Intelligent switching method of ship communication links based on intelligent analysis of multi-link states

Abstract

The present application relates to the field of wireless communication network technology, and specifically to a method for intelligent switching of ship communication links based on intelligent analysis of multi-link states, including: collecting the packet loss rate, signal strength and delay of each communication link, constructing the packet loss rate time series, signal strength time series, delay time series and service quality time series of each communication link at each sampling moment, obtaining the steady-state index of packet loss rate, signal strength stability reliability, and docking delay impact coefficient, calculating the communication interference index and communication link quality prediction adjustment parameters, obtaining the predicted service quality, determining the alternative switching link, and realizing the intelligent switching of ship communication links. The present application aims to solve the problem that the traditional exponential moving average algorithm does not take into account that the service quality of the ship communication link is affected by many factors, and the fixed smoothing coefficient is used to cause inaccurate prediction of the service quality of the communication link.

Description

Technical Field

[0001] The present application relates to the technical field of wireless communication networks, and in particular to a method for intelligent switching of ship communication links based on intelligent analysis of multi-link states. Background Art

[0002] With the development of science and technology, the connection methods used for ship communications are gradually increasing, such as VSAT, Iridium, 4G / 5G, etc. Different communication connection methods have different advantages, but they are also subject to certain limitations. For example, when the 4G / 5G communication equipment is far away from the base station, the signal strength will be significantly weakened; VSAT, Iridium and other satellite communications will be affected by the location of the satellite, the severity of the weather, etc., resulting in communication obstruction. During the navigation of the ship, there are inevitably various adverse communication conditions. In order to ensure smooth communication on the ship, it is necessary to selectively switch among multiple communication links.

[0003] The quality of service in a network refers to the ability of a network to provide better service for designated network communications using various basic technologies. Usually, the quality of service (QoS) is used to measure the quality of a communication link. The higher the quality of service of a communication link, the more suitable it is for communication. During the voyage of a ship, in order to switch to a better communication link faster and more accurately, the EMA exponential moving average algorithm is used to predict the service quality of each communication link, so that when switching is needed, it can switch to a better communication link to ensure smooth communication.

[0004] However, during the navigation of a ship, there are many internal and external factors that affect it. The use of a fixed smoothing coefficient does not take into account the changes in the communication link during the navigation of the ship, resulting in inaccurate prediction of the service quality of the communication link, which in turn has a certain impact on the switching of the link. Summary of the invention

[0005] In order to solve the above technical problems, the present application provides a method for intelligent switching of ship communication links based on intelligent analysis of multi-link states to solve the existing problems.

[0006] The ship communication link intelligent switching method based on multi-link status intelligent analysis of the present application adopts the following technical solutions:

[0007] An embodiment of the present application provides a method for intelligent switching of ship communication links based on intelligent analysis of multi-link states, the method comprising the following steps:

[0008] Collect the packet loss rate, signal strength and delay of each communication link, and construct the packet loss rate time series, signal strength time series, delay time series and service quality time series of each communication link at each sampling time;

[0009] Obtain the packet loss rate steady-state index according to the packet loss rate time series; obtain the signal strength steady reliability according to the signal strength time series; obtain the communication interference index according to the packet loss rate steady-state index and the signal strength steady reliability;

[0010] Obtain the delay peak interval of each peak point in the delay fitting curve according to the delay time series; obtain the docking delay influence coefficient according to the delay time series and the delay peak interval of all peak points in the delay fitting curve; obtain the communication prediction external interference coefficient according to the communication interference index of all communication links at each sampling time; obtain the communication link quality prediction adjustment parameter according to the docking delay influence coefficient and the communication prediction external interference coefficient;

[0011] The predicted service quality of each communication link at the next sampling moment is obtained according to the communication link quality prediction adjustment parameters of each communication link at each sampling moment and the service quality time series; the alternative switching link is obtained according to the predicted service quality of each communication link at the next sampling moment.

