Frequency hopping frequency point prediction method and device based on adjoint model, and electronic equipment

Abstract

The invention provides a frequency hopping frequency point prediction method and device based on an adjoint model, and electronic equipment. The method comprises the following steps: obtaining a to-be-predicted frequency hopping sequence; intercepting the last n frequency points of the to-be-predicted frequency hopping sequence according to different step lengths to obtain a plurality of frequency point sequences, and inputting the plurality of frequency point sequences into the adjoint model to obtain a plurality of predicted frequency points output by the adjoint model and probability values of the plurality of predicted frequency points; and determining a next frequency hopping frequency point of the to-be-predicted frequency hopping sequence from the plurality of predicted frequency points based on the probability value. By adopting the technical scheme, the problems of poor real-time performance and low accuracy of the frequency point prediction method in the prior art are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of frequency hopping communication, and in particular to a method, device and electronic equipment for predicting frequency hopping points based on an adjoint model. Background Art

[0002] Frequency hopping technology is a technology widely used in the field of wireless communications to counter interference and prevent monitoring. Global System for Mobile Communications (GSM), Bluetooth, and some special equipment all use wireless frequency hopping technology.

[0003] In the application of frequency hopping technology, in order to improve communication efficiency and ensure the communication quality of one's own party, it is usually necessary to avoid the frequency hopping sequences of other parties. However, since these frequency hopping sequences may not be directly obtained, it is necessary to predict the frequency hopping sequences of other parties to avoid interfering with one's own communication.

[0004] The current frequency hopping sequence prediction mainly relies on the actual experience of operators to predict the frequency hopping frequencies. However, this method of frequency hopping frequency prediction has poor real-time performance and low accuracy. Summary of the invention

[0005] The present invention provides a frequency hopping frequency prediction method, device and electronic equipment based on an adjoint model, which are used to solve the problems of poor real-time performance and low accuracy of the frequency prediction method in the prior art.

[0006] In the first aspect, the present invention provides a method for predicting frequency hopping points based on an accompanying model, comprising: obtaining a frequency hopping sequence to be predicted. Cutting the last n frequency points of the frequency hopping sequence to be predicted according to different step sizes to obtain multiple frequency point sequences, where the value of n is equal to the step size. Inputting the multiple frequency point sequences into the accompanying model to obtain multiple predicted frequency points and probability values ​​of the multiple predicted frequency points output by the accompanying model. Determining the next frequency hopping point of the frequency hopping sequence to be predicted from the multiple predicted frequency points based on the probability value. The accompanying model includes multiple accompanying sub-models, and the multiple accompanying sub-models correspond to different step sizes respectively. The input of the accompanying sub-model corresponding to the step size of X is a frequency point sequence composed of the last X frequency points of the frequency hopping sequence to be predicted, and the output is the predicted frequency point and its probability value corresponding to the frequency point sequence composed of the last X frequency points, where the value of X is an integer greater than or equal to 1.

[0007] Optionally, the companion sub-model is obtained in the following manner: obtain a sample frequency hopping sequence. The sample frequency hopping sequence is intercepted according to multiple step sizes, and a group of sample frequency point sequences are obtained for each step size. The companion sub-model corresponding to the target step size is determined based on a group of sample frequency point sequences corresponding to the target step size. Optionally, the predicted frequency point is determined based on the following manner: for a frequency point sequence composed of the last X frequency points of the frequency hopping sequence to be predicted, the reference frequency point sequence in the companion sub-model corresponding to the step size X is intercepted according to the step size X to obtain multiple sub-frequency point sequences. The reference frequency point sequence is obtained based on a group of sample frequency point sequences corresponding to the step size X. The similarity of the frequency point sequence and the multiple sub-frequency point sequences is calculated respectively, and a target sub-frequency point sequence whose similarity with the frequency point sequence is greater than the similarity threshold is determined. The next frequency point of the target sub-frequency point sequence in the reference frequency point sequence is determined as the predicted frequency point corresponding to the frequency point sequence.

[0008] Optionally, the probability value of the predicted frequency point is determined based on the following method: obtaining the number of occurrences of each predicted frequency point in the multiple predicted frequency points and the total number of occurrences of the multiple predicted frequency points. Based on the number of occurrences of each predicted frequency point and the total number of occurrences of the multiple predicted frequency points, the probability value of each predicted frequency point is determined.

[0009] Optionally, determining the next frequency hopping frequency of the to-be-predicted frequency hopping sequence from multiple predicted frequency points based on the probability value includes: determining the predicted frequency points whose probability values ​​of the multiple predicted frequency points are sorted in descending order and are in the first Q as the next frequency hopping frequency of the frequency hopping sequence, where Q is an integer greater than or equal to 1.

[0010] Optionally, the number of each frequency point in the reference frequency point sequence of any companion sub-model is greater than or equal to a frequency point number threshold.

