A communication and sensing integrated implementation method and device based on an existing WiFi communication system

Abstract

The application discloses a communication and sensing integrated implementation method and device based on an existing WiFi communication system, in order to solve the inconsistency problem of the CSI collected based on the communication packet, a calibration method is proposed to normalize the CSI under different communication modes to the same mode, and a compensation method is proposed to remove the influence of beam forming. In order to solve the uneven problem of the CSI collected based on the communication packet, an incentive strategy is proposed to ensure that the number of collected CSIs is sufficient and key sensing information is not missed, and a fitting resampling strategy is proposed to reconstruct the collected CSIs into equally spaced distributed CSIs. In combination with the above scheme, the communication and sensing integrated implementation method based on the existing WiFi communication system can realize the sensing application by collecting the existing communication data packets in the communication scene, and can effectively reduce the additional bandwidth resource overhead without affecting the communication performance as much as possible.

Description

Technical Field

[0001] This invention relates to a method for integrating communication and sensing, specifically, a method and apparatus for integrating communication and sensing based on an existing WiFi communication system. Background Technology

[0002] The development of traditional mobile communication technology aims to achieve higher communication speeds, lower latency, wider coverage, and more users. With the advent of the intelligent wave, mobile communication technology is no longer limited to pursuing communication performance alone, but rather to providing intelligent services. For example, sixth-generation mobile communication technology (6G) aims to propel communication from "human connectivity" and "Internet of Things" to "Intelligent Internet of Things," ushering in an intelligent society. However, intelligent services require not only the support of artificial intelligence (AI) foundations but also wireless sensing capabilities. On the one hand, the massive amounts of sensor data collected can serve as the training basis for AI algorithms; on the other hand, sensing capabilities enable mobile communication networks to perceive their surroundings in real time, thereby achieving real-time interaction and feedback. To simultaneously realize communication and sensing technologies, Integrated Communication and Sensing (ISAC) technology has emerged. This technology is based on the fact that communication and sensing have extremely similar hardware structures and use almost identical spectrum. Therefore, ISAC is expected to significantly improve spectrum efficiency and resource utilization by simultaneously achieving high-performance sensing and communication functions on the same hardware device and the same spectrum.

[0003] Existing research on integrated communication and sensing technologies aims to achieve communication and sensing functions by designing entirely new hardware, waveforms, architectures, and protocols. For example, the next-generation wireless local area network (WLAN) protocol, IEEE 802.11bf, aims to add sensing-oriented protocols to the existing WiFi protocol, thereby achieving sensing functions for next-generation WiFi. While redesigning network architecture, communication protocols, and physical layer design holds promise for reaching the limits of communication and sensing performance, the cost is the lengthy time and enormous expense required for iterative communication equipment iterations. For instance, from the commercial launch of 5G base stations in 2019 to the end of 2022, 5G base stations in China accounted for only about 20% of the total number of existing base stations. Meanwhile, WiFi 6 devices began commercial use in 2019, but by 2022, WiFi 6 devices only accounted for 24% of the WiFi market. Therefore, achieving integrated communication and sensing technologies in next-generation communication technologies remains a long and arduous process.

[0004] On the other hand, in WiFi communication systems, each WiFi data packet header contains pilot information. By listening to the pilot information, the channel state information (CSI) between the wireless WiFi router and the current device can be measured. Since this channel state information contains information about the surrounding environment and the sensing target, it is expected to be used to establish a mapping relationship from CSI to various sensing tasks such as positioning, tracking, and status of the sensing target. However, since the design of the communication system does not take into account the needs of sensing, using the CSI directly collected from the communication system for sensing will face two major problems. (1) Inconsistent CSI. The communication system will use diversity and multiplexing modes according to the user's needs, and will also use beamforming technology to provide throughput. These technologies will lead to inconsistent CSI collected in the same scene and under the same sensing target state, making it difficult to establish an inference model from CSI to the sensing target. (2) Uneven CSI. The density of communication packets depends entirely on the communication needs in the scene. When the communication demand is low, the number of CSI collected will also be low, which may lead to the loss of key sensing information and thus cause errors in the sensing results. Furthermore, sensing often requires consistent CSI time intervals, while CSI time intervals acquired through communication are unevenly distributed. Summary of the Invention

