Detection system, sensor terminal, and wireless base station
The detection system integrates with existing wireless communication systems by using sensor packets to detect state events, ensuring accurate monitoring with minimal disruption and easy retrofitting.
Patent Information
- Application Number
- JP2023078803
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-05-11
- Publication Date
- 2024-11-21
AI Technical Summary
Existing wireless communication systems lack the ability to detect and monitor state events without disrupting existing WiFi devices and communication systems, and there is a need for a system that can integrate sensing functions with minimal retrofitting.
A detection system comprising a first wireless device, a second wireless device, a sensor packet extraction unit, and a sensing processing unit that utilizes sensor packets with changing contents based on the environment to detect state events, allowing for integration with existing wireless communication systems.
Enables accurate detection and monitoring of state events with minimal impact on existing wireless devices and communication systems, facilitating easy retrofitting and high-accuracy sensing.
Smart Images

Figure 2024162861000001_ABST
Abstract
Description
[Technical field]
[0001] The present invention relates to a detection system, a sensor terminal and a wireless base station for monitoring and detecting status events. [Background technology]
[0002] In order to detect and monitor status events using radio, technologies for detecting events based on classification of channel status information are described in Patent Document 1, Non-Patent Document 1, and Non-Patent Document 2.
[0003] Patent Document 1 describes a configuration related to an apparatus for detecting an event. The apparatus described in Patent Document 1 includes a processor and a storage device that confirms instructions that, when executed by the processor, cause the processor to train a classifier for classifying channel state information (CSI) and detecting an event based on the classification of the CSI acquired during a monitoring stage. Training the classifier includes acquiring, for each known event to be detected, training CSI of a wireless multipath channel between a wireless transmitter and a wireless receiver in the venue, the training CSI being derived from one or more probing signals transmitted from the transmitter to the receiver via the wireless multipath channel, during a period during which the known event occurs in the venue, and training the classifier based on the known event and the training CSI associated with each event.
[0004] Non-Patent Document 1 describes a Compressed Beamforming Report packet (hereinafter, CBR packet). The CBR packet is a standard defined for WiFi5 and later, and transmits information on the V matrix obtained by singular value decomposition of the propagation matrix H used in MIMO communication. Non-Patent Document 2 describes a configuration related to a communication standard for realizing WLAN-Sencing by expanding the technology of Non-Patent Document 1. [Prior art documents] [Patent documents]
[0005] [Patent Document 1] International Publication No. 2017 / 100706 [Non-patent literature]
[0006] [Non-Patent Document 1] IEEE Std 802.11-2016, Part11:Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications [Non-Patent Document 2] IEEE802.11bf Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications Amendment: Enhancements for Wireless Local Area Network (WLAN) Sensing), https: / / standards.ieee.org / ieee / 802.11bf / 10365 / Summary of the Invention [Problem to be solved by the invention]
[0007] Patent document 1 describes a classifier for detecting events based on classification of channel state information (CSI) of radio waves emitted by a wireless communication system, but does not consider at all how to detect and monitor wireless events in the case of a wireless communication system in which CSI cannot be acquired.
[0008] Furthermore, when introducing the sensing system using the CBR packets described in Non-Patent Document 1 and Non-Patent Document 2 into an environment such as an office or factory, it is necessary to newly introduce a wireless base station or wireless terminal that acquires and stores the CBR packets. In this case, it may be difficult to introduce new WiFi equipment into an environment such as an office or factory where WiFi equipment has already been introduced. Therefore, it is desirable to install only equipment that realizes the sensing function so as not to affect the existing WiFi equipment and communication system.
[0009] The present invention has been made in consideration of the above-mentioned problems, and aims to provide a detection system, a sensor device, and a wireless base station that can monitor and detect status events even in a wireless communication system that cannot acquire CSI information. Another aim is to provide a detection system, a sensor device, and a wireless base station that can achieve a sensing function with simple retrofit installation work without affecting existing wireless devices and communication systems. [Means for solving the problem]
[0010] A representative example of the invention disclosed in the present application is as follows: That is, a detection system for detecting a status event includes a first wireless device that transmits wireless packets, a second wireless device that receives the wireless packets, a wireless base station capable of wirelessly communicating with at least the first wireless device, a sensor packet extraction unit that extracts a sensor packet transmitted by the first wireless device from a packet received by the second wireless device, and a sensing processing unit that executes a sensing process to detect a status event indicating information of the real world using the extracted sensor packet, wherein the sensor packet is a packet used for communication control and the content of which changes depending on the surrounding environment. Effect of the Invention
[0011] According to one aspect of the present invention, it is possible to accurately detect and monitor a wireless event. Problems, configurations and advantages other than those described above will become apparent from the following description of the embodiments. [Brief description of the drawings]
[0012] [Figure 1] 1 is a diagram illustrating a configuration of an event detection system according to a first embodiment. [Diagram 2] FIG. 2 is a diagram illustrating a configuration of a sensor terminal according to the first embodiment. [Diagram 3] FIG. 2 is a diagram illustrating a configuration of a radio base station according to the first embodiment. [Figure 4] FIG. 2 is a diagram illustrating a configuration of a sensor terminal according to the first embodiment. [Diagram 5] FIG. 2 is a diagram showing a sensing process executed by the event detection system according to the first embodiment. [Figure 6] FIG. 13 is a diagram showing an example of a monitor screen on which an event detection result is output. [Figure 7] FIG. 11 is a diagram illustrating a configuration of an event detection system according to a second embodiment. [Figure 8] FIG. 11 is a diagram showing a sensing process executed by the event detection system according to the second embodiment. [Figure 9] 10 is a flowchart of a process for controlling a transmission interval of a sensing packet in the second embodiment. [Figure 10] 10 is a sequence diagram of communication in which the transmission interval of the sensing packet shown in FIG. 9 is controlled. [Figure 11] FIG. 13 is a diagram illustrating a configuration of an event detection system according to a third embodiment. [Figure 12] FIG. 13 is a diagram showing a process of complementing missing data in the third embodiment. [Figure 13] FIG. 13 is a diagram illustrating a configuration of an event detection system according to a fourth embodiment. [Figure 14] FIG. 13 is a diagram showing a sensing process using an unmanaged sensor packet transmitted by an unmanaged wireless terminal in the fourth embodiment. [Figure 15] FIG. 13 is a diagram showing a sensing process using an unmanaged sensor packet transmitted by an unmanaged wireless terminal in the fourth embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0013] In the following embodiments, when necessary for convenience, the description will be divided into multiple sections or embodiments. In the following embodiments, when the number of elements (including the number, numerical value, amount, range, etc.) is mentioned, it is not limited to the specific number, and may be more than or less than the specific number, unless otherwise specified or clearly limited to a specific number in principle. Note that in the following embodiments, the components (including processing steps, etc.) are not necessarily essential, except when otherwise specified or clearly considered essential in principle.
