger threshold is met, select the channel with the higher overall score as the backup channel, based on the channel's overall score.

[0142] 3. After determining the backup channel, perform pre-verification first, and then perform a full switchover after the pre-verification is successful.

[0143] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the embodiments of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent disaster recovery of SMS channels for Internet of Things (IoT) devices, characterized in that, include: S1. Collect multi-dimensional SMS metrics for each SMS channel and preprocess the SMS metrics to remove outliers and make corrections. S2, the SMS metrics are standardized, and the weights of each SMS metric are determined using the analytic hierarchy process (AHP). The weights of each SMS metric are adjusted in real time according to the needs of the IoT scenario to establish a weighted comprehensive scoring model. This model is used to quantitatively evaluate the overall performance of each SMS channel, resulting in a comprehensive channel score for each SMS channel. The weighted comprehensive scoring model is as follows: S=(S_R×W_R)+(S_T×W_T)+(S_C×W_C)+(S_B×W_B) Where: S is the channel comprehensive score, S_R is the standardized score of channel success rate, S_T is the standardized score of channel latency, S_C is the standardized score of channel cost, S_B is the standardized score of channel bandwidth, W_R is the weight of channel success rate, W_T is the weight of channel latency, W_C is the weight of channel cost, and W_B is the weight of channel bandwidth. S3. Based on the IoT scenario and time period, dynamically adjust the trigger threshold for channel switching, and combine the ranking of the comprehensive channel scores to prioritize the SMS channel with the highest score as the backup channel; wherein, the trigger threshold for channel switching is the threshold of the SMS index. S4 introduces a pre-verification mechanism during the channel switching process. When the trigger threshold for channel switching is detected, an SMS is sent to the test device to pre-verify the backup channel before the formal channel switching. After the pre-verification is passed, the formal channel switching is performed, and the entire system is switched to the backup channel to achieve disaster recovery switching.

2. The intelligent disaster recovery method for SMS channels of IoT devices according to claim 1, characterized in that, In S1, the multi-dimensional SMS metrics include: channel success rate, channel latency, channel cost, and channel bandwidth.

3. The intelligent disaster recovery method for SMS channels of IoT devices according to claim 2, characterized in that, In S2, the channel success rate is preprocessed, including: when the channel success rate is detected to be greater than 100%, the channel success rate is corrected to 100%; when the channel success rate is detected to be less than 0%, the channel success rate is corrected to 0%; when the number of SMS messages sent in one minute is less than 10, the average success rate of the channel in the previous 5 minutes is used as the channel success rate. The channel latency is preprocessed, including: removing latency anomalies that are greater than the sum of the historical average latency of the SMS channel and 3 times the standard deviation, and then recalculating the average latency after the removal. The channel cost and channel bandwidth are preprocessed, including: when the channel cost or channel bandwidth data obtained from the API interface is detected to be empty or exceeds a preset reasonable range, the historical default value of the channel is used as the channel cost or channel bandwidth.

4. The intelligent disaster recovery method for SMS channels of IoT devices according to claim 1, characterized in that, In S2, the SMS indicators are standardized, including: using a linear normalization method to convert SMS indicators of different dimensions into a unified score of 100, and dividing the SMS indicators into two categories: positive indicators and negative indicators, and processing them with corresponding standardization formulas.

5. The intelligent disaster recovery method for SMS channels of IoT devices according to claim 4, characterized in that, The standardized formula for positive metrics of channel success rate and channel bandwidth is as follows: Among them, the larger the positive value of S_, the better the corresponding SMS indicator; The standardized formula for using negative indices for channel latency and channel cost is as follows: Among them, the smaller the negative value of S_, the better the corresponding SMS indicator.

6. The intelligent disaster recovery method for SMS channels of IoT devices according to claim 1, characterized in that, In S3, based on real-time monitoring of SMS channel operation data, the channel status is automatically determined. When the success rate of the target channel is lower than the trigger threshold, the channel switching mechanism is triggered to transfer the SMS sending task to the backup channel with a higher success rate.

7. The intelligent disaster recovery method for SMS channels of IoT devices according to claim 1, characterized in that, In S3, the weight allocation of various SMS indicators and the trigger threshold for channel switching are based on historical data.

8. The intelligent disaster recovery method for SMS channels of IoT devices according to claim 1, characterized in that, In S4, during the pre-verification phase, representative devices covering different regions and models are selected as test devices. Test SMS messages are sent to the test devices through a backup channel to verify various indicators of the backup channel.

9. The intelligent disaster recovery method for SMS channels of IoT devices according to claim 1, characterized in that, In S4, when the test results of the test equipment meet the requirements of the scenario, the pre-verification is considered successful and a full switchover of the backup channel is executed.

10. The intelligent disaster recovery method for SMS channels of IoT devices according to claim 9, characterized in that, In S4, the backup channel is automatically switched over in full by automatically modifying the SMS sending route configuration.

Images