[0011] In a second aspect, the present invention provides a frequency hopping point prediction device based on an accompanying model, the device comprising: an acquisition module, used to obtain a frequency hopping sequence to be predicted. An interception module, used to intercept the last n frequency points of the frequency hopping sequence to be predicted according to different step lengths to obtain multiple frequency point sequences, wherein the value of n is equal to the step length. A processing module, used to input multiple frequency point sequences into the accompanying model, and obtain multiple predicted frequency points and probability values ​​of the multiple predicted frequency points output by the accompanying model. A determination module, used to determine the next frequency hopping point of the frequency hopping sequence to be predicted from the multiple predicted frequency points based on the probability value. The accompanying model contains multiple accompanying sub-models, and the multiple accompanying sub-models correspond to different step lengths respectively. The input of the accompanying sub-model corresponding to the step length of X is the frequency point sequence composed of the last X frequency points of the frequency hopping sequence to be predicted, and the output is the predicted frequency point and its probability value corresponding to the frequency point sequence composed of the last X frequency points, wherein the value of X is a positive integer greater than or equal to 1.

[0012] In a third aspect, the present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, any of the frequency hopping frequency prediction methods based on the accompanying model described above is implemented.

[0013] In a fourth aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-mentioned frequency hopping frequency prediction methods based on the companion model.

[0014] In a fifth aspect, the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned frequency hopping frequency prediction methods based on the accompanying model.

[0015] The method, device and electronic device for predicting frequency hopping points based on the companion model provided by the present invention obtain multiple frequency point sequences by intercepting the frequency hopping sequence to be predicted according to different step sizes, and inputting the multiple frequency point sequences into the companion model to obtain multiple predicted frequency points output by the companion model and probability values ​​of the multiple predicted frequency points. The next frequency hopping point of the frequency hopping sequence to be predicted is determined from the multiple predicted frequency points based on the probability value. The prediction timeliness is improved by predicting the frequency point sequences of different step sizes through the companion sub-model in the companion model, and the frequency hopping point is further determined after outputting multiple predicted frequency points, thereby increasing the reliability and accuracy of the predicted frequency points. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0017] Figure 1 It is a flow chart of the frequency hopping frequency prediction method based on the adjoint model provided by the present invention.

[0018] Figure 2 It is a flow chart of the frequency hopping prediction provided by the present invention.

[0019] Figure 3 It is a schematic diagram of the output interface of the adjoint model provided by the present invention.

[0020] Figure 4 It is a structural schematic diagram of a frequency hopping frequency prediction device based on an adjoint model provided by the present invention.

[0021] Figure 5 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] Figure 1 FIG. 1 is a flow chart of a frequency hopping frequency prediction method based on an adjoint model provided by the present invention. Figure 1 As shown, the method includes the following steps S101-S104.

[0024] Step S101: obtaining a frequency hopping sequence to be predicted.

[0025] It should be noted that the frequency hopping sequence to be predicted refers to a known frequency hopping sequence of a non-self-communication device. The present invention does not further limit the method for obtaining the frequency hopping sequence to be predicted, nor does it further limit the length of the frequency hopping sequence to be predicted.

[0026] As an optional embodiment, a frequency hopping sequence to be predicted can be illustrated as {A, B, C, D, E, F, G, H, I}, where each letter represents a frequency point.

[0027] Step S102: the last n frequency points of the frequency hopping sequence to be predicted are intercepted according to different step lengths to obtain multiple frequency point sequences, where the value of n is equal to the step length.

[0028] As an optional embodiment, after the last n frequency points of the above-mentioned frequency hopping sequence to be predicted are intercepted according to a step size of 1-5, the obtained frequency point sequence is: The frequency point sequence obtained when the step length is 1 is {I}.

[0029] The frequency point sequence obtained when the step size is 2 is {H, I}.

[0030] The frequency point sequence obtained when the step length is 3 is {G, H, I}.

[0031] The frequency point sequence obtained when the step length is 4 is {F, G, H, I}.

[0032] The frequency point sequence obtained when the step size is 5 is {D, E, F, G, H, I}.

[0033] It should be noted that the step size setting is based on the step size setting of the companion sub-model in the companion model. That is to say, if there are 6 companion sub-models with 6 step sizes in the companion model, there are also 6 corresponding frequency point sequences, and the number of frequency points in each frequency point sequence is consistent with the step size of the companion sub-model.

[0034] Step S103: inputting a plurality of frequency point sequences into the adjoint model, and obtaining a plurality of predicted frequency points and probability values ​​of the plurality of predicted frequency points output by the adjoint model.

[0035] It should be noted that the companion model includes multiple companion sub-models, and the multiple companion sub-models correspond to different step sizes. The input of the companion sub-model corresponding to the step size of X is the frequency point sequence composed of the last X frequency points of the frequency hopping sequence to be predicted, and the output is the predicted frequency point and its probability value corresponding to the frequency point sequence composed of the last X frequency points. The value of X is an integer greater than or equal to 1.