[0012] Optionally, the method of obtaining the packet loss rate steady-state index according to the packet loss rate time series includes:

[0013] The first-order difference of the packet loss rate time series is taken as the packet loss rate difference sequence, and all mutation points in the packet loss rate difference sequence are extracted. The difference between the sampling time of each mutation point in the packet loss rate difference sequence and the adjacent previous mutation point is recorded as the packet loss rate mutation interval of each mutation point in the packet loss rate difference sequence. The average value of the packet loss rate mutation interval of all mutation points in the packet loss rate difference sequence is recorded as the packet loss rate mutation interval index.

[0014] Calculate the average value of all data in the packet loss rate time series, and record the sum of squares of differences between all data values ​​in the packet loss rate time series and the average value as a packet loss rate floating index;

[0015] The sum of the packet loss rate floating index and the preset parameter adjustment factor is recorded as the adjusted packet loss floating index, and the ratio of the packet loss rate mutation interval index to the adjusted packet loss floating index is taken as the packet loss rate steady-state index.

[0016] Optionally, the obtaining of the signal strength stability reliability according to the signal strength time series includes the following specific methods:

[0017] Perform curve fitting on the signal strength time series, output the signal strength fitting curve, and obtain all the peak points in the signal strength fitting curve;

[0018] The difference between the sampling time of each peak point and the adjacent previous peak point in the signal strength fitting curve is recorded as the signal peak interval of each peak point in the signal strength fitting curve, and the average value of the signal peak intervals of all peak points in the signal strength fitting curve is recorded as the signal fluctuation interval index;

[0019] The minimum value of all data in the signal strength time series is taken as the reference signal strength, the absolute value of the difference between each data value in the signal strength time series and the reference signal strength is recorded as the relative signal difference of each data in the signal strength time series, and the sum of the relative signal differences of all data in the signal strength time series is recorded as the signal floating index;

[0020] The approximate entropy of the signal strength time series is multiplied by the signal fluctuation index as the signal fluctuation complexity index, the sum of the signal fluctuation complexity index and the preset parameter adjustment factor is recorded as the adjusted signal fluctuation index, the product of the average value of all data in the signal strength time series and the signal fluctuation interval index is calculated, and the product is divided by the adjusted signal fluctuation index as the signal strength stability reliability.

[0021] Optionally, the communication interference index is the reciprocal of the sum of the packet loss rate steady-state index and the signal strength steady-state reliability.

[0022] Optionally, the method of obtaining the delay peak interval of each peak point in the delay fitting curve according to the delay time series includes the following specific methods:

[0023] Perform curve fitting on the delay time series, output the delay fitting curve, obtain all the peak points in the delay fitting curve, and record the absolute value of the difference between the sampling time of each peak point in the delay fitting curve and the adjacent subsequent peak point as the delay peak interval of each peak point in the delay fitting curve.

[0024] Optionally, the obtaining of the connection delay impact coefficient includes a specific method as follows:

[0025] Calculate the average value of the delay peak intervals of all peak points in the delay fitting curve, calculate the ratio of the delay peak interval of each peak point in the delay fitting curve to the average value, record the absolute value of the difference between the value 1 and the ratio as the delay interval relative index of each peak point in the delay fitting curve, and record the product of the delay interval relative index of each peak point in the delay fitting curve and the delay peak interval as the delay peak regularity index;

[0026] The average value of the delay peak law index of all peak points in the delay fitting curve is recorded as the peak law average coefficient, and an exponential function with a natural constant as the base and the opposite of the peak law average coefficient as the exponent is calculated. The calculation result of the exponential function is multiplied by the average value of all data in the delay time series to obtain the docking delay influence coefficient.

[0027] Optionally, obtaining a communication prediction external interference coefficient includes:

[0028] For each communication link, the sum of the absolute values ​​of the differences between the communication interference indexes of each communication link and the other communication links at each sampling time is recorded as the interference difference coefficient of each communication link at each sampling time;

[0029] The average value of the communication interference index of all communication links at each sampling moment is calculated, and the sum of the interference difference coefficient of each communication link at each sampling moment and the preset parameter adjustment factor is recorded as the adjusted interference difference, and the ratio of the average value to the adjusted interference difference is used as the communication prediction external interference coefficient.

