System for controlling radio frequency sensing

Abstract

This invention relates to a system for controlling radio frequency (RF) sensing of a network 100. The network is adapted to perform different RF sensing modes, such as RSSI, CSSI, or Doppler-based modes. System 130 includes a providing unit 131 for providing environmental information, wherein the environmental information indicates the physical properties of one or more surfaces in the network's environment. A selection unit 132 is adapted to select and / or modify the RF sensing mode to be performed by the network, wherein the RF sensing mode is selected and / or modified based on the environmental information, and a control unit 133 is adapted to control the network 100 to perform the selected and / or modified sensing mode. Taking into account the physical properties of surfaces in the environment and modifying and / or selecting the RF sensing mode used accordingly increases the reliability and accuracy of RF sensing.

Description

Technical Field

[0001] This invention relates to a system, method, and computer program product for controlling radio frequency (RF) sensing in a network. Furthermore, this invention relates to a network including a system for controlling RF sensing in that network. Background Technology

[0002] Currently, radio frequency (RF) sensing is typically performed using two different measurements that indicate the wireless RF communication link. One common method involves using Received Signal Strength Indicator (RSSI) to measure the total attenuation of the RF signal between two network devices. Another common method uses Channel State Information (CSI), which represents the propagation of the RF signal along multiple spatial paths between the two network devices. Typically, CSI provides very detailed information about the environment with which the RF signal has interacted, and thus allows for very accurate sensing of, for example, very small movements (such as breathing movements). However, it has been noted that while CSI-based measurements are theoretically very accurate, in many applications, accuracy is very low due to unintended interactions between the RF signal and the environment. On the other hand, RSSI-based sensing methods do not offer the same sensing performance as CSI-based methods because RSSI measurements provide less detailed information about the RF signal. However, RSSI-based measurements are less affected by environmental conditions and therefore offer very reliable sensing performance.

[0003] Therefore, it would be advantageous to provide an RF system that allows for reliable and / or accurate sensing, if necessary.

[0004] WO 2021023725A1 discloses a method for selecting a communication technology in a radio frequency (RF)-based sensing system having one or more nodes. The RF-based sensing system is configured to perform RF-based sensing using one or more of two or more different communication technologies. Based on one or more parameters related to RF-based sensing in the RF-based sensing system, a communication technology for performing RF-based sensing in the RF-based sensing system is selected for one or more nodes. These parameters may include sensing application parameters, sensing quality parameters, system resource parameters, and context parameters. Summary of the Invention

[0005] The purpose of this invention is to provide a system, network, method, and computer program product that allows for reliable and accurate radio frequency sensing.

[0006] In a first aspect of the invention, a system for controlling radio frequency (RF) sensing of a network comprising a plurality of network devices is provided, wherein the network is adapted to perform different RF sensing modes, wherein the system comprises a) an environmental information providing unit for providing environmental information, wherein the environmental information indicates physical properties of one or more surfaces in the environment of the network; b) a selection unit for selecting and / or modifying at least one of RF sensing modes to be performed by at least a portion of the network as RF sensing, wherein at least one RF sensing mode is selected and / or modified based on the environmental information; and c) a control unit for controlling at least a portion of the network to perform at least one selected and / or modified RF sensing mode.

[0007] Because the selection unit is adapted to select and / or modify at least one RF sensing mode based on environmental information indicating the physical properties of one or more surfaces in the network's environment, RF sensing can be performed based on the RF sensing mode (e.g., RSSI or CSI sensing mode) that provides optimal results in the environment of the RF sensing network. In particular, considering the physical properties of the surfaces in the network environment and modifying and / or selecting the RF sensing mode used accordingly increases the reliability and accuracy of RF sensing.

[0008] This system is suitable for controlling radio frequency (RF) sensing in a network comprising multiple network devices. Typically, the RF sensing network includes at least two network devices, more preferably at least three. In one embodiment, the network may be a lighting system or part of a lighting system, in which case at least some of the network devices may be lamps whose light output is controlled based on the sensing results of the RF sensing. Furthermore, the network may be a smart home system or part of a smart home system, in which case the network devices may be, for example, smart home devices that perform functions in a user's home or office based on the sensing results. The network is formed by the network devices of the network, particularly by communication between the network devices. Communication, and therefore the network, can be based on any known network communication protocol, such as WiFi, ZigBee, Bluetooth, etc.

[0009] Network devices preferably each include a communication unit adapted to receive and transmit wireless signals (particularly radio frequency signals) and / or wired signals for communication. The communication unit of a network device (which may also be referred to as a network device communication unit) may include a receiving unit and a transmitting unit, wherein the receiving unit and transmitting unit may also be integrated with each other, for example, in the form of transceivers. Radio frequency signals can generally be understood as wireless signals consisting of electromagnetic radiation in the radio frequency range. The network may include this system and possibly other systems. However, the network may also be considered not to include this system. Typically, the system can be implemented as hardware and / or software on any dedicated or general-purpose computing device. In particular, the hardware and / or software representing the system may be provided as part of one or more network devices, wherein the system is then formed through communication between one or more network devices. However, the hardware and / or software representing the system may also be provided as a computing device that is not part of the network, wherein in this case, the computing device preferably includes a communication unit adapted to receive and transmit wireless signals and / or wired signals for communication, particularly for communication with network devices of the network. In addition, the system can also be located on an external computing device (such as the cloud, especially one or more servers) that is adapted to communicate with a network, particularly with at least one network device.

[0010] The network is adapted to perform different radio frequency (RF) sensing modes. Specifically, at least a portion of the network devices are adapted to transmit and receive RF signals for different RF sensing modes, and at least one network device is adapted to perform RF sensing based on the RF signals transmitted and received with respect to the RF sensing mode. In particular, the network and therefore the network devices can be adapted to or communicate with devices adapted to process the received RF signals according to the corresponding RF sensing mode (e.g., by applying a corresponding analysis algorithm).

[0011] In a preferred embodiment, the network is adapted to perform at least one of a CSI-based sensing mode, an RSSI-based sensing mode, and a Doppler-based sensing mode as a radio frequency (RF) sensing mode. In the CSI-based sensing mode, RF sensing is performed based on the measurement of the CSI of the RF signal between the transmitter and receiver (i.e., between two network devices). In the RSSI-based sensing mode, RF sensing is performed based on the measurement of the RSSI of the RF signal between the transceiver and transmitter (i.e., between two network devices). In the Doppler sensing mode, Doppler analysis is performed based on the RF signal received by at least one network device to determine, for example, the movement of a person in the area where RF sensing is performed.

