Communication support'invisible 'anti-climbing system and method based on millimeter wave radar

Abstract

The invention relates to the technical field of communication support'invisible 'anti-climbing systems, and discloses a communication support'invisible' anti-climbing system and method based on millimeter-wave radar, and the anti-climbing system comprises a millimeter-wave radar module, a signal processing unit, a neural network reasoning unit, an alarm unit, a power management unit and a remote operation and maintenance platform. Through the system and the method, key time can be bought for emergency response, the target identification precision and the emergency response efficiency are improved, and the protection effect on the communication support is improved.

Description

Technical Field

[0001] This invention belongs to the technical field of "invisible" anti-climbing systems for communication brackets, specifically a "invisible" anti-climbing system and method for communication brackets based on millimeter-wave radar. Background Technology

[0002] With the accelerated construction of 5G networks, the distribution density of communication base stations has significantly increased, with base stations widely distributed across urban rooftops and suburban towers. While this widespread distribution ensures network coverage, it also exposes significant security risks. Furthermore, in recent years, unauthorized climbing and theft of communication equipment have been frequent, leading to serious consequences such as equipment damage, network outages, and even personal injury. Existing security measures have the following shortcomings: (1) Physical fences: affect the city's appearance and are costly; (2) Infrared beam: susceptible to interference from rain, snow, and birds, resulting in a high false alarm rate; (3) Visible light camera devices: require supplemental lighting, involve privacy, and have low nighttime recognition rate; (4) Vibration sensor: cannot distinguish between vibrations caused by people, animals and wind loads.

[0003] To address this, we propose an "invisible" anti-climbing system and method for communication brackets based on millimeter-wave radar. Summary of the Invention

[0004] The purpose of this invention is to improve the accuracy of target identification and emergency response efficiency, and to enhance the protective effect on communication brackets. This application provides a "stealth" anti-climbing system and method for communication brackets based on millimeter-wave radar.

[0005] The technical solution adopted in this invention is as follows: A "stealth" anti-climbing system and method for a communication bracket based on millimeter-wave radar, the anti-climbing system comprising a millimeter-wave radar module, a signal processing unit, a neural network inference unit, an alarm unit, a power management unit, and a remote operation and maintenance platform; The millimeter-wave radar module is installed at a height of 0.5 m–2 m on the communication support, with its antenna flush with the metal surface of the support. It transmits raw radar data to the signal processing unit in real time via a high-speed serial interface. The signal processing unit periodically acquires radar echoes and generates Range-Doppler maps and micro-Doppler spectra, transmitting the feature data to the neural network inference unit via a low-latency serial bus. The lightweight CNN model mounted on the neural network inference unit performs real-time classification of the input data, distinguishing between humans, animals, and inanimate targets. The alarm unit includes an audible and visual alarm, a LoRa and NB-IoT communication module. When the neural network inference unit detects climbing behavior, the classification result triggers the audible and visual alarm of the alarm unit. The alarm unit will upload the alarm information to the remote operation and maintenance platform through the LoRa and NB-IoT communication module. After receiving the alarm frame, the remote operation and maintenance platform can link the camera to start video evidence collection and provide multi-level alarm response through SMS or audible and visual alerts.

[0006] In a preferred embodiment of the invention, the power management unit draws power from a 48V DC power supply on the tower, and then steps it down to a stable voltage of 5V / 0.1A via a high-efficiency DC / DC converter, providing low-power power to the entire system.

[0007] In a preferred embodiment of the invention, the neural network inference unit is used to trigger an audible and visual alarm and generate an alarm frame when a person is identified as “human” in N_frames ≥ 3 consecutive frames and the distance d from the support is ≤ 30 cm.

[0008] In a preferred embodiment of the invention, the alarm unit is used to upload the alarm frame to the operation and maintenance platform via LoRa and NB-IoT.

[0009] In a preferred embodiment of the invention, the alarm frame includes at least a radar ID, a UTC timestamp, GPS coordinates, and a URL of a live photo.