[0014] In addition, each of the configurations, functions, processing units, processing means, etc. in the following embodiments may be realized in part or in whole as, for example, an integrated circuit or other hardware. Also, each of the configurations, functions, processing units, processing means, etc. described below may be realized as a program executed on a computer. In other words, they may be realized as software. Information such as programs, tables, files, etc. that realize each of the configurations, functions, processing units, processing means, etc. may be stored in a storage device such as a memory, a hard disk, or an SSD (Solid State Drive), or a storage medium such as an IC card, an SD card, or a DVD.
[0015] Hereinafter, the embodiments of the present invention will be described in detail with reference to the drawings. In all the drawings for explaining the embodiments, the same or related reference numerals are used for the members having the same functions, and the repeated explanations are omitted. In the following embodiments, the explanations of the same or similar parts will not be repeated as a rule unless it is particularly necessary.
[0016] First Embodiment [System Configuration] The event detection system 200 of the first embodiment is composed of a server 201, a sensor terminal 206, a wireless base station 204, and a wireless terminal 205. The event detection system 200 according to the first embodiment of the present invention will be described below with reference to the drawings.
[0017] FIG. 1 is a diagram showing a configuration of an event detection system 200. As shown in FIG.
[0018] The event detection system 200 is composed of a server 201, a wireless base station 204, a wireless terminal 205, and a sensor terminal 206. The server 201 and the wireless base station 204 are connected via an external network 202 and a network hub 203. The network hub 203 and the wireless base station 204 are connected by wire, such as Ethernet (207). The wireless base station 204 and the wireless terminal 205 communicate with wireless radio waves 208. The sensor terminal 206 detects an event by receiving and analyzing the wireless radio waves 208.
[0019] The server 201 is a control device that controls the event detection system 200 (e.g., the timing of the sensing process). The wireless terminal 205 is a device that transmits or receives user data between the wireless base station 204 and the sensor terminal 206. The sensor terminal 206 is a device that receives a sensor packet transmitted by the sensor terminal 206 itself or the wireless base station 204 to detect a status event. The sensor terminal 206 may transmit or receive user data.
[0020] Fig. 2 is a diagram showing the configuration of the sensor node 206 according to the first embodiment. In Fig. 2, a non-sensor packet transmitted by the sensor node 206 is received and analyzed.
[0021] The sensor terminal 206 includes a wireless device 1 (102) and a wireless device 2 (103). The wireless device 1 (102) wirelessly communicates with the wireless base station 100 and transmits and receives a sensor packet 104 and a non-sensor packet 105. The sensor packet 104 is a packet used for communication control, and the contents of which change depending on the surrounding environment. For example, a CBR packet including MIMO information in WiFi can be used as the sensor packet 104. The non-sensor packet 105 is a packet other than the sensor packet 104, and is, for example, a packet for transmitting the result of sensing by the sensor terminal 206 to a higher-level device or a packet for transferring user data. The wireless device 2 (103) receives the sensor packet 104 and the non-sensor packet 105 transmitted and received by the wireless device 1 (102), and stores them as packet data 106. After that, the sensor packet extraction unit 107 of the sensor terminal 206 extracts the sensor packet 104 from the stored packet data 106, and stores the sensor packet data 108 in a memory. The sensing processing unit 109 executes sensing processing on the sensor packet data 108 stored in the memory. The result output unit 111 outputs the sensing result 110 derived by the sensing processing. The sensing result is the state of the person or object to be sensed. Assuming that the above-described configuration is realized in an environment such as an office or a factory, the wireless base station 204 does not need to be introduced together with the sensor terminal 206, and if there is an existing wireless base station 204, the sensing function can be realized by introducing one sensor terminal 206. With this configuration, even in an environment where the introduction of a new wireless base station 204 is restricted, the sensing function can be realized by newly introducing the sensor terminal 206.
[0022] The sensing process shown in FIG. 2 may be executed by the server 201 instead of the sensor terminal 206 .
[0023] Fig. 3 is a diagram showing the configuration of the wireless base station 204 according to the first embodiment. In Fig. 3, the wireless base station 204 receives and analyzes a non-sensor packet 305 transmitted by a wireless terminal 205.
[0024] The wireless base station 204 includes a wireless device 1 (302) and a wireless device 2 (303). The wireless device 1 (302) wirelessly communicates with the wireless terminal 205, and transmits and receives a sensor packet 304 and a non-sensor packet 305. The sensor packet 304 is a packet used for communication control, and the contents of which change depending on the surrounding environment. For example, in WiFi, a CBRpkt including MIMO information can be used as the sensor packet 304. The non-sensor packet 305 is a packet other than the sensor packet 304, and is, for example, a packet for transmitting the result of sensing by the sensor terminal 206 to a higher-level device. In addition, the wireless device 2 (303) receives the sensor packet 304 and the non-sensor packet 305 transmitted and received by the wireless device 1 (302), and stores them as packet data 306. Thereafter, a sensor packet extraction unit 307 of the wireless base station 204 extracts the non-sensor packet 305 from the stored packet data 306, and stores the sensor packet data 308 in a memory. The sensing processing unit 309 executes sensing processing on the sensor packet data 308 stored in the memory. The result output unit 311 outputs the sensing result 310 derived by the sensing processing. The configuration of the wireless base station 204 described above is obtained by applying the configuration of the sensor terminal 206 shown in Fig. 2 to the wireless base station 204. When it is possible to install a new wireless base station in an environment such as an office or factory, a sensing function can be realized by introducing the wireless base station 204 having the configuration shown in Fig. 3.