[0030] Optionally, the acquiring of the communication link quality prediction adjustment parameter includes:

[0031] An exponential function with a natural constant as base and an inverse of the connection delay influence coefficient as exponent is calculated, and a normalized value of the product of the communication prediction external interference coefficient and the calculation result of the exponential function is used as a communication link quality prediction adjustment parameter.

[0032] Optionally, the obtaining of the predicted service quality of each communication link at the next sampling moment includes the following specific methods:

[0033] The communication link quality prediction adjustment parameter of each communication link at each sampling moment is used as the smoothing coefficient of the exponential moving average method, and the service quality time series of each communication link at each sampling moment is used as the input of the exponential moving average method, and the predicted service quality of each communication link at the next sampling moment is output.

[0034] Optionally, the obtaining of the candidate switching link includes a specific method as follows:

[0035] Sort the predicted service qualities from large to small, and select the communication links corresponding to the first w predicted service qualities as the candidate switching links at each sampling moment, where w is the number of pre-selected links;

[0036] When the service quality of the current communication link is greater than or equal to the minimum value of the predicted service qualities of all alternative switching links, continue to use the current communication link; when the service quality of the current communication link is less than the minimum value of the predicted service qualities of all alternative switching links, switch the current communication link to the alternative communication link with the highest predicted service quality.

[0037] The beneficial effects of the present application are as follows: first, a steady-state index of packet loss rate and a stable reliability of signal strength are obtained according to a packet loss rate time series and a signal strength time series, and a communication interference index is calculated to characterize the interference of a communication link of a ship during navigation, and then the causes of fluctuations in the communication link are further analyzed, and a communication prediction external interference coefficient is determined according to the communication interference index of all communication links at each sampling moment, and a docking delay influence coefficient is obtained according to a delay time series to distinguish the influence of regular fluctuations caused by docking work of staff members during navigation and irregular fluctuations caused by concentrated use of the same communication link by other personnel on the communication link, and a communication link quality prediction adjustment parameter is calculated in combination with the communication prediction external interference coefficient as a smoothing coefficient of an exponential moving average method, and a predicted service quality at the next sampling moment is obtained according to a service quality time series, and the smoothing coefficient of the exponential moving average method is adaptively set according to the cause of fluctuations in the communication link, thereby improving the accuracy of predicting the service quality of the communication link, and obtaining an alternative switching link at each sampling moment according to the predicted service quality of all communication links at the next sampling moment, thereby realizing intelligent switching of ship communication links and improving the efficiency of intelligent switching of ship communication links. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0039] Figure 1 A schematic flow chart of a method for intelligent switching of ship communication links based on intelligent analysis of multi-link states provided in one embodiment of the present application;

[0040] Figure 2 Schematic diagram for obtaining predicted service quality. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0042] See also Figure 1 , which shows a flow chart of a method for intelligent switching of ship communication links based on intelligent analysis of multi-link states provided by an embodiment of the present application, the method comprising the following steps:

[0043] Step S001: Collect the packet loss rate, signal strength and delay of each communication link, and construct the packet loss rate time series, signal strength time series, delay time series and service quality time series of each communication link at each sampling time.

[0044] The service quality of a communication link depends on the packet loss rate, signal strength, delay, etc. This embodiment uses the packet loss rate, signal strength, and delay of a communication link as an example for analysis. In the signal receiving device on the ship, the network monitoring software in the signal receiving device is used to collect the packet loss rate, signal strength, and delay of each communication link once every interval T. In order to unify the dimensions, the packet loss rate, signal strength, and delay of each communication link are normalized respectively, and the service quality at each sampling moment is expressed as the signal strength divided by the product of the packet loss rate and the delay, that is, the service quality , where a, b, and c are packet loss rate, signal strength, and delay, respectively.

