OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals

Abstract

The present invention discloses an OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals. The method divides subcarriers with larger CSI amplitudes into two sets based on the sensitivity of subcarriers to gesture changes and the role of data transmission. Then, the correlation between adjacent subcarriers is considered, and the subcarrier interval is set and grouped to maintain the common change trend embodied by the original CSI and the correlation between adjacent CSIs. Finally, three selection strategies for processing each group of CSI are provided, which select the average CSI value in each subcarrier group to form a set; select the maximum CSI value in each subcarrier group to form a set; and select the CSI value corresponding to the first subcarrier in each group to form a set.

Description

Technical Field

[0001] The present invention relates to the field of gesture recognition, and in particular to an OFDM subcarrier grouping and selection method for intelligent recognition of WIFI signals. Background Art

[0002] With the rapid development of wireless communication technology, OFDM (Orthogonal Frequency Division Multiplexing) has been widely used in various wireless communication systems, such as WiFi, LTE, 5G, etc. OFDM has become a key technology in modern wireless communication systems with its high spectrum efficiency and superior performance in resisting multipath interference. Especially in WiFi technology, OFDM is used to improve data transmission rate and communication quality. In recent years, WiFi signals have not only been used for data transmission and communication, but have also been widely studied for environmental perception. It has the advantages of no need for additional hardware, easy deployment, and a wide range of applications, such as using WiFi signals for gesture recognition.

[0003] In the gesture recognition system based on WiFi signals, gestures are detected and recognized based on the analysis of CSI (channel state information). CSI contains rich information on the signal propagation path, such as reflection, refraction, etc. By analyzing this information, the perception and recognition of moving targets in the environment can be achieved. However, traditional gesture recognition methods usually require a lot of computing resources and storage space, which is a big challenge for devices with limited computing resources such as mobile terminals. When processing CSI data, traditional methods usually need to fully analyze the data of all subcarriers, which not only increases the computational complexity and storage requirements, but also may affect the real-time and accuracy of gesture recognition. Therefore, how to perform effective CSI data processing and achieve efficient gesture recognition under the condition of limited computing resources has become an important topic of current research. Summary of the invention

[0004] The purpose of the present invention is to overcome the shortcomings and deficiencies of the prior art, and to provide an OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals, which can compress the channel state information (CSI) dimension, reduce file storage space, and effectively complete gesture recognition tasks, thereby solving the problem that traditional methods do not take into account the limited computing resources of mobile terminals.

[0005] The purpose of the present invention is achieved through the following technical solutions:

[0006] The OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals includes the following steps:

[0007] S1. According to the corresponding CSI amplitudes on different subcarriers, subcarriers with CSI amplitudes greater than a first preset value are selected from all subcarriers, and their corresponding CSI amplitudes constitute a set , the number of subcarriers included is , the subcarrier feature dimension of CSI data is reduced to In this embodiment, subcarriers with indexes No.126 to No.187 are selected from 256 subcarriers. ; The subcarrier feature dimension of CSI data is reduced from 256 to .

[0008] S2. Collection CSI amplitude in time frame The subcarrier set that changes within constitutes a set , the remaining subcarrier sets constitute the set ;gather ;in, ,and , , For collection The number of subcarriers included, For collection The number of subcarriers included;

[0009] Based on the different transmission functions of subcarriers in OFDM systems and the characteristics of CSI, the collection Divide into sets , .gather The subcarrier group is used for actual data transmission and is very sensitive to environmental changes. The subcarrier group is used as a protection subcarrier near the center frequency to reduce adjacent channel interference. The CSI amplitude of the subcarrier in the time frame changes within, and the collection The corresponding CSI amplitude remains constant.

[0010] S3, according to the subcarrier grouping interval , Separately for the set ,gather Further grouping, corresponding to the formation of sets ,gather ,gather ;

[0011] S4. For the collection The CSI amplitude corresponding to the subcarrier is used to operate the CSI in each group using a data selection strategy and generate the final CSI data input ; The data selection strategy is one of the selection strategies of taking the average value, taking the maximum value, and discrete selection.

[0012] According to the correlation between subcarriers and the system's resource and performance requirements, the present invention proposes three strategies for trade-off selection: taking the average value (A-CSIS) can better retain the overall change trend of the channel state and reflect the global characteristics of the channel information based on the strong correlation between subcarriers; taking the maximum value (M-CSIS) is suitable for situations that are sensitive to the peak value of the signal, and the system needs to respond quickly to drastic changes in the channel state; discrete selection (D-CSIS) has a lower data processing volume, which can better ensure real-time performance and high efficiency, and meet the system requirements under limited computing resources. For the specific application of gesture recognition, A-CSIS has the highest accuracy among the three strategies when the CSI dimension is the same.

[0013] In step S4, the selection strategy for taking the average value is specifically:

[0014] definition , ,in Indicates in the collection The group index in ;

[0015] definition For the The average value of CSI in the subcarrier groups, The calculation formula is:

[0016]

[0017] The obtained average CSI set is: ,So Will be used as the final CSI data input.