[0012] An environmental information providing unit is adapted to provide environmental information. For example, an environmental information providing unit may be a receiving unit adapted to receive environmental information, for example, from user input or from another device used to provide environmental information. However, an environmental information providing unit may also be a storage unit that has already stored environmental information. The environmental information indicates the physical properties of one or more surfaces in the network's environment. One or more surfaces in the network's environment refer to any kind of surface of an object. Furthermore, the term "surface" is defined in the context of this invention as any part of an object that can interact with a radio frequency signal provided by one of the network devices. In most cases, this definition actually refers to the general surface of an object, i.e., the boundary surface or interface between the object and its environment. However, in particular, for objects that are at least partially transparent to radio frequency signals, the surface may also be located within the object, or it may refer to one or more surface layers that have different effects on radio frequency signals. For example, a radio frequency signal may penetrate only the fabric outer surface of a bench with low interaction and then interact primarily with the surface of the metal springs in the bench. In another example, some modern windows include a heat-reflective metallic film within a layer of the window, where radio frequency signals may also interact with the reflective metallic film but not with the glass layer surrounding the metallic film. The environmental information then indicates physical properties that can be any property of one or more surfaces in the network's environment. For example, physical properties can indicate or refer to the material and / or texture of a surface. However, physical properties can also indicate or refer to the position, size, and / or orientation of a surface relative to at least a portion of the network (e.g., relative to at least one network device). Furthermore, particularly if a surface within an object interacts with a radio frequency (RF) signal, the physical properties of the object's surface can also refer to the physical properties of one or more layers above the interacting surface that also interact with the RF signal, preferably the attenuation coefficients and / or thicknesses of one or more layers that affect the amount of RF signal reaching that surface within the object. For example, the number of surfaces in the environment can also be derived from this information and can also be provided as environmental information. Preferably, the provided physical properties allow for the determination of the expected interaction between the RF signal and the surface. However, if not all physical properties for determining the expected interaction are provided, the missing information can be replaced by predetermined general values. For example, if only position and orientation are provided, but material information is not provided, the material information can be replaced by a predetermined material (e.g., concrete). Preferably, the provided environmental information is independent of the sensing target of the RF sensing. For example, if the goal of radio frequency sensing is to determine the presence or absence of humans, it is preferable that the environmental information is unrelated to the presence of humans in the sensing area, i.e., it does not involve the physical attributes of humans. However, in other applications, the provided environmental information may also be related to the sensing target.For example, environmental information can be correlated with the expected attenuation of the sensing target, where the expected attenuation is low if the sensing target is a small mammal (such as a cat or dog), and the selection unit can be adapted to select a Doppler sensing mode for sensing the movement of the sensing target. Furthermore, physical properties can also indicate or represent the state of a surface. For example, environmental information can indicate whether a surface is wet (i.e., includes a thin layer of water), as this state information can also affect the interaction between the surface and the radio frequency signal.

[0013] The selection unit is adapted to select and / or modify at least one radio frequency (RF) sensing mode to be performed by at least a portion of the network as RF sensing. For example, the selection unit may be adapted to select whether the network should perform a CSI-based sensing mode, an RSSI-based sensing mode, and / or a Doppler-based sensing mode. Preferably, the selection unit selects only one of the RF sensing modes. However, depending on the specific situation, the selection unit may also be adapted to select more than one RF sensing mode to be performed. For example, two RF sensing modes may be performed in parallel. The selection unit may be adapted to determine the selection of more than one RF sensing mode preferably based on information about the execution capabilities of the network or one or more network devices. For example, different RF sensing modes may require different execution capabilities (such as computing or storage capabilities), and if the execution capabilities allow for the simultaneous execution of the selected RF sensing modes, the selection unit may be adapted to select more than one RF sensing mode. Furthermore, the selection unit may be adapted to modify the currently performed sensing mode or the selected sensing mode, for example, by customizing the corresponding sensing mode. This modification may refer to, for example, adapting the sensing mode to the current application or the situation where RF sensing should be performed. For example, modification could refer to adjusting the weights or thresholds provided as part of the sensing mode, or it could refer to selecting from multiple network devices the network device that should perform the radio frequency sensing mode.

[0014] The selection unit is adapted to select and / or modify the radio frequency (RF) sensing mode based on environmental information. For example, if information is provided that the surface has physical properties known to cause interference in a CSI-based sensing mode, the selection unit may be adapted to select an RSSI-based sensing mode instead of a CSI-based sensing mode for RF sensing, as the RSSI-based sensing mode is less affected by the interaction between the RF signal and the surface in the environment. The selection unit may be adapted to use predetermined rules that indicate which RF sensing mode should be selected and / or modified based on which physical properties of the surface. These rules may be predetermined by the user, for example, based on experience, or based on calibration, for example, during network setup. Furthermore, the selection unit may be adapted to learn rules itself; for example, the selection unit may be adapted to always select a first mode, wherein if the results of the first sensing mode seem unsuitable for the user, then the user is allowed to change the first mode, wherein the selection unit may then be adapted to learn in which cases the user selects a sensing mode different from the first sensing mode. Typically, the rules may refer to very simple rules, for example, instructing that if a specific material or a certain number of surfaces are present in the network environment, the corresponding sensing mode should always be performed. However, the rules can also be more complex, including multiple conditions and considering more than one physical property or more than one surface physical property to select and / or modify the radio frequency sensing mode.

[0015] The control unit is adapted to control at least a portion of the network to execute at least one selected and / or modified radio frequency (RF) sensing mode. For example, the control unit may be adapted to control network devices belonging to at least a portion of the network to transmit and receive RF signals suitable for the selected and / or modified RF sensing mode. Furthermore, the control unit may be adapted to cause the network or at least a portion of the network to apply a corresponding RF sensing algorithm corresponding to the selected and / or modified RF sensing mode to the received RF signals of at least a portion of the network.

[0016] In one embodiment, physical properties indicate the interaction of a surface with a radio frequency (RF) signal used for RF sensing. Preferably, the physical properties indicate the material and / or texture, as well as the location, physical size, and / or orientation of one or more surfaces relative to at least a portion of a plurality of network devices. Providing information indicating the location, physical size, and / or orientation of one or more surfaces relative to at least a portion of the network (i.e., relative to at least one network device), in addition to information indicating the material and / or texture of the surface in the network environment, allows for a more accurate determination of the surface's influence on RF sensing, and therefore also allows for a more accurate selection and / or modification of the corresponding RF sensing mode. Providing information about the location of the surface relative to at least a portion of the network allows, for example, determining whether reflections or scattering on the surface are likely to be experienced by RF signals along their path between two network devices. Furthermore, the physical size of the surface allows for determining, depending on its wavelength, whether the RF signal fully interacts with the surface, or what type of interaction is expected (e.g., scattering or reflection of the RF signal). Additionally, the orientation of one or more surfaces affects, for example, the direction in which the surface reflects the RF signal. Therefore, the selection unit can be adapted to apply rules (which also take into account all the information provided on the surface) to select and / or modify the RF sensing mode. For example, the rules applied by the selection unit may also include conditions regarding location, size, and / or orientation. For example, a rule may include a condition that states that the selection of the sensing mode requires the absence of reflective surfaces at a specific orientation and distance to one or more network devices.

[0017] Environmental information, such as location, physical dimensions, and / or orientation, as well as other information about physical properties, can be obtained using LiDAR scanning of the environment. However, environmental information can also be obtained from images of the environment. For example, an environmental information providing unit can be adapted to prompt a user to create overlay images from different perspectives within the room, and can then be adapted to infer the material, texture, location, size, and / or orientation of surfaces within the room from these images. Furthermore, blueprints, architectural plans, or structural designs can also be used to infer environmental information about the room, if feasible.