[0036] As an optional embodiment, Figure 2 It is a flow chart of the frequency hopping prediction provided by the present invention. The measured frequency hopping sequence in the figure is the frequency hopping sequence to be predicted in the above embodiment. The measured frequency hopping sequence is truncated according to the step length to obtain multiple frequency point sequences which are input into the companion model. The companion model will input the frequency point sequence of the corresponding step length into the corresponding companion sub-model for frequency point prediction.

[0037] It should be noted that Figure 2 The maximum step length in can be flexibly set according to the length of the frequency hopping sequence to be predicted, and the maximum step length is generally set to 10.

[0038] It should be noted that, since the companion model includes multiple companion sub-models, frequency point sequences with different step lengths are input into the companion sub-models with corresponding step lengths. For example, the input of the companion sub-model with a step length of 1 is the last frequency point {I} of the frequency hopping sequence to be predicted, and the input of the companion sub-model with a step length of 2 is the last two frequency points {H, I} of the frequency hopping sequence to be predicted, and so on.

[0039] It should be noted that the present invention only defines the companion model as including multiple companion sub-models with different step sizes, and does not further define the structure of the companion model, and other functions can be integrated into the model. In addition, after multiple frequency point sequences are input into the companion model, the companion model will send the frequency point sequences with different step sizes to the corresponding companion sub-models for processing according to the step size.

[0040] Frequency prediction is performed through the adjoint model without manual operation. The received frequency sequence can be quickly predicted, which improves the timeliness of frequency prediction.

[0041] In an exemplary embodiment, the adjoint sub-model is obtained by following the steps S11-S13: Step S11: Acquire a sample frequency hopping sequence.

[0042] Step S12: intercepting the sample frequency hopping sequence according to multiple step lengths respectively, and obtaining a set of sample frequency point sequences corresponding to each step length.

[0043] As an optional embodiment, it is assumed that the sample frequency hopping sequence is {1, 2, 3, 4, 5, 6}. It should be noted that each number represents a frequency point.

[0044] After the above sample frequency hopping sequence is intercepted according to the step length, a set of sample frequency point sequences can be expressed as: A step size of 1 corresponds to a set of frequency point sequences {1}, {2}, {3}, {4}, {5}, {6} obtained by interception.

[0045] The frequency point sequence obtained when the step size is 2 is {1, 2}, {2, 3}, {3, 4}, {4, 5}, and {5, 6}.

[0046] The frequency point sequences obtained when the step length is 3 are {1, 2, 3}, {2, 3, 4}, and {4, 5, 6}.

[0047] Step S13: Determine the companion sub-model corresponding to the target step length based on a set of sample frequency point sequences corresponding to the target step length.

[0048] In an exemplary embodiment, the number of each frequency point in the reference frequency point sequence of any companion sub-model is greater than or equal to a frequency point number threshold.

[0049] It should be noted that in order to make the reference frequency points in the companion sub-model more referenceable and cover more sequence combinations of frequency hopping frequencies, the number of each frequency point in the reference frequency point sequence needs to be greater than or equal to the frequency point number threshold. The frequency point number threshold can be flexibly set. If the frequency point number threshold is 3, the reference frequency point sequence of a companion sub-model can be expressed as {2, 3, 1, 1, 2, 3, 3, 2, 1}.

[0050] It should be noted that a group of frequency point sequences corresponding to the step length constitutes a reference frequency point sequence in the companion sub-model. In the embodiment of the present invention, each companion sub-model must include a reference frequency point sequence corresponding to the step length. The present invention does not further limit the structure of the companion sub-model, and other functions can also be integrated in the companion sub-model.

[0051] It should be noted that in actual application, the number of each frequency point in the reference frequency point sequence of any accompanying sub-model needs to be greater than or equal to the frequency point number threshold. In the following embodiments, numbers are used to represent frequency points. For the convenience of writing, the number of each frequency point in the embodiments is not strictly set to be greater than the frequency point number threshold, and is only used to illustrate the prediction method of frequency hopping frequencies.

[0052] As an optional embodiment, a reference frequency point sequence in a companion sub-model determined by a set of sample frequency point sequences with a target step size of 1 can be expressed as: {1, 2, 3, 4, 5, 6}.

[0053] The reference frequency point sequence in the companion sub-model determined by a set of sample frequency point sequences with a target step size of 2 can be expressed as: {1, 2, 2, 3, 3, 4, 4, 5, 5, 6}.

[0054] In an exemplary embodiment, the predicted frequency point is determined by the following steps S21-S23.

[0055] Step S21: for the frequency point sequence composed of the last X frequency points of the frequency hopping sequence to be predicted, the reference frequency point sequence in the companion sub-model corresponding to the step length X is intercepted according to the step length X to obtain multiple sub-frequency point sequences. The reference frequency point sequence is obtained based on a group of sample frequency point sequences corresponding to the step length X.