[0045] The packet loss rates of each communication link at each sampling moment and the n-1 sampling moments before it are arranged in ascending order according to the time sequence of collection, and a packet loss rate time series at each sampling moment is constructed. In order to prevent data loss during transmission, this embodiment uses a mean filling method to fill the packet loss rate time series with data. Mean filling is a well-known technology and will not be described in detail in this embodiment.

[0046] The acquisition methods of the signal strength time series, the delay time series, the service quality time series and the packet loss rate time series are the same. In this embodiment, n is 50 and T is 0.5 seconds.

[0047] Step S002: Obtain the steady-state index of packet loss rate and the stable reliability of signal strength according to the packet loss rate time series and the signal strength time series, calculate the communication interference index, determine the communication prediction external interference coefficient, obtain the docking delay influence coefficient according to the delay time series, calculate the communication link quality prediction adjustment parameter in combination with the communication prediction external interference coefficient as the smoothing coefficient of the exponential moving average method, and obtain the predicted service quality at the next sampling moment according to the service quality time series.

[0048] According to the differences in the peak distribution of packet loss rate and communication intensity of the ship during navigation, a communication interference index is constructed; according to the changing state of the transmission queuing delay of each communication link when the crew and passengers communicate during the navigation, the docking delay influence coefficient is constructed, and the communication link quality prediction adjustment parameters are further constructed. The smoothing coefficient in the EMA algorithm is improved, and the EMA prediction algorithm is used to predict the service quality time series of each communication link.

[0049] During the voyage of a ship, each communication link will produce different service quality changes due to different situations. When the ship encounters weather changes, areas with weak satellite signal coverage, etc. during the voyage, the service quality on the ship may decrease, which in turn leads to increased delays or even communication interruptions. At this time, it is necessary to replace the communication link based on the actual service quality of each link to ensure the service quality to the greatest extent. However, when the EMA prediction algorithm is used to predict the link service quality under normal circumstances, it is difficult to consider the reasons that really affect the service quality, which leads to a certain degree of deviation in the prediction. Therefore, it is necessary to make certain corrections based on the actual situation when making the prediction.

[0050] There are many factors that affect the service quality of the communication link during the navigation of the ship, and when it is affected, it will change to a certain extent within the time range of the influence of the factor. Perform a first-order difference on the packet loss rate time series, construct a packet loss rate difference sequence, and use the Pettitt mutation point detection algorithm to extract all mutation points in the packet loss rate difference sequence. Use the least squares method to perform curve fitting on the signal strength time series, output the signal strength fitting curve, and obtain all the peak points in the signal strength fitting curve. Approximate entropy is used to measure the irregularity and complexity of the signal sequence. In order to analyze the regularity of the signal strength change, the approximate entropy of the signal strength time series is calculated, where the Pettitt mutation point detection algorithm, the least squares method, and the calculation of approximate entropy are all well-known technologies, which will not be described in detail in this embodiment. Construct a communication interference index:

[0051] ;

[0052] ;

[0053] In the formula, is the communication interference index of the mth communication link at the i-th sampling time, is the steady-state index of packet loss rate of the mth communication link at the i-th sampling time, is the total number of mutation points in the packet loss rate difference sequence of the mth communication link at the i-th sampling time, , are the sampling times of the jth mutation point and the j-1th mutation point in the packet loss rate difference sequence of the mth communication link at the ith sampling time, respectively. is the yth data value in the packet loss rate time series of the mth communication link at the ith sampling time, is the total number of data in each packet loss rate time series and signal strength time series, is the average value of all data in the packet loss rate time series of the mth communication link at the i-th sampling time, is a parameter adjustment factor greater than 0, the purpose is to prevent the denominator from being 0. The value is 0.001.

[0054] is the signal strength stability reliability of the mth communication link at the i-th sampling time, is the average value of all data in the signal strength time series of the mth communication link at the i-th sampling time, , are the sampling moments of the xth and x-1th peak points in the signal strength fitting curve of the mth communication link at the ith sampling moment, respectively. is the total number of peak points in the signal strength fitting curve of the mth communication link at the i-th sampling time, is the approximate entropy of the signal strength time series of the mth communication link at the i-th sampling time, is the vth data value in the signal strength time series of the mth communication link at the ith sampling moment, It is the minimum value of all data in the signal strength time series of the mth communication link at the i-th sampling moment.