[0018] In a preferred embodiment, selection and / or modification involves simulating the propagation paths of radio frequency (RF) signals between network devices based on environmental information. The simulation of the RF signal propagation paths can be based on known physical laws governing electromagnetic wave propagation and its interaction with different materials and surfaces. Environmental information can then be provided as input to the simulation, and the corresponding propagation paths can be provided as output. The selection unit can then be adapted to select and / or modify the RF sensing mode based on the simulated propagation paths. For example, if the simulated propagation paths indicate that a certain amount of RF signal can interact with and is therefore affected by objects outside the room to be sensed, the selection unit can be adapted to select a CSI-based sensing mode as the RF sensing mode and modify the CSI-based sensing mode such that channels (i.e., propagation paths) that may have already interacted with objects outside the room are less strongly weighted during RF sensing inside the room.

[0019] In one embodiment, selection and / or modification are also based on region of interest (ROI) information, which indicates the location of an object of interest in the environment. ROI information may indicate the location or area of ​​an object (e.g., a person) and / or object in the environment in which radio frequency (RF) sensing is of interest. For example, ROI information may indicate a location, such as an area, in which a person is expected to perform certain activities that should be monitored by RF sensing. Therefore, the selection unit can use this information, for example, to determine the corresponding RF signal propagation in the ROI, and then select and / or modify the RF sensing mode accordingly. Specifically, it may be that a first RF sensing mode is best suited for RF sensing relative to the entire environment of the network, such as the entire environment of the room where the network is installed, while the RF selection unit can determine that a different second RF sensing mode may be more suitable relative to a specific ROI within the room. Therefore, the selection and / or modification of the RF sensing mode can be particularly suited to the specific intended application of RF sensing. Furthermore, in alternative or additional embodiments, selection and / or modification may be further based on non-ROI information, which indicates the location or area and / or object in the environment in which RF sensing is not of interest. For example, if radio frequency (RF) sensing is to be applied to detect the presence of a person in a garden, the region of non-interest could refer to the location in the garden where an ornamental pond is provided, since no one is expected to be present in the ornamental pond. The selection unit can then be adapted to select and / or modify the RF sensing mode while ignoring the performance of the RF sensing mode in the region of non-interest during this process, for example, by ignoring the influence of surfaces on the propagation path in the region of non-interest, or by ignoring the influence of certain surfaces on the detection of a person in the region of non-interest. This allows for more accurate selection and / or modification of the RF sensing mode relative to the area of ​​greater interest to the user.

[0020] In one embodiment, radio frequency (RF) sensing is applied for the detection of vital signs, gait, and / or posture of one or more individuals, and the region of interest (ROI) information refers to the area in which one or more individuals are present during a predetermined time period. In a preferred example, RF sensing is applied for the detection of vital signs of one or more sleepers, such as respiratory detection of children or the elderly. In this preferred example, the ROI information refers to the bedroom area where the person's bed is located, allowing the selection unit to select and / or modify the RF sensing mode to achieve accurate and reliable respiratory detection in the ROI. Specifically, in this case, for highly accurate vital sign detection, the selection unit may be adapted to select a CSI-based sensing mode and then to modify the CSI-based sensing mode such that signal paths from or into the ROI are given a higher weight than other signal paths, such as signal paths of RF signals reflected from opposite walls. In another aspect of this preferred example, the vital signs of more than two people (e.g., two people sleeping close to each other) should be monitored. In this case, for each of the two people, the ROI information refers to the area of ​​the bed in which the person is expected to spend most of the night. Furthermore, in this configuration, the selection unit is adapted to individually select and / or modify the RF sensing mode for each of the two regions of interest. Therefore, selecting and / or modifying the RF sensing mode based on region of interest information allows the system to be applied even when highly accurate differentiation of the sensed signals is necessary.

[0021] In one embodiment, if environmental information indicates at least one possible radio frequency (RF) signal propagation path involving shallow reflections on a horizontal plane between network devices in at least a portion of the network, the selection unit is adapted to select an RSSI-based sensing mode for at least that portion of the network as the RF sensing mode. For example, if the environmental information indicates the presence of a flat reflective surface (which can result in, for example, shallow reflections on a flat reflective surface), it can be shown that an RSSI-based sensing mode will provide more accurate sensing results. In this case, shallow reflection can be defined as a reflection of an RF signal that is not allowed to be detected by a network device that has already transmitted the RF signal, i.e., a reflection of an RF signal that is not allowed to return to the direction of the incident RF signal. In particular, this condition is satisfied if the incident RF signal and the reflected RF signal beams form an angle greater than 90°. In other words, this condition is considered satisfied if the angle between the incident RF signal beam and the reflective surface is less than 45° (preferably less than 30°). Typically, the selection unit can be adapted to determine whether the condition is met, for example by utilizing the simulation described above, or by utilizing other rules, such as a lookup table that indicates the relative position and physical properties of surfaces associated with at least a portion of the network (i.e., associated with at least a portion of the network devices that cause the condition to be met).

[0022] In one embodiment, if environmental information indicates high RF signal attenuation in the network's environment, the selection unit is adapted to select an RSSI-based sensing mode as the RF sensing mode. In this case, high attenuation can be defined relative to the signal-to-noise ratio (SNR) of the received RF signal. Specifically, for a particular application of the system, attenuation can be considered high if attenuation in the network's environment causes the SNR of the received RF signal to be unsuitable for the intended application. For example, if the system is intended for motion detection, a lower SNR and therefore higher attenuation are permissible regardless of whether the system is intended for respiratory or heart rate detection.

[0023] In one embodiment, if the environmental information indicates non-uniform surface material in the network's environment, the selection unit is adapted to select a CSI-based sensing mode as the radio frequency (RF) sensing mode. Non-uniform surface material can refer to, for example, a surface material comprising materials with different physical properties over a length range below the RF signal wavelength. For example, a wall on which multiple decorative objects (such as paintings, mirrors, etc.) are hung is composed of multiple materials with different properties; for example, the surface of a mirror has properties different from both a concrete wall surface and a painting surface. In such embodiments, the inventors have found that utilizing a CSI-based sensing mode (particularly a modified CSI-based sensing mode) results in more reliable and accurate RF sensing in environments with such walls. In a preferred embodiment, if the environmental information indicates leakage of the RF signal through a portion of the non-uniform wall material, the selection unit is adapted to modify the CSI-based sensing mode. For example, a portion of a wall including glass (such as a window) can cause the RF sensing signal to propagate to areas outside the intended sensing area, thus causing RF signal leakage. This leakage can lead to the influence of objects outside the intended sensing area on the RF signal, and thus inaccurate sensing results. In this context, utilizing a CSI-based sensing mode allows for the identification of potentially affected propagation paths, and the CSI-based sensing mode can be modified accordingly. Preferably, this modification includes identifying radio frequency (RF) signals with signal strengths below a predetermined signal strength and ignoring these RF signals during RF sensing.

[0024] In a preferred embodiment, the surface providing environmental information refers to the surface of the enclosure of at least a portion of the area providing the network. For example, the enclosure may refer to the walls of a room containing, for example, windows or doors, ceilings, and floors, but it may also refer to walls that do not reach the entire ceiling height, such as the walls of a dressing room in a fashion retail store, the walls of a toilet cubicle, a cubicle partition, etc. However, additionally or alternatively, the surface providing environmental information may also refer to the surface of objects in the network's environment, such as the surface of a table, desk, sofa, chair, screen, etc. Preferably, the environmental information includes at least one of a LiDAR scan of the environment, a panoramic image scan, a building floor plan, and / or an image of at least a portion of the network environment.