[0025] Fig. 4 is a diagram showing the configuration of a sensor terminal 206 according to the first embodiment. In Fig. 4, the sensor terminal 206 receives and analyzes a non-sensor packet transmitted by a wireless terminal 205. In general, the wireless terminals 205 are installed widely within the communication range of the wireless base station 204, and therefore have the feature that the state of the wide area in which the wireless terminals 205 are installed can be known.
[0026] The sensor terminal 206 includes a wireless device 403. The wireless terminal 205 wirelessly communicates with the wireless base station 204 and transmits and receives a sensor packet 404 and a non-sensor packet 405. The sensor packet 404 is a packet used for communication control, and the contents of which change depending on the surrounding environment. For example, a CBRpkt including MIMO information in WiFi can be used as the sensor packet 404. The non-sensor packet 405 is a packet other than the sensor packet 404, and is, for example, a packet for transmitting the result of sensing by the sensor terminal 206 to a higher-level device. The wireless device 403 receives the sensor packet 404 and the non-sensor packet 405 transmitted and received by the wireless terminal 205, and stores them as packet data 406. Thereafter, the sensor packet extraction unit 407 of the sensor terminal 206 extracts the sensor packet 404 from the stored packet data 406 (407), and stores the sensor packet data 408 in a memory. The sensing processing unit 409 executes a sensing process on the sensor packet data 408 stored in the memory. The result output unit 411 outputs the sensing result 410 derived in the sensing process. The configuration of the sensor terminal 206 described above is a configuration in which the wireless communication function and the sensor function of the sensor terminal 206 shown in Fig. 2 are separated, and the sensor function is removed. In this configuration, it is desirable that the sensor terminal 206 is installed in a fixed location. If the sensor terminal 206 is a mobile device, the sensor information included in the sensor packet changes as the sensor terminal 206 moves, and it becomes unclear whether the change in the sensor packet is due to the movement of the sensor terminal 206 or a change in the sensing target, making it difficult to realize the sensing function.
[0027] In addition, the configurations of the sensor terminal 206, the wireless base station 204, and the sensor terminal 206 shown in Figures 2, 3, and 4 may be such that sensing processing is performed by multiple devices, but it is not necessary for sensing processing to be performed by all devices as long as the sensing processing is performed by at least one device.
[0028] FIG. 5 is a diagram showing the sensing process executed by the event detection system 200 according to the first embodiment.
[0029] The sensing process is broadly divided into a learning mode 500 and an operation mode 501. In the learning mode 500, a set of label information 502 and learning sensor packet data 503 is input to a sensing machine learning model creation unit 504 as teacher data. The label information 502 is information that indicates the state of a person or object to be sensed. The label information 502 and the sensor packet data transmitted and received in that state are used as learning sensor packet data 503. The sensing machine learning model creation unit 504 learns these two and creates a machine learning model 505 to be used for sensing.
[0030] In the operation mode 501, the machine learning model 505 created in the learning mode 500 is applied to the sensor packet to estimate the classification result. In the operation mode 501, the operation sensor packet data 506 received during operation is input to the sensing machine learning classification processing unit 507. The state of the operation sensor packet data 506 at the time of transmission / reception is unknown, and the sensing machine learning classification processing unit 507 estimates the state of the person or object to be sensed at the time of transmission / reception of the operation sensor packet data 506 using the trained machine learning model 505. The obtained sensing result 508 is the detection result of the state event. The sensing process is executed by the process described above.
[0031] FIG. 6 is a diagram showing an example of a monitor screen 600 on which the event detection results output by the result output units 111, 311, and 411 are displayed.
[0032] The result output unit 111, 311, 411 outputs display data for displaying a sensing result 601, a sensing history 602, and sensing packet data details 603 on a monitor screen 600. The sensing result 601 is the state of a person or object to be sensed. The sensing history 602 is the past state of a person or object to be sensed, and is a set of data of a timestamp indicating the detection date and time and the state of the sensing object. The sensing packet data details 603 is a characteristic value (e.g., signal-to-noise ratio, amplitude, phase) measured in a received non-sensor packet, and may be displayed so that the amplitude value and phase value for each subcarrier can be known. Also, the time change of the signal-to-noise ratio may be displayed. These pieces of information may be displayed in a state where a display device is connected to the sensor terminal 206. Also, information on the sensing result may be transmitted from the sensor terminal 206 to the server 201 and displayed on a display device connected to the server 201.
[0033] As described above, according to the first embodiment of the present invention, the sensor terminal 206 or the wireless base station 204 receives a sensor packet transmitted by the sensor terminal 206 or the wireless terminal 205, and can sense the state of a person or object that is a sensing target. In particular, when the sensor terminal 206 receives a sensor packet, the wireless base station 204 can use an existing device, and the sensing function can be realized by installing at least one sensor terminal 206 at the site. Furthermore, the sensing function can be realized without building a new wireless network for sensing, and with minimal impact on the existing system.
[0034] <Second embodiment> Next, a second embodiment of the present invention will be described. The event detection system of the second embodiment is composed of a server 201, a sensor terminal 206, a wireless base station 204, and a wireless terminal 205, similar to the first embodiment (FIG. 1). In the second embodiment, one sensor terminal 206 and multiple wireless terminals are provided. In the second embodiment, differences from the first embodiment described above will be mainly described, and the same configurations and processes as in the first embodiment will be assigned the same reference numerals, and their description will be omitted.
[0035] [System Configuration] FIG. 7 is a diagram showing the configuration of an event detection system according to the second embodiment.
[0036] In the second embodiment, a sensor terminal 1 (206), a wireless terminal 2 (205), and a wireless terminal N (205) are connected to a wireless base station 700. A plurality of wireless terminals 205 are connected to the wireless base station 204, and the number of connected terminals is not limited. The sensor terminal 206 has the same configuration as that shown in FIG. 2, and includes a wireless device 1 (102) and a wireless device 2 (103). The wireless device 1 (102) transmits a sensor packet 1 (708), and the wireless device 2 (705) receives the sensor packet 1 (708). The wireless terminal 2 (205) includes a wireless device 706, which transmits a sensor packet 2 (709). The sensor packet 2 (709) is received by the sensor terminal 206. Similarly, the wireless terminal N (205) includes a wireless device 707, which transmits a sensor packet N (710). The sensor packet N (710) is received by the sensor terminal 206. In the second embodiment, the sensor terminal 206 receives all of the sensor packets 1 to N.