[0055] When the time interval between the sampling moments of adjacent mutation points in the packet loss rate difference sequence is larger, and the data in the packet loss rate time series fluctuates around its mean value, The bigger, The smaller it is, the smaller the overall fluctuation of the packet loss rate of the communication link during navigation, the smaller the interference, and the larger the steady-state index value of the packet loss rate; when the average value of all data in the signal strength time series is larger, the time interval between adjacent peak points in the signal strength fitting curve is larger, the difference between each data in the signal strength time series and its minimum value is smaller, and the approximate entropy of the signal strength time series is smaller, and The bigger, and The smaller it is, the stronger the signal is, the less fluctuation of signal strength is, the gentler the fluctuation amplitude is, the more regular the fluctuation of signal strength is, and the larger the signal strength stability reliability value is; when the packet loss rate steady-state index and signal strength stability reliability are larger, it means that the packet loss rate of the communication link is less disturbed during the ship's navigation, the fluctuation of signal strength is more stable, and the communication interference index value of the communication link is smaller.

[0056] So far, the communication interference index of each communication link at the i-th sampling time is calculated by the above method.

[0057] The service quality of the communication link is not only affected by the communication link itself, but also by certain human factors. For example, during the voyage of a ship, the crew on the ship will use communication equipment to transmit various navigation information for navigation information docking at certain specified times. However, due to the large amount of content transmitted, a certain load pressure will be generated on the communication link used. At this time, if other crew members or passengers use the communication link for communication, it may result in a long response time and large delay. Therefore, in the process of considering link switching, in addition to considering the impact of the communication link itself, it is also necessary to consider the congestion caused by human factors on the target link when switching the link.

[0058] Specifically, the congestion of the communication link is mainly manifested as an increase in the transmission queuing delay. Since there is a prescribed docking time for the information transmission of the staff during the voyage, the increase in the transmission queuing delay will also have a certain periodicity. However, in addition to the docking work of the crew, the congestion of the communication link may also be caused by the fact that other personnel on the ship use the same communication link for communication, which leads to congestion. Finally, the transmission queuing delay presents an irregular multi-peak situation, that is, when the transmission queuing delay peak in the communication link appears more regularly, it is more likely that it is the communication link used by the crew for docking work, and the less regular the peak is, it is more likely that other crew members or passengers on the ship concentrate on using the communication link for communication, resulting in congestion.

[0059] In order to distinguish the regular fluctuations caused by the staff's docking work from the irregular fluctuations caused by other personnel using the same communication link, the least squares method is used to perform curve fitting on the delay time series, output the delay fitting curve, obtain all the peak points in the delay fitting curve, and record the absolute value of the difference between the sampling time of each peak point in the delay fitting curve and the adjacent subsequent peak point as the delay peak interval of each peak point in the delay fitting curve. Among them, the least squares method is a well-known technology and will not be described in detail in this embodiment. Construct communication link quality prediction adjustment parameters:

[0060] ;

[0061] ;

[0062] In the formula, The communication link quality prediction adjustment parameter for the mth communication link at the i-th sampling time, is the normalization function, is the connection delay influence coefficient of the mth communication link at the i-th sampling time, is the average value of all data in the delay time series of the mth communication link at the i-th sampling time, is the total number of peak points in the delay fitting curve of the mth communication link at the i-th sampling time, is the delay peak interval of the rth peak point in the delay fitting curve of the mth communication link at the i-th sampling time, is the average value of the delay peak intervals of all peak points in the delay fitting curve of the mth communication link at the i-th sampling time, is an exponential function with a natural constant as its base.

[0063] Predict the external interference coefficient for the communication of the mth communication link at the i-th sampling time, is the average value of the communication interference index of all communication links at the i-th sampling time, , are the communication interference indexes of the mth and pth communication links at the i-th sampling time, respectively, M is the total number of communication links, is a parameter adjustment factor greater than 0, the purpose is to prevent the denominator from being 0. The value is 0.001.