[0056] As an optional embodiment, the frequency point sequence to be predicted is {1, 2, 3, 4, 5, 6}. Then the frequency point sequence composed of the last X frequency points of the frequency point sequence to be predicted can be expressed as: The frequency point sequence composed of the last frequency point of the frequency point sequence to be predicted is {6}.

[0057] The frequency point sequence composed of the last two frequency points of the frequency point sequence to be predicted is {5, 6}.

[0058] The frequency point sequence composed of the last three frequency points of the frequency point sequence to be predicted is {4, 5, 6}.

[0059] The frequency point sequence composed of the last 4 frequency points of the frequency point sequence to be predicted is {3, 4, 5, 6}.

[0060] The frequency point sequence composed of the last 5 frequency points of the frequency point sequence to be predicted is {2, 3, 4, 5, 6}.

[0061] It should be noted that, since there are only 5 companion sub-models in the companion model at this time, the frequency point sequences of 5 step lengths can be cut and input into the corresponding companion sub-models. If the number of sub-models is greater than the number of frequency point sequences cut, the frequency point sequences of the corresponding step lengths can be input into the corresponding companion sub-models according to the corresponding relationship of the step lengths. The companion sub-models without corresponding frequency point sequences have no input and are not used for the prediction of frequency hopping frequencies.

[0062] The reference frequency point sequence in the companion sub-model with a step size of 1 is {1, 2, 3, 4, 5, 6, 7, 8, 9, 10}.

[0063] The reference frequency point sequence in the companion sub-model with a step size of 2 is {1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8, 8, 9, 9, 10}.

[0064] The reference frequency point sequence in the companion submodel with a step size of 3 is {1, 2, 3, 2, 3, 4, 3, 4, 5, 4, 5, 6, 5, 6, 7, 6, 7, 8, 7, 8, 9, 8, 9, 10}.

[0065] The reference frequency point sequence in the companion submodel with a step size of 4 is {1, 2, 3, 4, 2, 3, 4, 5, 3, 4, 5, 6, 4, 5, 6, 7, 5, 6, 7, 8, 6, 7, 8, 9, 7, 8, 9, 10}.

[0066] The reference frequency sequence in the companion submodel with a step size of 5 is {1, 2, 3, 4, 5, 2, 3, 4, 5, 6, 3, 4, 5, 6, 7, 4, 5, 6, 7, 8, 5, 6, 7, 8, 9, 6, 7, 8, 9, 10}.

[0067] Since the input of the companion sub-model with a step size of 1 is the frequency point sequence {6} composed of the last frequency point of the frequency point sequence to be predicted, the sub-frequency point sequence obtained by truncating the reference frequency point sequence in the companion sub-model with a step size of 1 according to the step size can be illustrated as {1}, {2}, {3}, {4}, {5}, {6}, {7}, {8}, {9}, {10}.

[0068] The input of the companion sub-model with a step size of 2 is the frequency point sequence {5, 6} composed of the last two frequency points of the frequency point sequence to be predicted. Then, for the reference frequency point sequence in the companion sub-model with a step size of 2, the sub-frequency point sequence obtained after truncating according to the step size can be illustrated as {1, 2}, {2, 2}, {2, 3}, {3, 3}, {3, 4}, {4, 4}, {4, 5}, {5, 5}, {5, 6}, {6, 6}, {6, 7}, {7, 7}, {7, 8}, {8, 8}, {8, 9}, {9, 9}, {9, 10}.

[0069] The manner in which the companion sub-models of other step sizes are cut to obtain the sub-frequency point sequence is consistent with the above embodiment, and will not be described in detail here.

[0070] Step S22: calculating similarities between the frequency point sequence and a plurality of sub-frequency point sequences respectively, and determining a target sub-frequency point sequence whose similarity with the frequency point sequence is greater than a similarity threshold.

[0071] It should be noted that the similarity between frequency point sequences can be calculated according to the frequency difference method, signal correlation method or spectrum feature method. A neural network can also be integrated in the adjoint model to learn the similarity between frequency points, and the similarity between frequency points is determined by inputting the characteristic value model of the frequency point signal according to the similarity score output. The specific method for calculating the similarity is not further limited in the embodiment of the present application.

[0072] Taking the companion sub-model with a step size of 2 as an example, the input of the companion sub-model with a step size of 2 is the frequency point sequence {5, 6} composed of the last two frequency points of the frequency point sequence to be predicted. The similarity of {5, 6} is calculated with the obtained sub-frequency point sequences {1, 2}, {2, 2}, {2, 3}, {3, 3}, {3, 4}, {4, 4}, {4, 5}, {5, 5}, {5, 6}, {6, 6}, {6, 7}, {7, 7}, {7, 8}, {8, 8}, {8, 9}, {9, 9}, {9, 10} respectively.