[0025] In another aspect of the invention, a network is proposed, comprising a) a plurality of network devices adapted to perform radio frequency sensing, wherein these network devices are adapted to perform different radio frequency sensing modes, and b) the system as described above.

[0026] In another aspect of the invention, a method for controlling radio frequency (RF) sensing of a network comprising a plurality of network devices is provided, wherein the network is adapted to perform different RF sensing modes, wherein the method includes a) providing environmental information, wherein the environmental information indicates physical properties of one or more surfaces in the environment of the network, b) selecting and / or modifying at least one RF sensing mode to be performed by at least a portion of the network as RF sensing, wherein the at least one RF sensing mode is selected and / or modified based on the environmental information, and c) controlling at least a portion of the network to perform at least one selected and / or modified RF sensing mode.

[0027] In another aspect of the invention, a computer program product for controlling radio frequency sensing of a network is provided, wherein the computer program product includes program code means that causes the system described above to perform the method described above.

[0028] It should be understood that the systems, networks, methods, and computer program products described above have similar and / or identical preferred embodiments, particularly as defined in the dependent claims.

[0029] It should be understood that the preferred embodiments of the present invention may also be any combination of the dependent claims or the above embodiments with the corresponding independent claims.

[0030] These and other aspects of the invention will become clear and explained with reference to the embodiments described below. Attached Figure Description

[0031] In the following figures:

[0032] Figure 1 An embodiment of a network including a system for controlling the network is illustrated schematically and exemplary.

[0033] Figure 2 Embodiments of a method for controlling a network are illustrated schematically and exemplary.

[0034] Figure 3 The attenuation of radio frequency signals from different materials is illustrated exemplarily. Detailed Implementation

[0035] Figure 1 A network 100 is illustrated schematically and exemplary, including a system 130 for controlling radio frequency (RF) sensing within the network 100. The network 100 includes network devices 120, 121, 122, 123, and 124 distributed throughout a room 110. In this exemplary embodiment, network devices 120, 121, 122, 123, and 124 refer to lights provided within the room for illuminating the room, such as light fixtures on the room's ceiling. In addition to their lighting function, network devices 120, 121, 122, 123, and 124 are adapted to perform RF sensing within the room 110. Specifically, for performing RF sensing, network devices 120, 121, 122, 123, and 124 are adapted to transmit and receive RF signals. Preferably, the RF signal refers to a communication signal used for communication within the network formed by network devices 120, 121, 122, 123, and 124; however, the RF signal may also refer to a signal transmitted solely for the purpose of RF sensing, without additional communication functionality. Typically, network 100, i.e., at least some of network devices 120, 121, 122, 123, and 124, is adapted to perform different radio frequency (RF) sensing modes. The executable RF sensing modes preferably refer to at least one of CSI-based sensing modes, RSSI-based sensing modes, and Doppler-based sensing modes. Preferably, the sensing mode of network 100 is a CSI-based sensing mode, as this allows for the most accurate sensing. However, the sensing mode can then be changed from a CSI-based sensing mode to any other sensing mode.

[0036] In this example, network 100 also includes a system 130 for controlling radio frequency sensing of network 100. In this exemplary embodiment, system 130 is provided as a standalone device, i.e., not integrated into one or more of network devices 120, 121, 122, 123, 124. In this case, system 130 is adapted to communicate with at least one of network devices 120, 121, 122, 123, 124, for example, preferably by communicating with network device 120 using radio frequency signals 140, 141. However, system 130 may also be adapted to communicate with all network devices 120, 121, 122, 123, 124 using any wired or wireless communication method. Furthermore, in another embodiment, system 130 may also be integrated as part of the hardware and / or software of one or more of network devices 120, 121, 122, 123, 124.

[0037] System 130 includes an environmental information providing unit 131, a selection unit 132, and a control unit 133. The environmental information providing unit 131 is adapted to provide environmental information. The environmental information provided by the environmental information providing unit 131 indicates the physical properties of one or more surfaces in the environment of network 100. For example, the environmental information provided by the environmental information providing unit 131 may indicate the physical properties of surfaces in room 110. For example, room 110 includes a wall surface 113 made of concrete, a door surface 112 made of wood, and two window surfaces 111 made of insulated glass. The environmental information provided by the environmental information providing unit 131 may then indicate, for example, the different materials used to manufacture these surfaces. However, the environmental information may also include the corresponding attenuation coefficient or reflectance coefficient of each of these materials. Furthermore, in a preferred embodiment, the environmental information also includes the location, size, and orientation of the surfaces in room 110. For example, the environmental information may be generated based on a LiDAR scan of room 110 and may then be provided by the environmental information providing unit 131. However, in other examples, such as a floor plan of room 110 or an image of room 110 provided by the user, can also be used as environmental information or for deriving environmental information. In addition to environmental information about the surfaces of the enclosure of room 110, environmental information about objects within room 110 can also be provided. For example, environmental information indicating the physical properties of the surface of table 114 provided in room 110 can be provided. However, it is preferable that the environmental information is unrelated to the radio frequency sensing target, for example, unrelated to person 115 who should be monitored using radio frequency sensing (e.g., by applying vital sign monitoring).

[0038] Selection unit 132 is adapted to select and / or modify one or more radio frequency (RF) sensing modes to be performed by at least a portion of network 100 (i.e., by at least some of network devices 120, 121, 122, 123, 124) as RF sensing. Network 100 (i.e., network devices 120, 121, 122, 123, 124) may be adapted to perform at least one of CSI-based sensing modes, RSSI-based sensing modes, and Doppler-based sensing modes as RF sensing modes. Typically, the same RF signal can be used to apply most different RF sensing modes, such that the main difference between different RF sensing modes lies in how the received RF signal is further processed for RF sensing, e.g., whether RSSI values, CSI information, or the spectrum are determined from the received RF signal. However, different RF sensing modes can also be modified to utilize different RF sensing signals, e.g., RF sensing signals within different wavelength ranges. Selection unit 132 is then adapted to select and / or modify the RF sensing mode that should be performed by at least a portion of network 100 based on environmental information provided by environmental information providing unit 131. For example, selection unit 132 may use predetermined rules that specify the conditions under which network 100 should change from one RF sensing mode to another, and how the RF sensing mode should be modified under specific conditions. In a preferred embodiment, selection unit 132 may utilize simulation of the propagation path of RF signals between network devices 120, 121, 122, 123, and 124. For example, such simulation may use known physical laws of wave propagation and interaction with corresponding surfaces to simulate the propagation path. For example, environmental information may be integrated into such simulation to provide the corresponding physical properties of surfaces on which RF signals are reflected, scattered, and / or attenuated, such as material, location, orientation, and size. For example, a LiDAR scan of room 110 may be used to identify all surfaces in room 110, and optionally, an image of room 110 may be used to provide information about the materials of these surfaces. The simulation may then use known physical laws to simulate possible propagation paths of RF signals, for example, transmitted from network device (exemplarily network device 120) to network device (exemplarily network device 124). The simulated propagation path allows for the derivation of, for example, whether CSI-based sensing is possible in room 110, or whether the attenuation of some simulated propagation path is expected to be higher than a predetermined threshold for which RSSI-based sensing has been observed to provide better sensing results. Therefore, in this case, selection unit 132 can be adapted to select and / or modify the RF sensing mode based on the simulated propagation path.