[0037] FIG. 8 is a diagram showing the sensing process executed by the event detection system 200 according to the second embodiment.
[0038] Although the sensing process is described as being executed by the server 201, a device other than the server 201 (for example, a sensor terminal) may execute the sensing process. The sensing process is broadly divided into a learning mode and an operation mode. In the learning mode, a set of label information 801 and learning sensor 1 packet data 802, a set of label information 801 and learning sensor 2 packet data 803, and a set of label information 801 and learning sensor N packet data 804 are input to a sensing machine learning model creation unit 805 as teacher data. The label information 801 is information that represents the state of a person or object to be sensed. The label information 801 and the sensor packet data 802 to 804 transmitted and received in that state are used as learning sensor packet data. The sensing machine learning model creation unit 805 learns these data and creates machine learning models 806 to 808 to be used for sensing. At this time, a learning model 1 (806) is created for the learning sensor 1 packet data 802, a learning model 2 (807) is created for the learning sensor 2 packet data 803, and a learning model N (808) is created for the learning sensor N packet data 804. If the label information 801 including the identification information of the sensor terminal 206 is used as teacher data, a learning model that can derive the sensing result for each sensor terminal 206 can be created without creating a learning model for each wireless terminal 205 and sensor terminal 206.
[0039] In the operation mode, the sensing processing unit 109 applies a plurality of machine learning models 806 to 808 created in the learning mode to each sensor packet to estimate a classification result. For example, when the sensor 1 packet data 708 is given as an input, the classifier 1 (811) uses the learning model 1 (806) to estimate the state of the sensing target (person, object) at the time of transmitting and receiving the sensor 1 packet data 708, and outputs the sensing result 1 (812). The learning model 1 (806) is a machine learning model created using the learning sensor 1 packet data 802. Similarly, when the sensor 2 packet data 709 is given as an input, the classifier 2 (815) uses the learning model 2 (807) to estimate the state of the sensing target (person, object) at the time of transmitting and receiving the sensor 2 packet data 709, and outputs the sensing result 2 (816). The learning model 2 (807) is a machine learning model created using the learning sensor 2 packet data 803. Furthermore, when the sensor N packet data 710 is given as an input, the classifier 3 (819) uses the learning model N (808) to estimate the state of the sensing target (person, object) at the time of transmitting and receiving the sensor N packet data 710, and outputs the sensing result N (820). The learning model N (808) is a machine learning model created using the learning sensor N packet data 804.
[0040] The overall judgment unit 821 makes an overall judgment on the multiple sensing results 812, 816, 820 obtained in this manner, and obtains one sensing result 822. For example, the overall judgment can be made by majority judgment or judgment taking into account a weighting coefficient based on the accuracy rate calculated using the correct answer data.
[0041] In the second embodiment, in a configuration in which multiple wireless terminals and sensor terminals are installed, a learning model is created for each wireless terminal and sensor terminal, and sensor packet data is classified. The contents of the sensor packets acquired by the sensor terminals differ depending on the installation location and the installation environment of the wireless propagation environment. For this reason, if one learning model is created from sensor packets obtained from multiple sensor terminals, there are cases in which the state cannot be classified correctly. By creating a learning model for each sensor, a learning model that can correctly classify the state can be created, and sensing can be performed with high accuracy.
[0042] Fig. 9 is a flowchart of a process for controlling the transmission interval of a sensing packet in the second embodiment. The process shown in Fig. 9 can also be applied to the first embodiment. The process shown in Fig. 9 may be performed by any of a server, a wireless base station, and a communication terminal.
[0043] After starting the process (900), the transmission interval of the sensor packets is set (901). After that, if there is an instruction to start sensing (902), a sensor request packet is transmitted (903).
[0044] After the sensor request packet is transmitted, when the set transmission interval has elapsed (904), the next sensor request packet is transmitted (903).
[0045] If an instruction to stop sensing is received (905), the transmission of the sensor request packet is stopped.
[0046] By repeating this process, if the transmission interval is set to, for example, one second, a sensor request packet can be transmitted at one second intervals, and a sensor packet can be transmitted at one second intervals.
[0047] FIG. 10 is a sequence diagram of communication in which the transmission interval of the sensing packets shown in FIG. 9 is controlled.
[0048] In the sequence shown in FIG. 10, communication is performed between the server 201, wireless base station 204, wireless terminal 1 (205), wireless terminal N (205), and sensor terminal 1 (206), and sensor packets are transmitted and received. After setting a transmission interval, the server 201 transmits a sensor request 1 to the wireless terminal 1 (205). The sensor request 1 is transmitted to the wireless terminal 1 (205) via the wireless base station 204. The wireless terminal 1 (205) receives the sensor request 1 and returns a sensor packet 1. The sensor packet 1 is received by the sensor terminal 1 (206). The sensor terminal 1 (206) transmits the received sensor packet 1 to the server 201 via the wireless base station 204. The server 201 executes a sensing process on the received sensor packet 1.
[0049] Similarly, the server 201 transmits a sensor request 2 to the wireless terminal 2 (205). The sensor request 2 is transmitted to the wireless terminal 2 (205) via the wireless base station 204. The wireless terminal 2 (205) receives the sensor request 2 and returns a sensor packet 2. The sensor packet 2 is received by the sensor terminal 1 (206). The sensor terminal 1 (206) transmits the received sensor packet 2 to the server 201 via the wireless base station 204. The server 201 executes a sensing process on the received sensor packet 2.
[0050] Then, after the set transmission interval has elapsed, the server 201 transmits the sensor request 1.
[0051] By repeating the above process, the sensing interval can be controlled at any timing. Furthermore, the sensing process does not have to be repeated at a fixed cycle, and sensing can be done just once at any timing. The sensing frequency described here is set according to the purpose of how responsively a change in physical state should be detected when the state changes. For applications that require detection in a short time (such as intrusion detection), it is best to set the transmission interval to a short value.