[0005] This invention addresses the challenge of implementing integrated communication sensing technology within existing communication systems by providing a method and apparatus for achieving integrated communication sensing based on an existing WiFi communication system. By monitoring communication packets emitted by a wireless WiFi router, the integrated communication sensing device can estimate the Communication Signal Indicator (CSI) from the router to the device using pilot data in the packet header, and then use the CSI to perceive the surrounding environment. Specifically, to address the inconsistency problem, calibration and compensation relationships are designed; to address the non-uniformity problem, excitation and fitting resampling strategies are designed. This invention is achieved through the following technical solutions:

[0006] A method for integrating communication and sensing based on an existing WiFi communication system includes a mapping relationship establishment stage and a sensing stage; the mapping relationship establishment stage includes the following steps:

[0007] 1) Within the coherent time, obtain multiple sets of communication data packets under diversity and multiplexing communication modes, and use the pilots in the obtained communication data packets to calculate channel state information (CSI);

[0008] 2) Calculate the calibration matrix from the CSI corresponding to the multiplexing mode to the CSI corresponding to the diversity mode;

[0009] 3) Obtain a large number of CSIs under different perception target states, and use algorithms such as neural networks to establish an inference model from CSIs to perception target states;

[0010] The perception phase includes the following steps:

[0011] 4) The incentive strategy controls the frequency of communication with the WiFi router, thereby incentivizing the WiFi router to emit a sufficient number of communication packets;

[0012] 5) Obtain feedback on communication packets and beamforming. Calculate CSI based on the pilot signals in the collected communication packets. Determine whether the WiFi router uses beamforming technology based on the information in the communication packet header. If it does, compensate for the CSI using the obtained beamforming feedback; otherwise, proceed directly to the next step.

[0013] 6) Based on the obtained communication packet header information, determine whether the currently obtained CSI is in diversity communication mode or multiplexing communication mode. If it is in diversity communication mode, proceed directly to the next step; if it is in multiplexing communication mode, use the calibration matrix obtained in step 2) to calibrate the obtained CSI to diversity mode.

[0014] 7) Repeat steps 4)-6) to obtain a large number of CSIs, and fit and resample the CSIs in the time dimension;

[0015] 8) Use the inference model established in step 3) to process the fitted resampled CSI to obtain the perceived target state information.

[0016] As a further improvement, step 2) of this invention calculates the calibration matrix between the CSI corresponding to the multiplexing mode and the CSI corresponding to the diversity mode, thereby ensuring the consistency of the collected CSIs, specifically as follows:

[0017] CSI corresponding to the reuse mode, g k,2 , and the CSI corresponding to diversity mode, g k,1 The relationship between them can be constructed as

[0018] g k,1 =g k,2 P k ,

[0019] Among them, P k To obtain the calibration matrix, where k represents the k-th subcarrier, the following problem can be solved using the least squares method:

[0020]

[0021] As a further improvement, in step 4) of this invention, the incentive strategy controls the frequency of communication with the WiFi router, thereby incentivizing the WiFi router to transmit a sufficient number of communication packets, specifically:

[0022] The incentive strategy has two states: a silent state and an incentive state. When the actual CSI sampling rate is higher than the required CSI sampling rate, it is in the silent state; when the actual CSI sampling rate is lower than the required CSI sampling rate, it is in the incentive state, communicating with the WiFi router, and the frequency of communication is the difference between the actual CSI sampling rate and the required CSI sampling rate.

[0023] As a further improvement, in step 5) of the present invention, the obtained beamforming feedback is used to compensate for CSI, specifically as follows:

[0024] First, based on the WiFi protocol (IEEE 802.11 series), the channel state information between the communication receiver and the WiFi router is reconstructed from the collected beamforming feedback, denoted as... Based on classic beamforming schemes (such as zero-forcing beamforming, etc.), utilizing Reconstruct the beamforming matrix V at the WiFi router end k Finally, regarding the CSI,h obtained using communication data packets k Compensation will be provided, specifically as follows:

[0025]

[0026] This invention also discloses a communication sensing integrated implementation device based on an existing WiFi communication system, comprising:

[0027] The first acquisition module: within the coherence time, it acquires multiple sets of communication data packets under diversity and multiplexing communication modes, and uses the pilots in the acquired communication data packets to calculate the channel state information (CSI).