[0039] Control unit 133 is then adapted to control at least a portion of the network, i.e., network devices belonging to at least a portion of network 100, to execute selected and / or modified radio frequency sensing modes. For example, control unit 133 may communicate with network devices 120, 121, 122, 123, 124 using signals 141, 140, causing network devices 120, 121, 122, 123, 124 to process received radio frequency signals according to the selected and / or modified radio frequency sensing mode.

[0040] Figure 2 An embodiment of a method 200 for controlling radio frequency (RF) sensing of a network (such as network 100) is illustrated schematically and exemplary. Method 200 includes a first step 210: providing environmental information, wherein the environmental information indicates the physical properties of one or more surfaces in the environment of the network (such as network 100). For example, as explained above, the environmental information may refer to providing the physical properties of surfaces within room 110 where network 100 is located. In a second step 220, method 200 includes selecting and / or modifying one of the RF sensing modes to be performed by at least a portion of the network as an RF sensing mode, wherein the RF sensing mode is selected and / or modified based on the environmental information. For example, the selection and / or modification may be performed according to selection unit 132 relative to the example explained above. In a final step 230, method 200 includes controlling at least a portion of the network to perform the selected and / or modified RF sensing mode.

[0041] Below, some preferred embodiments of the invention will be described in detail. In a first embodiment, the system can be applied to simultaneously track the sleep of two people sharing a bed. In this embodiment, environmental information indicates the type of building material and its corresponding location in the room. Based on the environmental information, the dominant radio frequency signal multipath can be determined, for example, by utilizing a corresponding simulation. The selection unit is then adapted to select and modify a CSI-based sensing mode for respiration detection, which is capable of simultaneously tracking the sleep of two people sharing a bed. Generally, it has been found that environmental information indicating, for example, building materials, allows for improved context-aware radio frequency sensing applications, such as long-range respiration detection, gait detection, posture detection, and heart rate detection. For example, current state-of-the-art respiration detection / sleep monitoring solutions are unable to perform radio frequency sensing simultaneously on two people sharing a bed.

[0042] For this specific application, network devices could refer, for example, to first and second lamps located near the heads of the first and second individuals, respectively, and a third central lamp (e.g., a ceiling-mounted lamp) installed further away from these two individuals. The environmental information then preferably indicates the type of building materials and their corresponding positions in the bedroom relative to these three lamps. This environmental information can be provided, for example, by prompting the user to initiate a panoramic image scan or by directly inputting the information. For instance, specific building material types (e.g., wood versus brick versus metal) can be inferred from the panoramic image by their shape, texture, color, or size, or through user input via a selection unit.

[0043] Subsequently, the selection unit can be adapted to estimate the dominant radio frequency (RF) signal multipath between the first and third beams (i.e., for the "first channel") and between the second and third beams (i.e., for the "second channel"). Furthermore, information about regions of interest (e.g., the left and right sides of the bed) can be provided, for example, via user input. The selection unit can then utilize, for example, a lookup table detailing how each common building material interacts with the RF signal. Finally, based on environmental information indicating the building materials and spatial arrangement, the selection unit can be adapted to select a CSI-based sensing mode as the RF sensing mode and / or modify the CSI-based sensing mode such that: a first subset of the wireless multipath from the first channel is selected, associated with the first person lying on the first side of the bed, but interference from the second person is avoided or limited; and a second subset of the wireless multipath from the second channel is selected, primarily associated with the second person lying on the second side of the bed. Specifically, based on an understanding of the spatial arrangement, building materials, and their interaction with the wireless signal, the selection unit can be adapted to correlate which multipaths extracted by the CSI measurements are associated with which multipaths in the physical world based on attenuation and signal delay. Optionally, the selection unit may be adapted to visualize multipath during the setting of a sensing mode, such as on a smartphone display, allowing the user to adapt or modify the simulated multipath.

[0044] Typically, it is preferable that the control unit is adapted to identify the presence of sensing blind spots within the sensing area, for example, based on information provided from a walking test during the execution of a sensing mode. In this case, the control unit may be adapted to prompt the user to begin providing environmental information, for example, by suggesting an analysis of building materials near the blind spot. The selection unit can then modify the CSI-based sensing mode, for example, based on the provided environmental information, to fine-tune the RF sensing performance.

[0045] In another preferred embodiment, if the RF signal path between the two network devices includes shallow-angle signal reflections containing the majority of the total transmitted radio power, the selection unit is adapted to select an RSSI-based sensing mode instead of a CSSI-based sensing mode. Typically, CSI-based sensing can isolate a specific multipath of interest, for example, by modifying the CSI-based sensing mode accordingly. For example, the RF signal channel between the first and second network devices may consist of 12 different signal multipaths. Generally, a direct signal path is preferred for RF sensing because it provides the best signal-to-noise ratio for human presence sensing, even though it actually receives only 1 / 12 of the total transmitted signal power emitted by the network devices. Conversely, by definition, an RSSI-based sensing mode utilizes the average of the received signals from all 12 multipaths between the two network devices and is therefore heavily influenced by reflections and losses due to building materials and physical obstacles. However, it can be shown that a CSI-based sensing mode typically provides reduced sensing accuracy when RF sensing covers strong, shallow-angle reflections, for example, from a floor or a large marble kitchen countertop. For example, a radio frequency (RF) signal path in a living room might begin with a floor lamp acting as a network device next to the sofa, and the signal arrives at a shallow angle on the concrete floor a few meters away in front of the dining table. The signal is then reflected by the concrete floor with minimal attenuation and then reaches a receiving table lamp, which is also a network device located at the other end of the living room. This specific floor-based path (i.e., the "second path") is unique compared to other RF signal multipaths in a living room that potentially involve multiple reflections, because the total path length is only slightly longer than the direct path (i.e., the "first path"), and there is no reflection between the two lamps. Therefore, a CSI-based sensing pattern might have difficulty distinguishing the second path via the floor from the first path, as the first and second paths have nearly identical arrival times and similar signal amplitudes. In this case in a CSI-based sensing pattern, the first and second paths might unexpectedly converge. Thus, if the floor reflection of the second path is altered, for example due to an object lying on a marble kitchen countertop or a wet surface of the floor, a CSI-based sensing pattern looking at the unexpected convergence of the first plus second path might notice an increase beyond the baseline and incorrectly infer the presence of a stationary person. When using the RSSI-based sensing mode, the radio frequency signal reaching the second lamp still increases due to modified reflection caused by the presence of a physical object on the surface. However, since the "odd" first wireless path only contributes a small amount to the overall accumulated RSSI signal reaching the receiver lamp, it does not significantly contribute to the overall decision-making process in the RSSI-based sensing mode. Therefore, in this case, it is preferable that the selection unit is adapted to determine this shallow-angle reflection condition based on environmental information and select the RSSI-based sensing mode as the radio frequency sensing mode.