[0052] Also, as one implementation example of the sensor request and sensor packet described above, the sensor request may be Ping data. When the sensor request is Ping data, the wireless base station transmits a Compressed Beamforming Report Request (CBRP) packet defined in the WiFi5 or higher standard immediately before transmitting the Ping packet. When the wireless terminal 205 receives the CBRR packet, it transmits a Compressed Beamforming Report (CBR) packet. This CBR packet is used as a sensor packet. In this way, the sensor request and sensor packet may be realized by a Ping packet in the WiFi5 protocol. Also, the same function can be realized by implementing a protocol corresponding to the sensor request and sensor packet.
[0053] <Third embodiment> Next, a third embodiment of the present invention will be described. The event detection system of the third embodiment is composed of a server 201, a sensor terminal 206, and a wireless base station 204, similar to the first embodiment (FIG. 1). In the third embodiment, a plurality of sensor terminals 206 are provided, but a wireless terminal 205 is not provided. In the third embodiment, differences from the first embodiment described above will be mainly described, and the same configurations and processes as those of the first or second embodiment will be assigned the same reference numerals, and their description will be omitted.
[0054] [System Configuration] FIG. 11 is a diagram showing the configuration of an event detection system according to the third embodiment.
[0055] In the third embodiment, a sensor terminal 1 (206), a sensor terminal 2 (206), and a sensor terminal N (206) are connected to the wireless base station 1100. A plurality of sensor terminals 206 are connected to the wireless base station 204, and the number of connected terminals is not limited. The sensor terminals 1 to N (206) have the same configuration as that shown in FIG. 2, and include a wireless device 1 (102) and a wireless device 2 (103). The wireless device 1 (102) of the sensor terminal 1 (206) transmits a sensor packet 1 (1110), and the wireless device 2 (103) receives the sensor packet 1 (1110). At this time, the sensor terminal 2 (206) and the sensor terminal N (206) receive the sensor packet 1 (1110). Similarly, the sensor terminal 2 (206) includes a wireless device 1 (102) and a wireless device 2 (103). The wireless device 1 (102) of the sensor terminal 2 (206) transmits a sensor packet 2 (1111), and the wireless device 2 (103) receives the sensor packet 2 (1111). At this time, the sensor terminal 1 (206) and the sensor terminal N (206) receive the sensor packet 2 (1111). In addition, the sensor terminal N (206) has a wireless device 1 (102) and a wireless device 2 (103). The wireless device 1 (102) of the sensor terminal N (206) transmits a sensor packet 3 (1112), and the wireless device 2 (103) receives the sensor packet 3 (1112). At this time, the sensor terminal 1 (206) and the sensor terminal 2 (206) receive the sensor packet 3 (1112). In the third embodiment, each sensor terminal receives the sensor packet transmitted by each sensor terminal.
[0056] FIG. 12 is a diagram showing a process of complementing missing data in the third embodiment.
[0057] In the following description, the complementation process is described as being executed by the server 201, but the complementation process may be executed by a device other than the server 201 (for example, a sensor terminal). In the third embodiment, in order to execute the complementation process, a sensor 1 table 1201 and a sensor 2 table 1202 are provided. The sensor 1 table 1201 records a sensor packet acquired by the sensor terminal 1, and the sensor 2 table 1202 records a sensor packet acquired by the sensor terminal 2. As described in FIG. 11, the sensor 1 table 1201 records the sensor packet 1 and the sensor packet 2 received by the sensor terminal 1 (206) and the time information T1 to T3 at which each sensor packet was acquired. Similarly, the sensor 2 table 1202 records the sensor packet 1 and the sensor packet 2 received by the sensor terminal 2 (206) and the time information T1 to T3 at which each sensor packet was acquired. Note that other sensor packets (for example, a sensor packet N) may be recorded in the sensor 1 tables 1201 and 1202.
[0058] Since a sensor packet is acquired via wireless communication, it may not be possible to acquire the sensor packet for some reason. For example, a sensor packet may be acquired at time T1, but not at time T2 for some reason, and the sensor packet may be acquired at times T3 and T4. In this case, if the sensor packet acquired at time T2 is recorded in the sensor 1 table, it is advisable to complement the data in the sensor 2 table 1202 using the sensor packet at time T2 recorded in the sensor 1 table 1201 and execute the sensing process.
[0059] An example of the processing procedure of the third embodiment is shown in Fig. 12. First, data in the sensor 1 table 1201 and data in the sensor 2 table 1202 are compared (1203), and one of the sensor packets having the same reception timing is deleted (1204). Since the sensor packets having different reception timing are sensor packets that have been lost for some reason, learning sensor packet data 503 in which the missing sensor packets are complemented is generated using the sensor packets recorded in the other sensor table, and a learning model is created. In the operation mode, the recorded sensor packets are used to generate operation sensor packet data 506 in which the missing sensor packets are complemented, and a sensing machine learning classification process is executed to classify the data using a classifier (1205).
[0060] As described above, with the configuration and processing of the third embodiment, even if a sensor packet is lost, the lost sensor packet can be complemented and the sensing process can be continued without interruption. Also, sensing can be performed with high accuracy.
[0061] <Fourth embodiment> Next, a fourth embodiment of the present invention will be described. As in the first embodiment (FIG. 1), the event detection system of the fourth embodiment is composed of a server 201, a wireless terminal 205, a sensor terminal 206, and a wireless base station 204. In the fourth embodiment, differences from the first embodiment described above will be mainly described, and the same configurations and processes as those of the first to third embodiments will be given the same reference numerals, and descriptions thereof will be omitted.
[0062] [System Configuration] The wireless terminals in the fourth embodiment include unmanaged wireless terminals 1302 and 1303 that are not managed by the event detection system, in addition to the wireless terminal 205. As described above, the wireless terminal 205 transmits a sensor packet 404 in response to a sensor request from the wireless base station 204, but the unmanaged wireless terminal is a wireless terminal whose operation mode in the installation environment of the event detection system is unknown and which does not respond to the sensor request from the wireless base station 204. For example, when a new sensor terminal is introduced, it is a wireless terminal that has already been installed and operated in that environment.
[0063] FIG. 13 is a diagram showing the configuration of an event detection system according to the fourth embodiment.
[0064] In the fourth embodiment, a sensor terminal 206, an unmanaged sensor mobile terminal 1302 as an unmanaged wireless terminal, and an unmanaged sensor fixed terminal 1303 are connected to a wireless base station 204. The sensor terminal 206 has the same configuration as that shown in Fig. 2, and includes a wireless device 1 (102) and a wireless device 2 (103). The wireless device 1 (102) transmits a sensor packet 1 (1306), and the wireless device 2 (103) receives the sensor packet 1 (1306).