[0073] It should be noted that, on the basis of matching, a similarity threshold is added, and it is allowed that there are less than 10% erroneous frequencies between the frequency point sequence and the sub-frequency point sequence, and it is considered that the similarity between the frequency point sequence and the sub-frequency point sequence is greater than 90% (i.e., the similarity threshold in the above embodiment) to be determined as the target sub-frequency point sequence. The specific similarity threshold setting can be flexibly set according to demand, and the higher the similarity threshold, the fewer sub-frequency point sequences are determined to be the target sub-frequency point sequence, and the higher the accuracy of the predicted frequency.

[0074] It should be further explained that, since the companion model includes multiple companion sub-models, the step sizes of the companion sub-models are not equal, and the similarity threshold of each companion sub-model can be set differently. For the companion sub-models with a smaller step size, the similarity threshold can be exactly the same, that is, the similarity threshold is 1. The similarity threshold of the companion sub-model with a longer step size can be set relatively low, so as to determine more target sub-frequency point sequences and obtain as many predicted frequency points as possible.

[0075] For the companion sub-model with a step size of 2, the target sub-frequency point sequence can be determined in a completely similar manner, that is, {5, 6} in the sub-frequency point sequence are all target sub-frequency point sequences.

[0076] Step S23: determining the next frequency point of the target sub-frequency point sequence in the reference frequency point sequence as the predicted frequency point corresponding to the frequency point sequence.

[0077] Through the above steps S21 to S23, the frequency prediction is performed in combination with the similarity, thereby further improving the prediction accuracy.

[0078] In combination with the above embodiment, the sub-frequency point sequences {5, 6} in the companion sub-model with a step length of 2 are all target sub-frequency point sequences, and the next frequency point of {5, 6} in the reference frequency point sequence is determined as the predicted frequency point corresponding to the frequency point sequence, the predicted frequency point obtained by the companion sub-model with a step length of 1 is 7, the predicted frequency point obtained by the companion sub-model with a step length of 2 is 6, the predicted frequency point obtained by the companion sub-model with a step length of 3 is 5, the predicted frequency point obtained by the companion sub-model with a step length of 4 is 4, and the predicted frequency point obtained by the companion sub-model with a step length of 5 is 3. The multiple predicted frequency points finally determined are 7, 6, 5, 4, and 3.

[0079] Step S104: determining the next frequency hopping frequency of the to-be-predicted frequency hopping sequence from the multiple predicted frequency frequencies based on the probability value.

[0080] In an exemplary embodiment, the probability value of the predicted frequency point is determined by the following steps S31-S32.

[0081] Step S31: Obtain the number of occurrences of each predicted frequency point among multiple predicted frequency points and the total number of occurrences of multiple predicted frequency points.

[0082] It should be noted that the number of occurrences of the predicted frequency point refers to the number of times the input frequency point sequence is used as the next frequency point of the same or similar sub-frequency point sequence in the accompanying sub-model. The total number of occurrences of multiple predicted frequency points refers to the sum of the number of occurrences of each predicted frequency point.

[0083] For example, the reference frequency sequence of the companion submodel is {1, 2, 6, 3, 7, 5, 3, 4, 5, 6, 3, 9, 3, 7, 6, 3, 7}, and the input frequency sequence is {3}, then the predicted frequency points are 7, 4, and 9. Frequency point 7 appears 3 times, frequency point 4 appears 1 time, frequency point 9 appears 1 time, and the total number of occurrences of multiple predicted frequency points is 5 times.

[0084] Step S32: Determine the probability value of each predicted frequency point based on the number of occurrences of each predicted frequency point and the total number of occurrences of multiple predicted frequency points.

[0085] As an optional embodiment, assume that the multiple predicted frequency points are: 2, 5, and 10, and the number of occurrences of these three frequency points is 20 times, of which frequency point 2 appears 11 times, frequency point 5 appears 6 times, and frequency point 10 appears 3 times. Then the probability value of predicting frequency point 2 is 11 / 20, the probability value of predicting frequency point 5 is 6 / 20, and the probability value of predicting frequency point 10 is 3 / 20.

[0086] In an exemplary embodiment, determining the next frequency hopping point of the to-be-predicted frequency hopping sequence from a plurality of predicted frequency points based on the probability value includes: Determine that the first Q predicted frequency points in descending order of probability values ​​of the plurality of predicted frequency points are the next frequency hopping frequency point of the frequency hopping sequence, where Q is an integer greater than or equal to 1.

[0087] It should be noted that the specific value of Q can be set flexibly. Generally speaking, the three frequencies with the highest probability values ​​are taken as the next frequency hopping frequencies of the predicted frequency hopping sequence. The communication equipment of the own party only needs to bypass these three predicted frequency hopping frequencies for communication.

[0088] The probability value corresponding to frequency 1 is 90%, the probability value corresponding to frequency 2 is 85%, the probability value corresponding to frequency 3 is 12%, and the probability value corresponding to frequency 4 is 77%. Assuming Q is 3, frequency 1, frequency 2, and frequency 4 are used as the next frequency hopping frequency of the predicted frequency hopping sequence.