[0046] On the other hand, if the primary reflection of the wireless radio frequency signal occurs not on a horizontal floor, but on a vertically oriented obstacle (such as a wall, a glass mirror, or the side of a bookshelf), this will result in a significant difference in path length between the reflected signal path and the direct signal path between the transmitter and receiver network devices. Therefore, also in CSI-based sensing mode, it is easy to distinguish, for example, via time-of-arrival analysis, which received signal belongs to the direct signal path and which belongs to the reflected signal path via a sidewall. Thus, in this case, the CSI-based sensing mode does not suffer from the problems experienced by floor reflections and is therefore as reliable as the RSSI-based sensing mode. In this case, the selection unit can be adapted to select the CSI-based sensing mode as the sensing mode, as this generally provides higher sensing accuracy.

[0047] In one embodiment, the selection unit is adapted to select an RSSI-based sensing mode for a space with high RF signal attenuation due to the presence of certain building materials in the RF signal path between two network devices. The RSSI-based sensing mode relies on the accumulated (i.e., aggregated) signal composed of all wireless signal multipaths between the network devices. Therefore, the RSSI-based sensing mode works more reliably than the CSI-based sensing mode whenever the RF sensing signal between the two network devices is attenuated to such a degree that it jeopardizes the CSI-based sensing mode. In practice, this can occur for various reasons. For example, if the RF signal must be transmitted over long distances (e.g., in a garden environment), or if inter-floor sensing is utilized (e.g., in an office application), if network devices on the first floor are used to sense people walking on the second floor above, and furthermore, if highly absorbing building materials are present in the sensing area (which can effectively render a subset of the signal multipaths useless for CSI-based sensing). In the above cases, as done by CSI-based sensing, attempting to analyze each multipath signal component of the CSI signal individually may lead to indeterminate results due to insufficient signal-to-noise ratio of most available signal multipaths. Even if sensing algorithms can provide useful data, they may be more prone to false alarms or false negatives, or incorrect estimates of metrics such as respiratory rate or heart rate, and thus include reduced accuracy and reliability.

[0048] In contrast, RSSI-based sensing mode analysis aggregates the received radio frequency signals, which, due to a sufficient signal-to-noise ratio, allows for definitive results, albeit with less rich information. Therefore, it is preferable that the selection unit is adapted to select an RSSI-based sensing mode, rather than a CSI-based sensing mode, for rooms where environmental information indicates the presence of high-attenuation building materials.

[0049] In one embodiment, the selection unit is adapted to select a CSI-based sensing mode for a room with non-uniform wall material, for example, to avoid “corridor problems” in an office. In a CSI-based sensing mode, it is possible to intentionally exclude a first subset of unwanted signal multipaths; for example, the CSI sensing mode can be modified to intentionally exclude those signal multipaths that are determined, for example, based on proportionally low arrival signal strength at the network device, and have clearly passed through the wall twice. This allows for the mitigation of false triggering associated with properly named “corridor problems” in a modified CSI-based sensing mode. Therefore, it can be avoided that radio frequency sensing signals leaking from the room into the corridor and back into the room could cause passersby in the corridor to generate false occupancy triggers within the room.

[0050] Typically, the attenuation of undesirable multipath propagation, such as that involving the exterior of a room, depends on the building materials. For example, a glass wall in a conference room attenuates a 2.4 GHz radio frequency signal only slightly, while a brick wall attenuates it significantly. The table below shows further examples of material attenuation for radio frequency signals. Specifically, it shows transmission and reflection measurements of window glass and a red brick wall composed of 17 bricks measuring 203 mm (w) × 51 mm (h) × 102 mm (d) at different commonly used radio frequency signal frequencies.

[0051]

[0052] Therefore, for example, if a meeting room has a glass wall leading to a corridor, an RSSI-based sensing mode will be erroneously triggered by passersby, while if the wall is made of metallic-coated hot glass or brick, the RSSI-based sensing mode will work reliably. Thus, the selection unit can be adapted to utilize environmental information to determine whether a situation might occur that could lead to radio frequency signal leakage, and then select a modified CSI-based sensing mode. Typically, environmental information indicating the spatial arrangement of the room's building materials can be obtained from Building Information Modeling (BIM) files or from user input. Preferably, the selection unit is adapted to utilize environmental information to select an RSSI-based sensing mode for the portion of network equipment closest to the enclosure containing the glass-like material, while selecting a CSI-based sensing mode for the portion of the network covering a sensing area far from such an enclosure. In this example, modifying the CSI-based sensing mode during system setup can be avoided.

[0053] Currently, radio frequency sensing is divided into two groups: a first group of systems using CSI-based sensing modes and a second group of systems using RSSI-based sensing modes. The inventors recognize that even with advanced wireless equipment capable of CSI-based sensing, there may be specific application scenarios where building materials have such a significant impact on the accuracy of CSI-based sensing that simple RSSI-based sensing is superior to CSI-based sensing. At the other extreme, the inventors recognize that CSI-based sensing may occasionally require detailed customization, such as for simultaneously performing respiratory detection on two people sharing a bed.

[0054] RSSI is a widely used measurement of the overall attenuation of wireless communication signals between two network devices. On the other hand, CSI has emerged as an alternative measurement in recent years, where CSI measurement represents how a wireless signal propagates from the transmitter to the receiver along multiple spatial signal paths at a specific carrier frequency in the radio frequency signal channel. Therefore, CSI-based sensing captures the signal characteristics of the surrounding environment because the amplitude and phase of the CSI measurement are affected by multipath effects, including amplitude attenuation and phase shift of the radio frequency signal. For example, for modern WiFi systems with MIMO-OFDM, CSI measurement provides a 3D matrix representing complex values ​​of amplitude attenuation and phase shift.

[0055] Time series of multiple CSI measurements capture how wireless signals propagate through surrounding physical objects and people in the time, frequency, and spatial domains. Therefore, existing techniques teach the transmission of wireless signals via wireless multipath channels and the use of artificial intelligence (AI) algorithms to analyze the time series of CSI measurements, enabling a wide variety of different RF sensing applications. For example, the variation in CSI amplitude in the time domain exhibits different patterns for different people, activities, postures, etc., which can be used for human presence detection, motion detection, activity recognition, posture recognition, and human identification.

[0056] Existing techniques also teach that the CSI phase shift observed in the spatial and frequency domains (i.e., transmit / receive antennas and carrier frequencies) is related to signal transmission delay and direction of arrival. Besides occupancy and activity detection, this can be used for personnel location and tracking across building spaces using radio frequency sensing. The CSI phase shift in the time domain can have different dominant frequency components, which have been used, for example, in existing techniques to estimate respiratory rates using CSI-based radio frequency sensing.

[0057] Since both CSI-based and RSSI-based sensing modes rely on the same physical network devices, assuming both are derived from the same protocol (e.g., WiFi) and use the same radio frequency signals to "probe" the space, the physical properties of signal propagation are identical for both modes. That is, the RF multipath channels within a room do not change simply because we want to extract CSI or RSSI data from the received RF signals. Furthermore, both RSSI-based and CSI-based sensing modes analyze the time series of RF signal channels between two network devices. However, compared to RSSI-based sensing modes, CSI-based sensing modes can extract metrics that can be correlated with the multipath characteristics of the physical building space from each distinct signal subchannel. CSI amplitude and phase are affected by RF signals from multiple paths within the room, not just a single signal path. Therefore, unlike the simple measurement mentioned in RSSI, which is performed locally by any RF receiver for "housekeeping" purposes, for CSI-based sensing modes, multipath information must first be derived from the raw CSI data measured by the RF receiver. For example, a 20MHz WiFi channel can have 64 CSI subcarrier frequencies. Since each of these sub-frequency interactions with material objects (such as the brick walls of a room or the upholstery of a sofa) is at least somewhat different, an analysis of how the 64 sub-components behave as a whole (e.g., their relative differences) indicates the multipath behavior of the architectural space.