[0065] The unmanaged mobile sensor terminal 1302 is a moving wireless terminal that is not operated as a sensor terminal under the event detection system of this embodiment. The unmanaged mobile sensor terminal 1302 is connected to the wireless base station 204, and transmits an unmanaged sensor packet M (1308). The unmanaged fixed sensor terminal 1303 is a fixedly installed wireless terminal that is not operated as a sensor terminal under the event detection system of this embodiment. The unmanaged fixed sensor terminal 1303 is connected to the wireless base station 204, and transmits an unmanaged sensor packet F (1307).
[0066] The sensor terminal 206 receives the unmanaged sensor packet M (1308) and the unmanaged sensor packet F (1307). In the fourth embodiment, the sensor terminal 206 receives the sensor packet 404 transmitted by the wireless terminal 205, and further receives the unmanaged sensor packets transmitted by the unmanaged wireless terminals 1302 and 1303 operating in the installation environment. The unmanaged sensor packets of the fourth embodiment are packets transmitted by terminals other than the wireless terminal 205, and the contents of the unmanaged sensor packets 1307 and 1308 and the sensor packet 404 are the same and are used for sensing.
[0067] FIG. 14 is a diagram showing a sensing process using unmanaged sensor packets transmitted by unmanaged wireless terminals 1302 and 1303 in the fourth embodiment.
[0068] Although the sensing process using the unmanaged sensor packet will be described as being executed by the server 201, the sensing process may be executed by a device other than the server 201 (for example, a sensor terminal). In the embodiment shown in Fig. 14, the server 201 has a sensor table 1401. The sensor table 1401 records the sensor packets acquired by the sensor terminal 206. For example, the sensor table 1401 records the sensor packet 404 transmitted by the wireless terminal 205 and the unmanaged sensor packets transmitted by the unmanaged wireless terminals 1302 and 1303, which were received at reception times T1 to T3.
[0069] Next, in the learning mode, the server 201 inputs a pair of the label information 1402 and the learning sensor packet data 1403, and a pair of the label information 1402 and the learning unmanaged sensor packet data 1404 to the sensing machine learning model creation unit 1405 as teacher data. The label information 1402 is information that indicates the state of a person or object to be sensed. The label information 1402 and the sensor packet data 1403 to 1404 transmitted and received in that state are used as the learning sensor packet data. The sensing machine learning model creation unit 1405 learns these data and creates machine learning models 1406 to 1407 to be used for sensing. At this time, a learning model 1 (1406) is created for the learning sensor packet data 1403, and a learning model 2 (1407) is created for the learning unmanaged sensor packet data 1404.
[0070] In the operation mode, the sensing processing unit 109 applies a plurality of machine learning models 1406-1407 created in the learning mode to each sensor packet to estimate a classification result. For example, when the sensor packet data 408 is given as an input, the classifier 1 (1410) uses the learning model 1 (1406) to estimate the state of the sensing target (person, object) at the time of transmitting and receiving the sensor packet data 408, and outputs the sensing result 1 (1411). The learning model 1 (1406) is a machine learning model created using the learning sensor packet data 1403. Similarly, when the unmanaged sensor packet data 1307, 1308 are given as an input, the classifier 2 (1414) uses the learning model 2 (1407) to estimate the state of the sensing target (person, object) at the time of transmitting and receiving the unmanaged sensor packet data 1307, 1308, and outputs the sensing result 2 (1415). Learning model 2 (1407) is a machine learning model created using sensor 2 packet data.
[0071] A comprehensive judgment unit 1416 comprehensively judges the multiple sensing results 1411, 1415 obtained in this way to obtain one sensing result 1417. For example, the comprehensive judgment can be made by majority decision or judgment taking into account a weighting coefficient based on the accuracy rate calculated using the correct answer data.
[0072] Fig. 15 is a diagram showing sensing processing in the fourth embodiment, which uses unmanaged sensor packets transmitted by unmanaged wireless terminals 1302 and 1303. The difference from Fig. 14 described above is that the unmanaged wireless terminals 1302 and 1303 are treated as one type in the processing shown in Fig. 14, but the unmanaged wireless terminals 1302 and 1303 are distinguished and treated as multiple types in the processing shown in Fig. 15. For example, if there are two types of unmanaged wireless terminals, an unmanaged sensor mobile terminal 1302 and an unmanaged sensor fixed terminal 1303, the unmanaged sensor packets of each wireless terminal are identified, and sensing processing is performed using the unmanaged sensor packets of the fixed terminal.
[0073] Although the sensing process using the unmanaged sensor packet of the fixed terminal will be described as being executed by the server 201, the sensing process may be executed by a device other than the server 201 (for example, a sensor terminal). In the embodiment shown in FIG. 15, the server 201 has a sensor table 1501. The sensor table 1501 records the sensor packets acquired by the sensor terminal 206. For example, the sensor table 1501 records the sensor packet 404 transmitted by the wireless terminal 205, the unmanaged sensor packet M (1308) transmitted by the unmanaged sensor mobile terminal 1302, and the unmanaged sensor packet F (1307) transmitted by the unmanaged sensor fixed terminal 1303, which are received at reception times T1 to T3.
[0074] Next, in the learning mode, the server 201 inputs a set of the label information 1502 and the learning sensor packet data 1503, a set of the label information 1502 and the learning unmanaged sensor packet M data (1504), and a set of the label information 1502 and the learning unmanaged sensor packet F data (1505) to the sensing machine learning model creation unit 1506 as teacher data. The label information 1502 is information that represents the state of a person or object to be sensed. The label information 1502 and the sensor packet data 1503 to 1505 transmitted and received in that state are used as learning sensor packet data. The sensing machine learning model creation unit 1506 learns these data and creates machine learning models 1507 to 1509 to be used for sensing. At this time, a learning model 1 (1507) is created for the learning sensor packet data 1503, a learning model M (1508) is created for the learning unmanaged sensor packet M data 1504, and a learning model F (1509) is created for the learning unmanaged sensor packet F data 1505.