[0089] As an optional embodiment, the probability value threshold may also be directly set, and the frequency point with a probability value greater than the probability value threshold is used as the next frequency hopping frequency point of the frequency hopping sequence.

[0090] As an optional embodiment, a graphical interface can be designed in the adjoint model to intuitively display the prediction results, such as Figure 3 The figure shows a schematic diagram of the output interface of the adjoint model provided by the present invention. Figure 3 In the formula, frequency 1-frequency X are the output predicted frequencies, the total number is the total number of occurrences of multiple predicted frequencies, the number of occurrences is the number of occurrences of each predicted frequency, the probability is the ratio of the number of occurrences of each predicted frequency to the total number, and the predicted frequency is the predicted frequency selected in combination with the probability value as the next frequency hopping frequency of the frequency hopping sequence (i.e., frequency 1, frequency 2, and frequency 4 in the above embodiment).

[0091] In the above steps S101-S104, multiple frequency point sequences are obtained by intercepting the frequency hopping sequence to be predicted according to different step sizes, and the multiple frequency point sequences are input into the companion model to obtain multiple predicted frequency points output by the companion model and the probability values ​​of the multiple predicted frequency points. The next frequency hopping frequency point of the frequency hopping sequence to be predicted is determined from the multiple predicted frequency points based on the probability value. The timeliness of the prediction is improved by predicting the frequency point sequences of different step sizes through the companion sub-model in the companion model, and the frequency hopping frequency point is further determined after outputting multiple predicted frequency points, which increases the reliability and accuracy of the predicted frequency points.

[0092] The execution subject of each step in the above method can be a frequency hopping frequency prediction device based on the accompanying model, which can be implemented by software and / or hardware. The device can be integrated in an electronic device. The electronic device can be a terminal device (such as a smart phone, a personal computer, etc.), or a server (such as a local server or a cloud server, or a server cluster, etc.), or a processor, or a chip, etc.

[0093] The following is a description of the frequency hopping frequency prediction device based on the companion model provided by the present invention. The frequency hopping frequency prediction device based on the companion model described below and the frequency hopping frequency prediction method based on the companion model described above can be referred to each other.

[0094] Figure 4 FIG. 1 is a schematic diagram of the structure of the frequency hopping frequency prediction device based on the adjoint model provided by the present invention. Figure 4 As shown, the device comprises: The acquisition module 410 is used to acquire the frequency hopping sequence to be predicted.

[0095] The interception module 420 is used to intercept the last n frequency points of the frequency hopping sequence to be predicted according to different step lengths to obtain multiple frequency point sequences, and the value of n is equal to the step length.

[0096] The processing module 430 is used to input the multiple frequency point sequences into the adjoint model to obtain the multiple predicted frequency points and the probability values ​​of the multiple predicted frequency points output by the adjoint model.

[0097] The determination module 440 is configured to determine the next frequency hopping frequency of the to-be-predicted frequency hopping sequence from the multiple predicted frequency points based on the probability value.

[0098] The companion model includes multiple companion sub-models, each of which corresponds to a different step size. The input of the companion sub-model corresponding to the step size X is a frequency point sequence consisting of the last X frequency points of the frequency hopping sequence to be predicted, and the output is a predicted frequency point and its probability value corresponding to the frequency point sequence consisting of the last X frequency points. The value of X is a positive integer greater than or equal to 1.

[0099] The above device obtains multiple frequency point sequences by intercepting the frequency hopping sequence to be predicted according to different step sizes, and inputs the multiple frequency point sequences into the companion model to obtain multiple predicted frequency points output by the companion model and the probability values ​​of the multiple predicted frequency points. The next frequency hopping frequency point of the frequency hopping sequence to be predicted is determined from the multiple predicted frequency points based on the probability value. The timeliness of the prediction is improved by predicting the frequency point sequences of different step sizes through the companion sub-model in the companion model, and the frequency hopping frequency point is further determined after outputting multiple predicted frequency points, thereby increasing the reliability and accuracy of the predicted frequency points.

[0100] In some embodiments, the companion sub-model is obtained by: obtaining a sample frequency hopping sequence, intercepting the sample frequency hopping sequence according to multiple step lengths, and obtaining a set of sample frequency point sequences corresponding to each step length, and determining the companion sub-model corresponding to the target step length based on the set of sample frequency point sequences corresponding to the target step length.

[0101] In some embodiments, the predicted frequency point is determined based on the following method: for a frequency point sequence composed of the last X frequency points of the frequency hopping sequence to be predicted, the reference frequency point sequence in the companion sub-model corresponding to the step length X is intercepted according to the step length X to obtain multiple sub-frequency point sequences. The reference frequency point sequence is obtained based on a group of sample frequency point sequences corresponding to the step length X. The similarity between the frequency point sequence and the multiple sub-frequency point sequences is calculated respectively, and a target sub-frequency point sequence whose similarity with the frequency point sequence is greater than the similarity threshold is determined. The next frequency point of the target sub-frequency point sequence in the reference frequency point sequence is determined as the predicted frequency point corresponding to the frequency point sequence.