[0058] To illustrate the profound impact of building materials on wireless channel metrics (crucial for accurate radio frequency sensing), a review below is provided of how radio waves interact with different commonly used building materials. Radio waves propagate through electromagnetic radiation and interact with the environment through reflection, refraction, diffraction, absorption, polarization, and scattering. Figure 3 The attenuation measurements of various building materials within a typical frequency range used in RF sensing applications are shown. Figure 3It can be concluded that the attenuation differences between different materials are very significant. Therefore, the room's construction, as well as the spatial arrangement and cumulative surface area of ​​each existing building material type, can affect the RF multipath signal characteristics of that particular room. For a typical wireless link budget of 90 dB (defined by the ratio of maximum transmit power to minimum receive sensitivity), the total wireless attenuation difference of only a few dB caused by building materials is well within the detectable range of an RF sensing system. The presence of metallic materials in the area strongly influences RF signal multipath propagation. In recent decades, metallic construction materials have been used more and more frequently, for example, to improve the thermal insulation of buildings. For example, in most modern homes today, aluminum foil is incorporated into foil-backed drywall and insulation panels to provide low thermal emissivity and vapor resistance while having minimal impact on room size. For example, "multifoil" building products are often nailed to roof eaves, hollow walls, or laid on attic floors. Because these multi-foil building materials typically consist of more than ten layers of aluminum foil and insulation, dormer windows or attics in the same house can exhibit very different RF sensing signal propagation patterns compared to a living room in the same house without aluminum building materials. Similarly, mirrors (e.g., in bathrooms or dressing rooms) are known to cause strong reflections. Likewise, most modern houses feature low-emissivity windows, improving thermal performance by adding a thin layer of metal or metal oxide to one of the window panes. This thin metal layer on the window glass affects the propagation of a subset of wireless signals involved in RF multipath transmission in the window area. Because changes in construction methods and materials have worsened the building penetration loss of cellular signals, modern buildings now intentionally create small sections in certain walls that allow cellular signals to penetrate freely into the interior of the house, and from one room to another. However, these “odd” spots in building walls also inadvertently affect the multipath propagation of RF sensing signals, for example, allowing RF sensing signals to inadvertently leak out of rooms (e.g., from the kitchen through a wall to an adjacent living room).

[0059] Therefore, it is advantageous for RF sensing systems to consider the building materials used in a room, specifically understanding how specific parts of the room reflect, absorb, or scatter RF signals. The two most readily understood interaction mechanisms between radio waves and building materials are reflection and absorption. For all practical purposes, large metal structures (such as steel beams and radiators in a room) can be considered perfect reflectors at the RF sensing frequencies of interest. Therefore, unlike the thin metal films described earlier, these thick metal structures do not allow significant RF sensing signals to pass through. Any reflection of the RF sensing signal (e.g., by the aforementioned large metal structures) introduces additional multipath components into the RF signal transmission channel between two RF sensing network devices. In practice, the strongest reflection effects are produced by large, smooth planar building bodies (such as walls, floors, ceilings, windows, and doors), because the surfaces of these objects are typically quite smooth at the typical RF sensing frequencies used, and therefore these objects reflect wireless RF sensing signals very strongly.

[0060] Diffraction is another mechanism that can affect multipath signals sensed by radio frequency (RF) within a room. Radio wave diffraction occurs where two different building materials intersect, or where there are abrupt changes in the surface shape of the materials. In practice, RF signal diffraction occurs in buildings, typically at corners and edges where two or more walls / ceilings intersect, and at the edges of windows and doors where wood or glass panels intersect with walls. While diffraction is generally a “weaker” mechanism than transmission for transmitting RF signals from one subspace to another within a building, existing technology has also taught that diffraction can sometimes be a primary mechanism for providing cellular or home Wi-Fi radio coverage in specific locations within a room or building. For example, diffraction can significantly help transmit wireless RF signals to blind spots behind high-attenuation metal walls or large metal objects. In this particular “blind spot” location, RF signals diffracted from other parts of the room may even be much stronger than those reaching the receiver directly through obstacles or via reflection. Existing technology has taught that the strength of diffracted wireless RF signals depends primarily on the path geometry, the shape of the diffraction edges, and the frequency. To some extent, it also depends on the electrical properties of the materials that include the diffraction edges, such as the corners of plaster walls in a room reinforced with long metal strips, but this dependence is usually weaker than other factors.

[0061] Another interaction mechanism between sensing radio waves and building materials is wireless scattering. Existing technology teaches that even at relatively low frequencies (such as 2.4 GHz), large clutter such as furniture and people in a room can often be modeled as scattering sources. Furthermore, wireless scattering occurs when radio waves strike rough surfaces. Whether a surface appears rough or smooth under RF depends on the relative size of the surface irregularity to the wavelength and the angle of incidence of the radio wave. At 2.4 GHz, the wavelength is approximately 12.5 cm, so if the irregularity is less than one-tenth of the wavelength (1.25 cm in the case of 2.4 GHz), the surface can be considered smooth at all angles of incidence. Therefore, at the wavelengths currently used for RF sensing by network devices, most interior and exterior walls can be considered smooth, and the effect of scattering is negligible. However, in the future, for 60 GHz WiFi (λ = 0.5 cm), a surface irregularity of 0.5 mm will already cause significant scattering in the propagation of RF signals. Therefore, it is preferable that the selection unit may also take into account the 60GHz WiFi sensing scattering effect, for example, in applications where the purpose is to detect clutter (such as kitchen tools) on a smooth tabletop.

[0062] Generally, CSI-based sensing patterns are preferred because they inherently offer more insights than RSSI-based sensing patterns. However, the inventors have discovered that RSSI-based sensing patterns are preferred in certain situations because the presence of certain building materials significantly affects the multichannel behavior of a room. Similarly, for some high-value sensing applications (such as respiration or heart rate detection), it may be necessary to purposefully modify the CSI-based sensing pattern by selecting a subset of the RF signal multipaths in the room to be used by the CSI-based sensing pattern.

[0063] Therefore, the inventors have noted that, in practice, the selection criteria for when to apply RSSI-based and CSI-based sensing modes strongly depend on the physical properties of the room, such as surface, material, and shape. Thus, the system described in the above embodiments is suitable for first utilizing environmental information, for example, to analyze and locate building materials present in the room, where the environmental information can be provided, for example, via panoramic scanning, LiDAR scanning, or user input during network initialization. Subsequently, the system can be adapted such that the collected building material information (e.g., as environmental information) is used as input to a selection unit that can utilize, for example, the selection criteria proposed above, to select when to apply a CSI-based sensing mode rather than a simple, less computationally intensive RSSI-based sensing mode and / or when to modify one of the radio frequency sensing modes. For example, as described above, in cases where the radio frequency sensing signal encompasses strong, shallow-angle reflections from a hard floor, the selection unit can be adapted to select an RSSI-based sensing mode. Furthermore, the selection and / or modification of the radio frequency sensing mode is also useful for determining the optimal radio frequency sensing setting for simultaneously monitoring the breathing of two people sharing a bed.