[0010] In a preferred embodiment of the invention, the millimeter-wave radar is a 60 GHz FMCW radar with a bandwidth ≥ 4 GHz.

[0011] In a preferred embodiment of the invention, the neural network is a lightweight CNN quantized with INT8, and the model size is ≤ 300 kB.

[0012] In a preferred embodiment of the invention, a method for making a communication support "invisible" and preventing it from climbing is provided, based on millimeter-wave radar: S1: Radar deployment steps: Conformally install a 60 GHz FMCW radar within a height range of 0.5 m–2 m on the communication support, with the radar antenna flush with the metal surface of the support to achieve "stealth". S2: Signal acquisition steps: acquire radar echoes in periods of 20 ms–100 ms to generate Range-Doppler diagrams and micro-Doppler spectra; S3: Target recognition step: Input the Range-Doppler graph into a lightweight convolutional neural network and output the target category and confidence level; S4: Alarm triggering steps: When N_frame ≥ 3 consecutive frames are identified as "human" and the distance d from the support is ≤ 30cm, the audible and visual alarm is triggered and a LoRa alarm frame is generated; S5: Information reporting steps: Upload alarm frames to the operation and maintenance platform via LoRa or NB-IoT. The alarm frame fields include: radar ID, UTC timestamp, GPS coordinates, and on-site photo URL.

[0013] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: 1. In this invention, a lightweight CNN model neural network inference unit is used to achieve millimeter-level differentiation between biological and non-biological targets through dual feature fusion of micro-Doppler spectrum and Range-Doppler image. A dual verification mechanism of 3-frame continuous recognition and distance threshold is adopted to effectively filter interfering targets such as birds and fallen leaves. Moreover, alarm information including GPS coordinates, timestamps and photo URLs is uploaded to the operation and maintenance platform through LoRa and NB-IoT, which buys critical time for emergency response, improves the accuracy of target recognition and the efficiency of emergency response, and enhances the protection effect of communication support.

[0014] 2. In this invention, by conformally mounting the millimeter-wave radar antenna to the metal surface of the bracket, the exposed structure of traditional infrared beam or vibration sensors is eliminated, achieving physical invisibility, avoiding human-caused damage, obstruction, or disassembly, improving the invisibility protection capability of the communication bracket, and thus improving the protection effect.

[0015] 3. In this invention, through 48V to 5V high-efficiency power management and LoRa and NB-IoT dual-mode communication, the system standby power consumption is less than 0.5W, supporting independent maintenance of radar unit and signal processing unit, unaffected by light, rain, snow and haze, and supporting long-term deployment of remote base stations without mains power support. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the anti-climbing system in this invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] The following will combine Figure 1This invention provides a detailed description of an "invisible" anti-climbing system and method for a communication bracket based on millimeter-wave radar, according to an embodiment of the present invention. Example 1

[0019] Reference Figure 1 A "stealth" anti-climbing system and method for communication brackets based on millimeter-wave radar is disclosed. The anti-climbing system includes a millimeter-wave radar module, a signal processing unit, a neural network inference unit, an alarm unit, a power management unit, and a remote operation and maintenance platform. The millimeter-wave radar module is installed at a height of 0.5 m–2 m on the communication bracket, with its antenna flush with the metal surface of the bracket. It transmits raw radar data to the signal processing unit in real time via a high-speed serial interface. The millimeter-wave radar is a 60 GHz FMCW radar with a bandwidth ≥ 4 GHz. Specifically, by conformally mounting the millimeter-wave radar antenna to the metal surface of the bracket, the exposed structure of traditional infrared beam or vibration sensors is eliminated, achieving physical stealth and avoiding human-caused damage, obstruction, or disassembly, thereby improving the stealth protection capability of the communication bracket and thus enhancing the protection effect. Example 2