[0028] First calculation module: Calculates the calibration matrix between the CSI corresponding to the multiplexing mode and the CSI corresponding to the diversity mode;

[0029] The second computing module: obtains a large number of CSIs under different perception target states, and uses algorithms such as neural networks to establish an inference model between CSIs and perception target states;

[0030] The perception phase includes the following steps:

[0031] First transmission module: used to run incentive strategies to control the frequency of communication with the WiFi router, thereby incentivizing the WiFi router to transmit a sufficient number of communication packets;

[0032] The second acquisition module is used to obtain communication packets and beamforming feedback. It calculates the CSI based on the pilot signals in the acquired communication packets and determines whether the WiFi router uses beamforming technology based on the information in the communication packet header. If it does, it compensates for the CSI using the obtained beamforming feedback; if it does not, it proceeds directly to the next step.

[0033] The third calculation module is used to determine whether the currently obtained CSI is in diversity communication mode or multiplexing communication mode based on the obtained communication packet header information. If it is in diversity communication mode, proceed directly to the next step; if it is in multiplexing communication mode, use the calibration matrix obtained in step 2) to calibrate the obtained CSI to diversity mode.

[0034] The fourth calculation module is used to repeat steps 4)-6) to obtain a large number of CSIs, and to fit and resample the CSIs in the time dimension;

[0035] The fifth calculation module is used to process the fitted resampled CSI using the inference model established in step 3) to obtain the perceived target state information.

[0036] The beneficial effects of this invention are as follows:

[0037] This invention utilizes an existing WiFi communication system to measure Communication-Sensing Indicators (CSIs) containing information about the target and its surroundings by monitoring communication packets within a scene, thus achieving integrated communication and sensing. Specifically, to address the inconsistency of CSIs acquired from communication packets, a calibration method is proposed to normalize CSIs from different communication modes to the same mode, and a compensation method is proposed to remove the influence of beamforming. To address the unevenness of CSIs acquired from communication packets, an excitation strategy is proposed to ensure a sufficient number of CSIs are acquired, preventing the omission of crucial sensing information, and a fitting resampling strategy is proposed to reconstruct the acquired CSIs into equally spaced CSIs. Combining the above solutions, the integrated communication and sensing implementation method based on an existing WiFi communication system proposed in this invention can achieve sensing applications by acquiring existing communication data packets within the communication scene, effectively reducing additional bandwidth resource overhead with minimal impact on communication performance. Attached Figure Description

[0038] Figure 1 This is a flowchart illustrating the method. Detailed Implementation

[0039] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments:

[0040] The purpose of this invention is to achieve integrated communication and sensing in existing WiFi communication systems by acquiring Communication Sensor Identity (CSI) through listening to communication packets. To address the inconsistency issue in CSI, calibration and compensation methods are proposed. Furthermore, to resolve the non-uniformity issue in CSI, excitation and fitting resampling strategies are proposed.

[0041] Figure 1 This is a flowchart illustrating the process, which includes two phases: the mapping relationship establishment phase and the perception phase. The mapping relationship establishment phase includes the following steps:

[0042] 1) Within the coherence time, acquire multiple sets of communication data packets in diversity and multiplexing communication modes, and calculate the CSI using the pilot signals in the acquired communication data packets. Specifically, for the 802.11a / g protocol, the L-STF and L-LTF fields can be used for CSI estimation; for the 802.11n protocol, the HT-STF and HT-LTF fields can be used; for the 802.11ac protocol, the VHT-STF and VHT-LTF fields can be used; and for the 802.11ax protocol, the HE-STF and HE-LTF fields can be used.