[0102] In some embodiments, the probability value of the predicted frequency point is determined based on the following method: obtaining the number of occurrences of each predicted frequency point in the multiple predicted frequency points and the total number of occurrences of the multiple predicted frequency points. Based on the number of occurrences of each predicted frequency point and the total number of occurrences of the multiple predicted frequency points, the probability value of each predicted frequency point is determined.

[0103] In some embodiments, determining the next frequency hopping frequency of a frequency hopping sequence to be predicted from a plurality of predicted frequency points based on probability values ​​includes: determining that the predicted frequency points of the first Q predicted frequency points in descending order of probability values ​​of the plurality of predicted frequency points are the next frequency hopping frequency points of the frequency hopping sequence, where Q is an integer greater than or equal to 1.

[0104] In some embodiments, the number of each frequency point in the reference frequency point sequence of any companion sub-model is greater than or equal to a frequency point number threshold.

[0105] Figure 5 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 5 As shown, the electronic device may include: a processor 510, a communication interface 520, a memory 530 and a communication bus 540, wherein the processor 510, the communication interface 520 and the memory 530 communicate with each other through the communication bus 540. The processor 510 may call the logic instructions in the memory 530 to execute the frequency hopping point prediction method based on the companion model, and the method includes: Get the frequency hopping sequence to be predicted.

[0106] The last n frequency points of the frequency hopping sequence to be predicted are intercepted according to different step lengths to obtain multiple frequency point sequences, and the value of n is equal to the step length.

[0107] A plurality of frequency point sequences are input into the adjoint model to obtain a plurality of predicted frequency points and probability values ​​of the plurality of predicted frequency points output by the adjoint model.

[0108] The next frequency hopping frequency of the frequency hopping sequence to be predicted is determined from a plurality of predicted frequency points based on the probability value.

[0109] The companion model includes multiple companion sub-models, each of which corresponds to a different step size. The input of the companion sub-model corresponding to the step size X is a frequency point sequence consisting of the last X frequency points of the frequency hopping sequence to be predicted, and the output is a predicted frequency point and its probability value corresponding to the frequency point sequence consisting of the last X frequency points. The value of X is an integer greater than or equal to 1.

[0110] In addition, the logic instructions in the above-mentioned memory 530 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0111] On the other hand, the present invention further provides a computer program product, the computer program product includes a computer program, the computer program can be stored in a non-transitory computer-readable storage medium, when the computer program is executed by a processor, the computer can execute the frequency hopping point prediction method based on the adjoint model provided by the above methods, the method includes: Get the frequency hopping sequence to be predicted.

[0112] The last n frequency points of the frequency hopping sequence to be predicted are intercepted according to different step lengths to obtain multiple frequency point sequences, and the value of n is equal to the step length.

[0113] A plurality of frequency point sequences are input into the adjoint model to obtain a plurality of predicted frequency points and probability values ​​of the plurality of predicted frequency points output by the adjoint model.

[0114] The next frequency hopping frequency of the frequency hopping sequence to be predicted is determined from a plurality of predicted frequency points based on the probability value.

[0115] The companion model includes multiple companion sub-models, each of which corresponds to a different step size. The input of the companion sub-model corresponding to the step size X is a frequency point sequence consisting of the last X frequency points of the frequency hopping sequence to be predicted, and the output is a predicted frequency point and its probability value corresponding to the frequency point sequence consisting of the last X frequency points. The value of X is an integer greater than or equal to 1.

[0116] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented when the computer program is executed by a processor to execute the frequency hopping frequency prediction method based on the companion model provided by the above methods, the method comprising: Get the frequency hopping sequence to be predicted.

[0117] The last n frequency points of the frequency hopping sequence to be predicted are intercepted according to different step lengths to obtain multiple frequency point sequences, and the value of n is equal to the step length.

[0118] A plurality of frequency point sequences are input into the adjoint model to obtain a plurality of predicted frequency points and probability values ​​of the plurality of predicted frequency points output by the adjoint model.

[0119] The next frequency hopping frequency of the frequency hopping sequence to be predicted is determined from a plurality of predicted frequency points based on the probability value.

[0120] The companion model includes multiple companion sub-models, each of which corresponds to a different step size. The input of the companion sub-model corresponding to the step size X is a frequency point sequence consisting of the last X frequency points of the frequency hopping sequence to be predicted, and the output is a predicted frequency point and its probability value corresponding to the frequency point sequence consisting of the last X frequency points. The value of X is an integer greater than or equal to 1.

[0121] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, i.e., they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative effort.