[0075] Next, it is determined whether the created learning model M (1508) and learning model F (1509) are based on a sensor packet of a mobile terminal or a sensor packet of a fixed terminal. In the form shown in Fig. 14, a mobile terminal and a fixed terminal are explicitly illustrated, but in an actual environment, it may be unclear whether an unmanaged wireless terminal is operated mobilely or fixedly. The processing flow described below is a procedure for estimating whether an unmanaged wireless terminal is operated mobile or fixedly from the characteristics of the unmanaged sensor packet and the learning model, even when it is unclear whether the unmanaged wireless terminal is operated mobile or fixedly, and adopting the unmanaged sensor packet and learning model estimated to be operated fixedly.
[0076] First, the score of the created learning model is calculated (1510). For example, correct answer data, the answers of which are known in advance, may be classified by a classifier, and the correct answer rate may be calculated to calculate the score.
[0077] Then, if the terminal is judged to be a mobile terminal because the score of the learning model is lower than a predetermined threshold, or if it is judged to be a mobile terminal based on the attributes of the terminal (1511), it is judged not to use the learning model for sensing in the operation mode (1512).If the terminal is judged to be a fixed terminal because the score of the learning model is higher than a predetermined threshold, or if it is judged to be a fixed terminal based on the attributes of the terminal (1511), the learning model is adopted (1513).
[0078] Also, characteristic values such as the phase, amplitude, and signal-to-noise ratio of the packet may be acquired, and the attributes of the unmanaged wireless terminal may be determined based on these characteristic values. For example, if the amount of fluctuation in the characteristic value is greater than a predetermined threshold, it may be determined that the unmanaged wireless terminal is a mobile terminal. Also, since the amplitude has a value for each carrier, it is possible to accurately determine whether the unmanaged wireless terminal is a fixed terminal or a mobile terminal.
[0079] Since the movement of the mobile terminal causes the unmanaged sensor packet data to contain large fluctuations, the classification accuracy of the learning model created based on the score sensor packet data is reduced. Therefore, the learning model created based on the unmanaged sensor packet data of the fixed terminal is expected to have a high score. The sensing process is performed using the learning model selected in this way. The sensing process for estimating the classification result in the operation mode is the same as the process described in Figure 8 and Figure 14.
[0080] In the fourth embodiment, in a configuration in which unmanaged wireless terminals 1302 and 1303 are installed, a learning model is created for each sensor terminal, and sensor packet data is classified. Therefore, even if the number of installed sensor terminals is small, highly accurate sensing is possible by using sensor packets transmitted by existing wireless terminals (unmanaged wireless terminals).
[0081] The present invention is not limited to the above-described embodiments, and includes various modified examples and equivalent configurations within the spirit of the appended claims. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to having all of the configurations described. Furthermore, a part of the configuration of one embodiment may be replaced with a configuration of another embodiment. Furthermore, a configuration of another embodiment may be added to a configuration of one embodiment. Furthermore, a part of the configuration of each embodiment may be added, deleted, or replaced with another configuration.
[0082] In addition, each of the above-mentioned configurations, functions, processing units, processing means, etc. may be realized in hardware, for example by designing some or all of them as an integrated circuit, or may be realized in software by a processor interpreting and executing a program that realizes each function.
[0083] Information such as programs, tables, and files that realize each function can be stored in a storage device such as a memory, a hard disk, or an SSD (Solid State Drive), or in a recording medium such as an IC card, an SD card, or a DVD.
[0084] In addition, the control lines and information lines shown are those considered necessary for the explanation, and do not necessarily show all the control lines and information lines necessary for implementation. In reality, it can be considered that almost all components are connected to each other. [Explanation of symbols]
[0085] 100 Radio base station 101 Sensor terminal 102, 103, 403 Wireless Devices 104, 304, 404 sensor packets 105, 305, 405 Non-sensor packets 106, 306, 406 Packet Data 107, 307, 407 Sensor packet extraction unit 108, 308, 408 Sensor packet data 109, 309, 409 Sensing processing unit 110, 310, 410 Sensing results 111, 311, 411 Result output section 200 Event Detection System 201 Server 202 External Network 203 Network Hub 204 Wireless base station 205 Wireless Terminal 206 Sensor Terminal 208 Radio Waves 500 Learning Mode 501 Operational Mode 502 Label Information 503 Sensor packet data for learning 504 Sensing Machine Learning Model Creation Department 505 Machine Learning Models 506 Operational Sensor Packet Data 507 Sensing Machine Learning Classification Processing Unit 508 Sensing Results 600 monitor screens 601 Sensing Results 602 Sensing History 603 Sensing packet data details 700 Wireless base stations 706 Wireless Devices 707 Wireless Devices 708 Sensor 1 packet data 709 Sensor 2 packet data 710 Sensor N packet data 801 Label Information 802 Learning sensor 1 packet data 803 Learning sensor 2 packet data 804 Learning sensor N packet data 805 Sensing Machine Learning Model Creation Department 806 Machine Learning Model 1 807 Machine Learning Model 2 808 Machine Learning Model N 812 Sensing result 1 816 Sensing result 2 820 Sensing Result N 821 General Judging Department 822 Sensing Results 1100 Wireless base station 1201 Sensor 1 Table 1202 Sensor 2 Table 1302 Unmanaged Sensor Mobile Terminal 1303 Unmanaged sensor fixed terminal 1307, 1308 Unmanaged sensor packet data 1401 Sensor Table 1402 Label Information 1403 Learning Sensor Packet Data 1404 Unmanaged Sensor Packet Data for Learning 1405 Sensing Machine Learning Model Creation Department 1406 Machine Learning Model 1 1407 Machine Learning Model 2 1411 Sensing result 1 1415 Sensing result 2 1416 General Judging Department 1417 Sensing Results 1501 Sensor Table 1502 Label Information 1503 Sensor packet data for learning 1504 Unmanaged Sensor Packet M Data for Learning 1505 Unmanaged Sensor Packet F Data for Learning 1506 Sensing Machine Learning Model Creation Department 1507 Machine Learning Model 1 1508 Machine Learning Model M 1509 Machine Learning Model F
Claims
1. 1. A detection system for detecting a status event, comprising: a first wireless device for transmitting wireless packets; a second wireless device that receives the wireless packet; a wireless base station capable of wirelessly communicating with at least the first wireless device; a sensor packet extraction unit that extracts a sensor packet transmitted by the first wireless device from a packet received by the second wireless device; a sensing processing unit that executes a sensing process to detect a state event indicating information of the real world by using the extracted sensor packet, The detection system is characterized in that the sensor packet is a packet used for communication control, and the content of the packet changes depending on the surrounding environment.