[0064] When the average value of all data in the delay time series is larger, and the delay peak interval of each peak point in the delay fitting curve is smaller and closer to its average value, it means that the delay of the communication link has more regular peak conditions, and it is more likely to be used frequently for docking work by crew members, and the docking delay impact coefficient value is larger; when the average value of the communication interference index of all communication links is larger, and the difference in the communication interference index of each communication link is smaller, The bigger, The smaller it is, the more likely the communication link is simultaneously subject to external interference and the higher the degree of interference, the more unstable the service quality of the communication link, and the larger the value of the communication prediction external interference coefficient.

[0065] When the docking delay influence coefficient is larger and the communication prediction external interference coefficient is smaller, it means that the communication link is more likely to be used for crew docking and the communication link is less subject to external interference during the current voyage. The fluctuation of the service quality of the communication link is more regular and the fluctuation degree is smaller during the current voyage. When making predictions, a smaller smoothing coefficient is set to reduce the correction amplitude so that the prediction model can contain information of a longer time series. When the docking delay influence coefficient is smaller and the communication prediction external interference coefficient is larger, it means that the communication link is less likely to be used for crew docking but the communication link is subject to greater external interference during the current voyage. The fluctuation of the service quality of the communication link is more irregular and the fluctuation degree is larger during the current voyage. When making predictions, a larger smoothing coefficient should be set to improve the sensitivity of the prediction model.

[0066] The communication link quality prediction adjustment parameter of each communication link at each sampling time is used as the smoothing coefficient of the EMA exponential moving average method, and the service quality time series of each communication link at each sampling time is used as the input of the EMA exponential moving average method, and the predicted service quality of each communication link at the next sampling time is output. Among them, the EMA exponential moving average method is a well-known technology and will not be described in detail in this embodiment. The schematic diagram of obtaining the predicted service quality is shown in FIG. Figure 2 shown.

[0067] At this point, the predicted service quality of each communication link at the next sampling moment is obtained.

[0068] Step S003: According to the predicted service quality of each communication link at the next sampling moment, the alternative switching link at each sampling moment is obtained to realize intelligent switching of the ship communication link.

[0069] Sort the predicted service qualities from large to small, and select the communication links corresponding to the first w predicted service qualities as the candidate switching links at each sampling time, where: , M is the total number of communication links, is the floor function.

[0070] When the service quality of the current communication link is greater than or equal to the minimum value of the predicted service quality of all alternative switching links, the current communication link continues to be used; when the service quality of the current communication link is less than the minimum value of the predicted service quality of all alternative switching links, the current communication link is switched to the alternative communication link with the highest predicted service quality, thereby realizing intelligent switching of ship communication links.