[0018] In step S4, the selection strategy for taking the maximum value is specifically:

[0019] definition , ,in Indicates in the collection The group index in ;

[0020] definition For the time frame In terms of dimension, The maximum CSI value in the subcarrier group, or The CSI value corresponding to the subcarrier with the highest fluctuation rate in the subcarrier group;

[0021] The specific calculation formula is:

[0022]

[0023] The maximum CSI set obtained is: ,So Will be used as the final CSI data input.

[0024] The subcarrier with the highest fluctuation rate is the subcarrier with the largest standard deviation or variance of the CSI data.

[0025] In step S4, the discrete selection strategy is specifically:

[0026] definition , ,in Indicates in the collection The group index in ;

[0027] definition For the The CSI value corresponding to a subcarrier in the group is calculated as follows:

[0028]

[0029] The obtained discrete CSI set is: ,So Will be used as the final CSI data input.

[0030] At the same time, the present invention provides:

[0031] A server includes a processor and a memory, wherein the memory stores at least one program, and the program is loaded and executed by the processor to implement the above-mentioned OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals.

[0032] A computer-readable storage medium stores at least one program, which is loaded and executed by a processor to implement the OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals.

[0033] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0034] 1. Based on the sensitivity of subcarriers to gesture changes and the role of data transmission, the present invention divides subcarriers with larger CSI amplitudes into two sets: Contains subcarriers with significant CSI amplitude changes, set Contains virtual subcarriers with stable CSI amplitudes that are not used for data transmission.

[0035] 2. The present invention takes into account the correlation between adjacent subcarriers and sets the subcarrier spacing. , right Grouping processing is performed to maintain the common change trend reflected by the original CSI and the correlation between adjacent CSIs.

[0036] 3. The present invention provides three selection strategies for processing each group of CSI, selecting the average CSI value in each subcarrier group to form a set ; Select the maximum CSI value in each subcarrier group to form a set ;Select the CSI value corresponding to the first subcarrier of each group to form a set .

[0037] 4. The present invention reduces the CSI input data dimension to a smaller number by subcarrier grouping and selecting a strategy for processing CSI. , so as to enable efficient model training and real-time processing on terminals with limited computing resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is a diagram of CSI amplitude information on 256 subcarriers.

[0039] Figure 2 For When, different Schematic diagram of gesture recognition accuracy and single input CSI file size under certain number.

[0040] Figure 3 For When, different Schematic diagram of gesture recognition accuracy and single input CSI file size under certain number. Figure 4 Schematic diagram of gesture recognition accuracy under different selection strategies. DETAILED DESCRIPTION

[0041] The present invention is further described in detail below in conjunction with embodiments and drawings, but the embodiments of the present invention are not limited thereto.

[0042] The OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals includes the following steps:

[0043] S1. According to the corresponding CSI amplitudes on different subcarriers, subcarriers with larger CSI amplitudes are selected from all subcarriers, and their corresponding CSI amplitudes constitute a set , the number of subcarriers included is .like Figure 1 In this example, from 256 subcarriers, subcarriers with indexes No.126 to No.187 are selected. subcarriers, the subcarrier feature dimension of CSI data is reduced from 256 to .

[0044] S2. According to the sensitivity of CSI amplitude on different subcarriers to gesture action response, the set Regroup the definitions.

[0045] S21, CSI amplitude in time frame The subcarrier set that changes within ,in , indicating the subcarriers in a time frame The CSI amplitude within In this example , and its subcarrier corresponding indexes are No.133 to No.187.

[0046] S22, CSI amplitude in time frame The subcarrier set that remains stable within ,in , indicating the subcarriers in a time frame The CSI amplitude within In this example , the subcarrier corresponding indexes are No.126 to No.132.

[0047] S23, according to the set , Definition, set It can be expressed as: ,in ,and , .

[0048] The grouping method here can also directly use a static grouping method to evenly group all subcarriers at fixed intervals; or calculate the correlation between subcarriers and group them according to the degree of correlation.

[0049] S3, according to the subcarrier grouping interval , Respectively , Further grouping From the subcarrier dimension, it can be divided into Group.

[0050] S31, setting subcarrier grouping interval ,Will Divide The specific calculation formula is: ,in Indicates rounding up.

[0051] S32, setting subcarrier grouping interval ,Will Divide The specific calculation formula is: ,in Indicates rounding down and the final subcarriers are merged into the previous group.

[0052] S4. For the above The CSI amplitude corresponding to the subcarrier of the group is operated by a data selection strategy to operate the CSI in each group and generate the final CSI data input . The dimension is , reducing the input dimension of the original CSI from 256 to .

[0053] S41. Definition , ,in Indicated in The group index in .

[0054] S42. If the selection strategy of taking the average value of the CSI in each group (A-CSIS) is adopted, define For the The average value of CSI in the subcarrier group, or The CSI value corresponding to the subcarrier with the highest volatility in the subcarrier group, where the subcarrier with the highest volatility is the subcarrier with the largest standard deviation or variance of the CSI data;

[0055] The calculation formula is:

[0056]

[0057] The obtained average CSI set is: ,So Will be used as the final CSI data input.