[0020] Reference Figure 1 The signal processing unit is used to periodically acquire radar echoes and generate Range-Doppler maps and micro-Doppler spectra, and transmit the feature data to the neural network inference unit through a low-latency serial bus. The lightweight CNN model on the neural network inference unit performs real-time classification of the input data, distinguishing between humans, animals, or non-biological targets. The neural network inference unit is used to trigger an audible and visual alarm and generate an alarm frame when a "human" is identified in N_frames ≥ 3 consecutive frames and the distance d from the support is ≤ 30 cm. The neural network is a lightweight CNN quantized with INT8, and the model size is ≤ 300 kB. Specifically, the system extracts the motion features of the target, such as speed, distance, and micro-motion frequency, through Range-Doppler maps and micro-Doppler spectra. It adopts a dual verification mechanism of 3 consecutive frames recognition and distance threshold to effectively filter out interfering targets such as birds and fallen leaves. Combined with the lightweight CNN model, it can achieve real-time classification of humans, animals, and non-biological targets.

[0021] Reference Figure 1The alarm unit includes an audible and visual alarm, a LoRa and NB-IoT communication module. When the neural network inference unit detects climbing behavior, the classification result triggers the audible and visual alarm of the alarm unit. The alarm unit then uploads the alarm information to the remote operation and maintenance platform via the LoRa and NB-IoT communication modules. The alarm unit is used to upload the alarm frame to the operation and maintenance platform via LoRa and NB-IoT. The alarm frame includes at least the radar ID, UTC timestamp, and GPS information. The system provides coordinates and URLs of on-site photos. Upon receiving an alarm frame, the remote operation and maintenance platform can activate the camera to initiate video evidence collection and provide multi-level alarm responses via SMS or audible and visual alerts. Specifically, when the neural network inference unit detects abnormal climbing behavior, the system can simultaneously trigger a local audible and visual alarm and upload alarm information containing GPS coordinates, timestamps, and photo URLs to the operation and maintenance platform via LoRa and NB-IoT communication modules. Upon receiving the alarm frame, the platform can automatically activate the camera to initiate video evidence collection and push SMS notifications or trigger audible and visual alerts, thus gaining crucial time for emergency response, improving target recognition accuracy and emergency response efficiency, and enhancing the protection of the communication support structure.

[0022] Reference Figure 1 A method for "invisible" anti-climbing of a communication support based on millimeter-wave radar: S1: Radar deployment step: A 60 GHz FMCW radar is conformally installed within a height range of 0.5 m–2 m on the communication support, with the radar antenna flush with the metal surface of the support to achieve "invisibility"; S2: Signal acquisition step: Radar echoes are acquired at a period of 20 ms–100 ms to generate a Range-Doppler image and a micro-Doppler spectrum; S3: Target recognition step: The Range-Doppler image is input into a lightweight convolutional neural network to output the target category and confidence level; S4: Alarm triggering step: When N_frame ≥ 3 consecutive frames are identified as "human" and the distance d from the support is ≤ 30 cm, an audible and visual alarm is triggered and a LoRa alarm frame is generated; S5: Information reporting step: The alarm frame is uploaded to the operation and maintenance platform via LoRa or NB-IoT. The alarm frame fields include: radar ID, UTC timestamp, GPS... Coordinates and URLs of on-site photos; specifically, this multi-level response mechanism ensures both the rapid intervention capability of on-site personnel and real-time early warning from the remote monitoring center, forming a complete closed loop of "local deterrence - remote evidence collection - multi-level alarm". Example 3

[0023] Reference Figure 1The power management unit draws power from the 48V DC power supply on the tower, and steps it down to a stable 5V / 0.1A voltage via a high-efficiency DC / DC converter, providing low-power power to the entire system. Specifically, through 48V to 5V high-efficiency power management and LoRa and NB-IoT dual-mode communication, the system's standby power consumption is less than 0.5W, supporting independent maintenance of the radar unit and signal processing unit, unaffected by sunlight, rain, snow, or fog, and supporting long-term deployment of remote base stations without mains power support.