[0043] 2) Calculate the calibration matrix from the CSI corresponding to the multiplexing mode to the CSI corresponding to the diversity mode. Specifically, let g be the CSI corresponding to the multiplexing mode. k,2 The CSI corresponding to diversity mode is g. k,1 Therefore, their relationship can be structured as follows:

[0044] g k,1 =g k,2 P k ,

[0045] Among them, P k Let k represent the k-th subcarrier, and k be the desired calibration matrix. To obtain the calibration matrix, the following problem can be solved using the least squares method:

[0046]

[0047] 3) Obtain a large number of CSIs under different perceived target states (e.g., walking, running, standing, sitting, standing up, etc.). Simultaneously, build deep learning-based neural networks such as CNNs and DNNs using tools like PyTorch and TensorFlow. Then, use the collected CSIs to fully train the network, thereby obtaining an inference model from CSIs to perceived target states. Generally, to ensure the effectiveness of training, at least 100 CSI samples should be collected for each action.

[0048] The perception phase includes the following steps:

[0049] 4) The incentive strategy controls the frequency of communication with the WiFi router, thereby incentivizing the WiFi router to transmit a sufficient number of communication packets. Specifically, the incentive strategy has two states: a silent state and an active state. When the actual CSI sampling rate is higher than the required CSI sampling rate, it is in the silent state. For example, when the actual CSI sampling rate is 200 Hz and the required CSI sampling rate is 100 Hz, it is in the silent state. When the actual CSI sampling rate is lower than the required CSI sampling rate, it is in the active state, communicating with the WiFi router at the frequency of the difference between the actual CSI sampling rate and the required CSI sampling rate. For example, when the actual CSI sampling rate is 70 Hz and the required CSI sampling rate is 100 Hz, the communication frequency with the WiFi router is 30 Hz.

[0050] 5) Obtain communication packet and beamforming feedback. Calculate CSI based on the pilot signals in the acquired communication packets. Determine whether the WiFi router uses beamforming technology based on the information in the packet header. Specifically, this can be determined based on the SIG field in the packet header. If it is used, then compensate for the CSI using the obtained beamforming feedback. Specifically, first, according to the WiFi protocol (IEEE 802.11a / g / n / ac / ax), reconstruct the channel state information between the communication receiver and the WiFi router from the acquired beamforming feedback, denoted as... Based on classic beamforming schemes, utilizing Reconstruct the beamforming matrix V at the WiFi router end k For example, when using zero-forcing beamforming, V k for in(·) H This represents the conjugate transpose operation. Finally, for the CSI obtained using communication data packets, denoted as h... k To provide compensation, that is If not used, proceed directly to the next step.

[0051] 6) Based on the obtained communication packet header information, determine whether the currently obtained CSI is in diversity communication mode or multiplexing communication mode. If it is in diversity communication mode, proceed directly to the next step; if it is in multiplexing communication mode, use the calibration matrix obtained in step 2) to calibrate the obtained CSI to diversity mode, i.e., g. k,1 =g k,2 P k .

[0052] 7) Repeat steps 4)-6) to obtain a large number of CSIs, and fit and resample the CSIs along the time dimension. For example, fitting and resampling can be done using cubic splines, with the formula y = at. 3 +bt 2 +ct+d.

[0053] 8) Use the CNN model established in step 3) to process the fitted resampled CSI to obtain the perceived target state information.

[0054] This invention also discloses a communication sensing integrated implementation device based on an existing WiFi communication system, comprising:

[0055] The first acquisition module: within the coherence time, it acquires multiple sets of communication data packets under diversity and multiplexing communication modes, and uses the pilots in the acquired communication data packets to calculate the channel state information (CSI).

[0056] First calculation module: Calculates the calibration matrix between the CSI corresponding to the multiplexing mode and the CSI corresponding to the diversity mode;

[0057] The second computing module: obtains a large number of CSIs under different perception target states, and uses algorithms such as neural networks to establish an inference model between CSIs and perception target states;

[0058] The perception phase includes the following steps:

[0059] First transmission module: used to run incentive strategies to control the frequency of communication with the WiFi router, thereby incentivizing the WiFi router to transmit a sufficient number of communication packets;

[0060] The second acquisition module is used to obtain communication packets and beamforming feedback. It calculates the CSI based on the pilot signals in the acquired communication packets and determines whether the WiFi router uses beamforming technology based on the information in the communication packet header. If it does, it compensates for the CSI using the obtained beamforming feedback; if it does not, it proceeds directly to the next step.