[0122] Through the description of the above implementation modes, those skilled in the art can clearly understand that each implementation mode can be implemented by means of software plus a necessary general hardware platform, or of course by hardware. Based on such an understanding, the above technical solution can essentially or in other words be embodied in the form of a software product that contributes to the prior art. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or some parts of the embodiment.

[0123] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the above embodiments, a person skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A frequency hopping frequency prediction method based on an adjoint model, characterized in that: include: Obtain a frequency hopping sequence to be predicted; Cutting the last n frequency points of the frequency hopping sequence to be predicted according to different step lengths to obtain multiple frequency point sequences, where the value of n is equal to the step length; Inputting the plurality of frequency point sequences into an adjoint model to obtain a plurality of predicted frequency points output by the adjoint model and probability values ​​of the plurality of predicted frequency points; Determine the next frequency hopping point of the to-be-predicted frequency hopping sequence from the multiple predicted frequency points based on the probability value; The companion model includes multiple companion sub-models, and the multiple companion sub-models correspond to different step sizes respectively. The input of the companion sub-model corresponding to the step size of X is a frequency point sequence composed of the last X frequency points of the to-be-predicted frequency hopping sequence, and the output is a predicted frequency point and its probability value corresponding to the frequency point sequence composed of the last X frequency points, and the value of X is an integer greater than or equal to 1.

2. The method for predicting frequency hopping frequency based on adjoint model according to claim 1, characterized in that: The adjoint submodel is obtained in the following way: Obtain a sample frequency hopping sequence; The sample frequency hopping sequence is intercepted according to a plurality of step lengths, and each step length corresponds to a set of sample frequency point sequences; Based on a set of sample frequency point sequences corresponding to the target step length, an accompanying sub-model corresponding to the target step length is determined.

3. The method for predicting frequency hopping frequency based on adjoint model according to claim 2, characterized in that: The predicted frequency is determined based on the following method: For a frequency point sequence composed of the last X frequency points of the frequency hopping sequence to be predicted, a reference frequency point sequence in the companion sub-model corresponding to the step length X is intercepted according to the step length X to obtain a plurality of sub-frequency point sequences; the reference frequency point sequence is obtained based on a group of sample frequency point sequences corresponding to the step length X; Calculating similarities between the frequency point sequence and the multiple sub-frequency point sequences respectively, and determining a target sub-frequency point sequence whose similarity with the frequency point sequence is greater than a similarity threshold; The next frequency point of the target sub-frequency point sequence in the reference frequency point sequence is determined as the predicted frequency point corresponding to the frequency point sequence.

4. The method for predicting frequency hopping frequency based on adjoint model according to claim 1 or 3, characterized in that: The probability value of the predicted frequency point is determined based on the following method: Obtaining the number of occurrences of each predicted frequency point in the multiple predicted frequency points and the total number of occurrences of the multiple predicted frequency points; Based on the number of occurrences of each predicted frequency point and the total number of occurrences of the multiple predicted frequency points, a probability value of each predicted frequency point is determined.

5. The method for predicting frequency hopping frequency based on adjoint model according to claim 1, characterized in that: The determining, based on the probability value, the next frequency hopping point of the to-be-predicted frequency hopping sequence from the multiple predicted frequency points comprises: Determine that the first Q predicted frequency points among the probability values ​​of the multiple predicted frequency points are sorted in descending order as the next frequency hopping frequency point of the frequency hopping sequence, where Q is an integer greater than or equal to 1.

6. The method for predicting frequency hopping frequency based on adjoint model according to claim 3, characterized in that: The number of each frequency point in the reference frequency point sequence of any of the companion sub-models is greater than or equal to a frequency point number threshold.

7. A frequency hopping frequency prediction device based on an adjoint model, characterized in that: include: An acquisition module, used for acquiring a frequency hopping sequence to be predicted; A truncation module, used for truncation of the last n frequency points of the frequency hopping sequence to be predicted according to different step lengths to obtain multiple frequency point sequences, where the value of n is equal to the step length; A processing module, used for inputting the plurality of frequency point sequences into an adjoint model to obtain a plurality of predicted frequency points output by the adjoint model and probability values ​​of the plurality of predicted frequency points; A determination module, configured to determine a next frequency hopping frequency of the to-be-predicted frequency hopping sequence from the multiple predicted frequency points based on the probability value; The companion model includes multiple companion sub-models, and the multiple companion sub-models correspond to different step sizes respectively. The input of the companion sub-model corresponding to the step size of X is a frequency point sequence composed of the last X frequency points of the to-be-predicted frequency hopping sequence, and the output is a predicted frequency point and its probability value corresponding to the frequency point sequence composed of the last X frequency points. The value of X is a positive integer greater than or equal to 1.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the frequency hopping frequency prediction method based on the adjoint model as claimed in any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for predicting frequency hopping frequency based on the adjoint model as claimed in any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for predicting frequency hopping frequency based on the adjoint model as claimed in any one of claims 1 to 6 is implemented.

Images