[0058] S43. If the selection strategy of taking the maximum value of the CSI in each group (M-CSIS) is adopted, define For the time frame In terms of dimension, The maximum CSI value in a subcarrier group is calculated as follows:

[0059]

[0060] The maximum CSI set obtained is: ,So Will be used as the final CSI data input.

[0061] S44. If a discrete selection strategy (D-CSIS) is adopted for CSIs in each group, define For the The CSI value corresponding to a subcarrier in the group (which can be the first subcarrier, the middle subcarrier, or the last subcarrier) is calculated as follows:

[0062]

[0063] The obtained discrete CSI set is: ,So Will be used as the final CSI data input.

[0064] like Figure 2 , in fixed In the case of The size of a single CSI data file is also gradually reduced, but the gesture recognition effect is The A-CSIS strategy is used in this comparison because the average value is the best in the comparison of different strategies.

[0065] like Figure 3 , in fixed In the case of The number of CSI data is further reduced. When the CSI dimension is reduced from 256 to 13, the best compromise between gesture recognition accuracy and CSI dimension can be obtained. The CSI dimension is reduced from 256 to 13, and the input file size is reduced from 101 KB to 6 KB, but the recognition accuracy can still be maintained at nearly 95%. The A-CSIS strategy is used in this comparison because it is found that the average value has the best effect in the comparison of different strategies.

[0066] like Figure 4 ,A-CSIS strategy significantly outperforms other strategies in gesture recognition accuracy because it better preserves the characteristics of channel changes.,M-CSIS and D-CSIS have similar performances, while random CSI selection (R-CSIS) performs the worst in terms of accuracy.

[0067] The present invention takes into account the correlation between adjacent subcarriers during grouping, and by making full use of the correlation information between subcarriers, it is possible to further reduce redundant data and optimize the grouping process of CSI data.

[0068] The prior art only statically selects 62 subcarriers, while the present invention introduces a variety of subcarrier selection strategies (A-CSIS, M-CSIS, D-CSIS), which can dynamically process CSI according to actual needs.

[0069] Experimental results show that the present invention further reduces the dimension of CSI input data from 62 to , which reduces the storage requirement while maintaining the high accuracy of the gesture recognition task.

[0070] At the same time, this embodiment also provides:

[0071] A server includes a processor and a memory, wherein the memory stores at least one program, and the program is loaded and executed by the processor to implement the above-mentioned OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals.

[0072] A computer-readable storage medium stores at least one program, which is loaded and executed by a processor to implement the OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals.

[0073] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals, characterized in that: The following steps are involved: S1. According to the corresponding CSI amplitudes on different subcarriers, subcarriers with CSI amplitudes greater than a first preset value are selected from all subcarriers, and their corresponding CSI amplitudes constitute a set , the number of subcarriers included is , the subcarrier feature dimension of CSI data is reduced to ; S2. Collection CSI amplitude in time frame The subcarrier set that changes within constitutes a set , the remaining subcarrier sets constitute the set ;gather ;in, ,and , , For collection The number of subcarriers included, For collection The number of subcarriers included; S3, according to the subcarrier grouping interval , Separately for the set ,gather Further grouping, corresponding to the formation of sets ,gather ,gather ; S4. For the collection The CSI amplitude corresponding to the subcarrier is used to operate the CSI in each group using a data selection strategy and generate the final CSI data input ; The data selection strategy is one of the selection strategies of taking the average value, taking the maximum value, and discrete selection.

2. The OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals according to claim 1, characterized in that: In step S4, the selection strategy for taking the average value is specifically: definition , ,in Indicates in the collection The group index in ; definition For the The average value of CSI in the subcarrier groups, The calculation formula is: The obtained average CSI set is: ,So Will be used as the final CSI data input.

3. The OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals according to claim 1, characterized in that: In step S4, the selection strategy for taking the maximum value is specifically: definition , ,in Indicates in the collection The group index in ; definition For the time frame In terms of dimension, The maximum CSI value in the subcarrier group, or The CSI value corresponding to the subcarrier with the highest fluctuation rate in the subcarrier group; The specific calculation formula is: ; The maximum CSI set obtained is: ,So Will be used as the final CSI data input.

4. The OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals according to claim 3 is characterized in that: The subcarrier with the highest fluctuation rate is the subcarrier with the largest standard deviation or variance of the CSI data.

5. The OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals according to claim 1, characterized in that: In step S4, the discrete selection strategy is specifically: definition , ,in Indicates in the collection The group index in ; definition For the The CSI value corresponding to a subcarrier in the group is calculated as follows: The obtained discrete CSI set is: ,So Will be used as the final CSI data input.

6. A server, comprising a processor and a memory, characterized in that: At least one program is stored in the memory, and the program is loaded and executed by the processor to implement the OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals as described in any one of claims 1 to 5.

7. A computer-readable storage medium, wherein at least one program is stored in the storage medium, characterized in that: The program is loaded and executed by a processor to implement the OFDM subcarrier grouping and selection method for intelligent identification of WIFI signals as described in any one of claims 1 to 5.

Images