[0024] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0025] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A "stealth" anti-climbing system for a communication bracket based on millimeter-wave radar, characterized in that: The anti-climbing system includes a millimeter-wave radar module, a signal processing unit, a neural network inference unit, an alarm unit, a power management unit, and a remote operation and maintenance platform. The millimeter-wave radar module is installed at a height of 0.5 m–2 m on the communication support, with its antenna flush with the metal surface of the support. It transmits raw radar data to the signal processing unit in real time via a high-speed serial interface. The signal processing unit periodically acquires radar echoes and generates Range-Doppler maps and micro-Doppler spectra, transmitting the feature data to the neural network inference unit via a low-latency serial bus. The lightweight CNN model mounted on the neural network inference unit performs real-time classification of the input data, distinguishing between humans, animals, and inanimate targets. The alarm unit includes an audible and visual alarm, a LoRa and NB-IoT communication module. When the neural network inference unit detects climbing behavior, the classification result triggers the audible and visual alarm of the alarm unit. The alarm unit will upload the alarm information to the remote operation and maintenance platform through the LoRa and NB-IoT communication module. After receiving the alarm frame, the remote operation and maintenance platform can link the camera to start video evidence collection and provide multi-level alarm response through SMS or audible and visual alerts.

2. The "stealth" anti-climbing system for a communication bracket based on millimeter-wave radar as described in claim 1, characterized in that: The power management unit draws power from the 48V DC power supply on the tower, and the voltage is stepped down to a stable 5V / 0.1A by a high-efficiency DC / DC converter, providing low-power power to the entire system.

3. The "invisible" anti-climbing system for a communication bracket based on millimeter-wave radar as described in claim 1, characterized in that: The neural network inference unit is used to trigger an audible and visual alarm and generate an alarm frame when a person is identified as "human" in N_frames ≥ 3 consecutive frames and the distance d from the support is ≤ 30 cm.

4. The "invisible" anti-climbing system for a communication bracket based on millimeter-wave radar as described in claim 3, characterized in that: The alarm unit is used to upload the alarm frame to the operation and maintenance platform via LoRa and NB-IoT.

5. The "invisible" anti-climbing system for a communication bracket based on millimeter-wave radar as described in claim 3, characterized in that: The alarm frame includes at least the radar ID, UTC timestamp, GPS coordinates, and URL of the on-site photo.

6. The "stealth" anti-climbing system for a communication bracket based on millimeter-wave radar as described in claim 1, characterized in that: The millimeter-wave radar is a 60 GHz FMCW radar with a bandwidth of ≥ 4 GHz.

7. The "stealth" anti-climbing system for a communication bracket based on millimeter-wave radar as described in claim 1, characterized in that: The neural network is a lightweight CNN quantized with INT8, and the model size is ≤ 300 kB.

8. A method for preventing the "invisible" climbing of a communication bracket based on millimeter-wave radar, characterized in that... The apparatus of any one of claims 1 to 7 is used: S1: Radar deployment steps: Conformally install a 60 GHz FMCW radar within a height range of 0.5 m–2 m on the communication support bracket, with the radar antenna flush with the metal surface of the bracket to achieve "stealth"; S2: Signal acquisition steps: acquire radar echoes at a period of 20 ms–100 ms to generate Range-Doppler diagrams and micro-Doppler spectra; S3: Target recognition step: Input the Range-Doppler graph into a lightweight convolutional neural network and output the target category and confidence level; S4: Alarm triggering steps: When N_frame ≥ 3 consecutive frames are identified as "human" and the distance d from the support is ≤ 30 cm, the audible and visual alarm is triggered and a LoRa alarm frame is generated; S5: Information reporting steps: Upload alarm frames to the operation and maintenance platform via LoRa or NB-IoT. The alarm frame fields include: radar ID, UTC timestamp, GPS coordinates, and on-site photo URL.

Images