2. 2. The detection system of claim 1, a wireless terminal having the first wireless device; a sensor terminal having the second wireless device, The sensor terminal receives a sensor packet transmitted from the wireless terminal, and executes a sensing process using the received sensor packet.
3. 3. The detection system of claim 2, The system includes one sensor terminal and a plurality of wireless terminals, the sensor terminal has the first wireless device and the second wireless device, the wireless terminal has the first wireless device; the first wireless device of the sensor terminal transmits a first sensor packet; the first wireless device of the wireless terminal transmits a second sensor packet; The detection system, wherein the second wireless device of the sensor terminal receives the first sensor packet and the second sensor packet.
4. 4. The detection system of claim 3, A model creation unit that creates a learning model used in the sensing process, The model creation unit is creating a first learning model to be used in a sensing process from the label information and the first sensor packet; creating a second learning model to be used in the sensing process for each of the wireless terminals from the label information and the second sensor packet; The sensing processing unit is In the sensing process using the first sensor packet, a sensing result is derived using the first learning model; In the sensing process using the second sensor packet, a sensing result is derived using the second learning model; A detection system comprising: a detection unit for detecting a status event based on the derived plurality of sensing results.
5. 5. The detection system of claim 4, The sensing processing unit performs majority decision on the derived multiple sensing results to detect one state event.
6. 2. The detection system of claim 1, a sensor terminal having the first wireless device and the second wireless device; The sensor terminal receives a sensor packet transmitted from the first wireless device by the second wireless device, and performs sensing processing using the received sensor packet.
7. 2. The detection system of claim 1, a wireless terminal having the first wireless device; a radio base station having the second radio device, A detection system characterized in that the wireless base station receives a sensor packet transmitted from the first wireless device at the second wireless device, and performs sensing processing using the received sensor packet.
8. 2. The detection system of claim 1, a result output unit that outputs display data for displaying the detected status event; The detection system is characterized in that the result output unit outputs display data for displaying at least one of the latest sensing result including the detected state event, the past sensing result, and information regarding the signal-to-noise ratio, the amplitude value, and the phase value included in the sensor packet.
9. 2. The detection system of claim 1, A control device for controlling the sensing process is provided, The control device includes: Transmitting a sensor request to the first wireless device according to a set timing; the first wireless device transmits the sensor packet in accordance with the sensor request; the second wireless device receives the sensor packet; The sensing processing unit executes the sensing process by using a sensor packet received by the second wireless device.
10. 2. The detection system of claim 1, a plurality of sensor terminals each having the first wireless device and the second wireless device; the first wireless device of the plurality of sensor terminals transmits the sensor packet; A detection system, characterized in that the second wireless device of the plurality of sensor terminals receives a sensor packet transmitted by the first wireless device of the plurality of sensor terminals.
11. 11. The detection system of claim 10, A control device for controlling the sensing process is provided, The control device includes: determining whether a sensor packet received by the plurality of sensor terminals is duplicated or missing according to a reception timing of the sensor packet; Delete duplicate sensor packets, Completing missing sensor packets, A learning model is created using the sensor packets after the duplicate deletion and the missing data completion, and the learning model is used for the sensing process. The sensing processing unit derives a sensing result using sensor packets after duplicate deletion and missing part completion.
12. 2. The detection system of claim 1, a sensor terminal having the first wireless device and the second wireless device, the first wireless device transmits a first sensor packet; the second wireless device receives the first sensor packet and an unmanaged sensor packet transmitted by an unmanaged wireless terminal that is not managed by the detection system; The sensing processing unit executes the sensing process using a first sensor packet and the unmanaged sensor packet received by the second wireless device.
13. 13. The detection system of claim 12, A model creation unit that creates a learning model to be used in the sensing process, The model creation unit is creating a first learning model to be used in a sensing process from the label information and the first sensor packet; creating a second learning model to be used in a sensing process from the label information and the unmanaged sensor packet; The sensing processing unit is In the sensing process using the first sensor packet, a sensing result is derived using the first learning model; In the sensing process using the unmanaged sensor packet, a sensing result is derived using the second learning model; A detection system comprising: a detection unit for detecting a status event based on the derived plurality of sensing results.
14. 14. The detection system of claim 13, The model creation unit is Using the correct answer data, a correct answer rate indicating the superiority or inferiority of the created second learning model is calculated; A detection system characterized by determining to reject a second learning model whose calculated accuracy rate is lower than a predetermined threshold.
15. 14. The detection system of claim 13, the model creation unit acquires a characteristic value of the unmanaged sensor packet, and determines to reject a second learning model that has been created using an unmanaged sensor packet transmitted by an unmanaged wireless terminal in which a fluctuation amount of the acquired characteristic value is greater than a predetermined threshold; The detection system, wherein the characteristic value is at least one of phase, amplitude, and signal-to-noise ratio.
16. A sensor terminal that detects a status event using a wireless packet, a first radio device that transmits the wireless packet to a wireless base station; a second wireless device that receives the wireless packet; a sensor packet extraction unit that extracts a sensor packet transmitted by the first wireless device from a packet received by the second wireless device; a sensing processing unit that executes a sensing process to detect a state event indicating information of the real world by using the extracted sensor packet, The sensor terminal is characterized in that the sensor packet is a packet used for communication control, and the content of the packet changes depending on the surrounding environment.
17. A wireless base station that detects a status event using a wireless packet, Communicating with a wireless terminal that transmits and receives the wireless packets; a sensor packet extraction unit that extracts a sensor packet to be used for a sensing process from the packet transmitted by the wireless terminal; a sensing processing unit that executes a sensing process to detect a state event indicating information of the real world by using the extracted sensor packet, The wireless base station is characterized in that the sensor packet is a packet used for communication control and the content of the packet changes depending on the surrounding environment.
Citation Information
Patent Citations
Method, apparatus, and systems for wireless event detection and monitoring
WO2017100706A1