[0071] It should be noted that the above sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. The above is a description of a specific embodiment of this specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0072] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0073] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for intelligent switching of ship communication links based on intelligent analysis of multi-link states, characterized in that: The method comprises the following steps: Collect the packet loss rate, signal strength and delay of each communication link, and construct the packet loss rate time series, signal strength time series, delay time series and service quality time series of each communication link at each sampling time; Obtain the packet loss rate steady-state index according to the packet loss rate time series; obtain the signal strength steady reliability according to the signal strength time series; obtain the communication interference index according to the packet loss rate steady-state index and the signal strength steady reliability; Obtain the delay peak interval of each peak point in the delay fitting curve according to the delay time series; obtain the docking delay influence coefficient according to the delay time series and the delay peak interval of all peak points in the delay fitting curve; obtain the communication prediction external interference coefficient according to the communication interference index of all communication links at each sampling time; obtain the communication link quality prediction adjustment parameter according to the docking delay influence coefficient and the communication prediction external interference coefficient; Obtain the predicted service quality of each communication link at the next sampling moment according to the communication link quality prediction adjustment parameter of each communication link at each sampling moment and the service quality time series; obtain the alternative switching link according to the predicted service quality of each communication link at the next sampling moment; The expression for obtaining the steady-state index of packet loss rate according to the packet loss rate time series is: In the formula, is the steady-state index of packet loss rate of the mth communication link at the i-th sampling time, is the total number of mutation points in the packet loss rate difference sequence of the mth communication link at the i-th sampling time, are the sampling times of the jth mutation point and the j-1th mutation point in the packet loss rate difference sequence of the mth communication link at the ith sampling time, respectively. is the yth data value in the packet loss rate time series of the mth communication link at the ith sampling moment, n is the total number of data in each packet loss rate time series and signal strength time series, is the average value of all data in the packet loss rate time series of the mth communication link at the i-th sampling time, and β is a parameter adjustment factor greater than 0; The expression for obtaining the signal strength stability reliability according to the signal strength time series is: In the formula, is the signal strength stability reliability of the mth communication link at the i-th sampling time, is the average value of all data in the signal strength time series of the mth communication link at the i-th sampling time, are the sampling moments of the xth and x-1th peak points in the signal strength fitting curve of the mth communication link at the ith sampling moment, respectively. is the total number of peak points in the signal strength fitting curve of the mth communication link at the i-th sampling time, is the approximate entropy of the signal strength time series of the mth communication link at the i-th sampling time, is the vth data value in the signal strength time series of the mth communication link at the ith sampling moment, is the minimum value of all data in the signal strength time series of the mth communication link at the i-th sampling time; The expression of the communication interference index is: In the formula, is the communication interference index of the mth communication link at the i-th sampling time; The expression of the connection delay influence coefficient is: In the formula, is the connection delay influence coefficient of the mth communication link at the i-th sampling time, is the average value of all data in the delay time series of the mth communication link at the i-th sampling time, is the total number of peak points in the delay fitting curve of the mth communication link at the i-th sampling time, is the delay peak interval of the rth peak point in the delay fitting curve of the mth communication link at the i-th sampling time, is the average value of the delay peak intervals of all peak points in the delay fitting curve of the mth communication link at the i-th sampling time, and exp() is an exponential function with a natural constant as the base; The expression of the communication prediction external interference coefficient is: In the formula, Predict the external interference coefficient for the communication of the mth communication link at the i-th sampling time, is the average value of the communication interference index of all communication links at the i-th sampling time, is the communication interference index of the pth communication link at the i-th sampling time, and M is the total number of communication links; The expression of the communication link quality prediction adjustment parameter is: In the formula, is the communication link quality prediction adjustment parameter of the mth communication link at the ith sampling time, norm() is the normalization function; The specific method of obtaining the predicted service quality of each communication link at the next sampling moment includes: The communication link quality prediction adjustment parameter of each communication link at each sampling moment is used as the smoothing coefficient of the exponential moving average method, and the service quality time series of each communication link at each sampling moment is used as the input of the exponential moving average method, and the predicted service quality of each communication link at the next sampling moment is output.

2. The method for intelligent switching of ship communication links based on intelligent analysis of multi-link states according to claim 1 is characterized in that: The specific method of obtaining the delay peak interval of each peak point in the delay fitting curve according to the delay time series includes: Perform curve fitting on the delay time series, output the delay fitting curve, obtain all the peak points in the delay fitting curve, and record the absolute value of the difference between the sampling time of each peak point in the delay fitting curve and the adjacent subsequent peak point as the delay peak interval of each peak point in the delay fitting curve.

3. The method for intelligent switching of ship communication links based on intelligent analysis of multi-link states according to claim 1 is characterized in that: The specific method of obtaining the candidate switching link includes: Sort the predicted service qualities from large to small, and select the communication links corresponding to the first w predicted service qualities as the candidate switching links at each sampling moment, where w is the number of pre-selected links; When the service quality of the current communication link is greater than or equal to the minimum value of the predicted service qualities of all alternative switching links, continue to use the current communication link; when the service quality of the current communication link is less than the minimum value of the predicted service qualities of all alternative switching links, switch the current communication link to the alternative communication link with the highest predicted service quality.

Images