[0061] The third calculation module is used to determine whether the currently obtained CSI is in diversity communication mode or multiplexing communication mode based on the obtained communication packet header information. If it is in diversity communication mode, proceed directly to the next step; if it is in multiplexing communication mode, use the calibration matrix obtained in step 2) to calibrate the obtained CSI to diversity mode.

[0062] The fourth calculation module is used to repeat steps 4)-6) to obtain a large number of CSIs, and to fit and resample the CSIs in the time dimension;

[0063] The fifth calculation module is used to process the fitted resampled CSI using the inference model established in step 3) to obtain the perceived target state information.

[0064] The above description is not intended to limit the present invention. It should be noted that, for those skilled in the art, various changes, modifications, additions or substitutions can be made without departing from the essential scope of the present invention, and these improvements and refinements should also be considered within the scope of protection of the present invention.

Claims

1. A method for integrating communication and sensing based on an existing WiFi communication system, characterized in that, This includes the mapping relationship establishment stage and the perception stage; The mapping relationship establishment phase includes the following steps: 1) Within the coherent time, obtain multiple sets of communication data packets under diversity and multiplexing communication modes, and use the pilots in the obtained communication data packets to calculate the channel state information (CSI); 2) Calculate the calibration matrix between the CSI corresponding to the multiplexing communication mode and the CSI corresponding to the diversity communication mode; 3) Obtain a large number of CSIs under different perception target states, and use neural network algorithms to establish an inference model from CSIs to perception target states; The perception phase includes the following steps: 4) The incentive strategy controls the frequency of communication with the WiFi router, thereby incentivizing the WiFi router to emit a sufficient number of communication packets; 5) Obtain feedback on communication packets and beamforming. Calculate CSI based on the pilot signals in the collected communication packets. Determine whether the WiFi router uses beamforming technology based on the information in the communication packet header. If it does, compensate for the CSI using the obtained beamforming feedback; otherwise, proceed directly to the next step. 6) Based on the obtained communication packet header information, determine whether the currently obtained CSI is in diversity communication mode or multiplexing communication mode. If it is in diversity communication mode, proceed directly to the next step. If it belongs to the multiplexing communication mode, the obtained CSI will be calibrated to the diversity communication mode using the calibration matrix obtained in step 2). 7) Repeat steps 4)-6) to obtain a large number of CSIs, and fit and resample the CSIs in the time dimension; 8) Use the inference model established in step 3) to process the fitted resampled CSI, thereby obtaining the perceived target state information; Step 2) calculates the calibration matrix between the CSI corresponding to the multiplexing communication mode and the CSI corresponding to the diversity communication mode, thereby ensuring the consistency of the collected CSIs. Specifically: CSI corresponding to multiplexing communication modes CSI, corresponding to diversity communication mode The relationship between them is constructed as follows: ； in, To obtain the calibration matrix, Representing the To obtain the calibration matrix using subcarriers, the following problem is solved using the least squares method: ； In step 4), the incentive strategy controls the frequency of communication with the WiFi router, thereby incentivizing the WiFi router to transmit a sufficient number of communication packets. Specifically: The incentive strategy has two states: a silent state and an incentive state. When the actual CSI sampling rate is higher than the required CSI sampling rate, it is in the silent state; when the actual CSI sampling rate is lower than the required CSI sampling rate, it is in the incentive state, communicating with the WiFi router, and the frequency of communication is the difference between the actual CSI sampling rate and the required CSI sampling rate. In step 5), the obtained beamforming feedback is used to compensate for CSI, specifically as follows: First, based on the WiFi protocol, the channel state information between the communication receiver and the WiFi router is reconstructed from the collected beamforming feedback, denoted as . Based on the classic beamforming scheme, utilizing Reconstruct the beamforming matrix at the WiFi router end Ultimately, for CSI obtained using communication data packets, Compensation will be provided, specifically as follows: 。

Images