[0064] While the exemplary embodiments described above primarily address the distinction between RSSI-based and CSI-based sensing, the same principles can be applied to other radio frequency sensing modes. For example, if the target to be sensed is expected to move substantially parallel to a determined spatial direction, the selection unit may be adapted to determine the expected spatial direction relative to the dominant radio frequency multipath signal based on environmental information and select a Doppler sensing mode, as this results in a relatively high Doppler frequency shift. This situation might occur, for example, in a narrow corridor of an office building, where the direction of movement is substantially determined by the corridor itself, with little possibility of movement in other directions. Where it is possible for the selection unit to determine that the target to be sensed is moving substantially perpendicular to the determined spatial direction, or in multiple directions, the selection unit may be adapted to select an RSSI-based and / or CSI-based sensing mode.

[0065] Furthermore, if the environmental information also includes information about the sensing target, indicating that the interaction between the moving sensing target and the radio frequency signal is relatively weak (e.g., due to low weight), the selection unit is preferably adapted to select a Doppler sensing mode. For example, tracking a cat via Doppler sensing is advantageous because its small weight results in minimal absorption, but its movement is relatively large. On the other hand, in cases where the aim is to detect respiration or heart rate, the selection unit may be adapted to determine, based on environmental information, whether a radio frequency signal with a suitable signal strength reflected from, for example, the chest of a person sleeping in bed can be detected. This determination may be based on, for example, a simulation of the radio frequency signal path as described above. If this condition is met, it is preferable that the selection unit is adapted to select a Doppler sensing mode; if this condition is not met, it is preferable that the selection unit selects a sensing mode based on RSSI or CSI.

[0066] By studying the accompanying drawings, the disclosure, and the appended claims, those skilled in the art can understand and implement other variations of the disclosed embodiments in practicing the claimed invention.

[0067] In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality.

[0068] A single unit or device can perform the functions of several items listed in the claims. The mere fact that certain measures are referenced in mutually different dependent claims does not indicate that a combination of these measures cannot be used advantageously.

[0069] Processes performed by one or more units or devices (e.g., providing environmental information, selecting and / or modifying RF sensing modes, controlling network devices, etc.) can be performed by any other number of units or devices. These processes can be implemented as program code devices of computer programs and / or as dedicated hardware.

[0070] Computer programs can be stored / distributed on suitable media, such as optical storage media or solid-state media, provided together with or as part of other hardware; but they can also be distributed in other forms, such as via the Internet or other wired or wireless telecommunications systems.

[0071] Any reference numerals in the claims should not be construed as limiting the scope.

[0072] This invention relates to a system for controlling radio frequency (RF) sensing in a network. The network is adapted to perform different RF sensing modes, such as RSSI, CSSI, or Doppler-based modes. The system includes a providing unit for providing environmental information, wherein the environmental information indicates the physical properties of one or more surfaces in the network's environment. A selection unit is adapted to select and / or modify the RF sensing mode to be performed by the network, wherein the RF sensing mode is selected and / or modified based on the environmental information, and a control unit is adapted to control the network to perform the selected and / or modified RF sensing mode. Taking into account the physical properties of surfaces in the environment and modifying and / or selecting the RF sensing mode used accordingly increases the reliability and accuracy of the RF sensing.

Claims

1. A system for controlling radio frequency sensing of a network (100); wherein the network (100) includes a plurality of network devices (120, 121, 122, 123, 124), and wherein the network (100) is adapted to perform different radio frequency sensing modes, wherein the system (130) includes: An environmental information providing unit (131) is configured to provide environmental information, wherein the environmental information indicates the physical properties of one or more surfaces in the environment of the network (100), wherein the physical properties indicate the material and / or texture of the one or more surfaces; Selection unit (132) for selecting and / or modifying at least one of radio frequency sensing modes to be performed by at least a portion of the network (100) as radio frequency sensing, wherein at least one radio frequency sensing mode is selected and / or modified based on the environmental information, and Control unit (133) for controlling at least a portion of the network (100) to perform at least one selected and / or modified radio frequency sensing mode.

2. The system according to claim 1, wherein, The network (100) is adapted to perform at least one of the following as radio frequency sensing modes: a sensing mode based on channel state information (CSI), a sensing mode based on received signal strength indicator (RSSI), and a sensing mode based on Doppler.

3. The system according to claim 1, wherein, The physical properties indicate the interaction between the surface and the radio frequency signal used for radio frequency sensing.

4. The system according to claim 3, wherein, The selection and / or modification includes simulating the propagation path of radio frequency signals between network devices (120, 121, 122, 123, 124) of at least a portion of the network (100) based on the environmental information.

5. The system according to claim 3, wherein, The selection and / or modification are further based on region of interest information, which indicates the location of the object and / or object in the environment in which the radio frequency sensing is interested.

6. The system according to claim 5, wherein, The selection unit (132) is arranged to select and / or modify at least one of the radio frequency sensing modes to be performed by at least a portion of the network (100) as radio frequency sensing for the detection of vital signs, gait and / or posture of one or more persons, and the region of interest information refers to the region in which the one or more persons exist within a predetermined time period.

7. The system according to claim 1, wherein, If the environmental information indicates at least one possible radio frequency signal propagation path involving shallow reflections on a horizontal plane between network devices (120, 121, 122, 123, 124) of at least a portion of the network (100), the selection unit is adapted to select an RSSI-based sensing mode as the radio frequency sensing mode for at least that portion of the network (100); wherein the shallow reflections include reflections of radio frequency signals that are not allowed to be detected by network devices (120, 121, 122, 123, 124) that have already transmitted radio frequency signals.

8. The system according to claim 2, wherein, If the environmental information indicates high radio frequency signal attenuation in the environment of the network (100), the selection unit is adapted to select an RSSI-based sensing mode as the radio frequency sensing mode.

9. The system according to claim 1, wherein, If the environmental information indicates non-uniform surface material in the environment of the network (100), the selection unit (132) is adapted to select a CSI-based sensing mode as the radio frequency sensing mode.

10. The system according to claim 9, wherein, If the environmental information indicates that the radio frequency signal is leaking through a portion of a non-uniform wall material, the selection unit (132) is adapted to modify the CSI-based sensing mode.

11. The system according to claim 10, wherein, The modification includes identifying radio frequency signals with signal paths having signal strengths lower than a predetermined signal strength, and ignoring these radio frequency signals during radio frequency sensing.

12. The system according to claim 1, wherein, The environmental information includes at least one of the following: LiDAR scan of the environment, panoramic image scan, building floor plan, and / or an image of at least a portion of the environment of the network (100).

13. A network (100) includes: Multiple network devices (120, 121, 122, 123, 124) suitable for performing radio frequency sensing, wherein the network devices (120, 121, 122, 123, 124) are adapted to perform different radio frequency sensing modes, and The system (130) according to any one of claims 1 to 12.

14. A method for controlling radio frequency sensing of a network (100); wherein the network (100) includes a plurality of network devices (120, 121, 122, 123, 124), and wherein the network (100) is adapted to perform different radio frequency sensing modes, wherein the method includes: Provide environmental information, wherein the environmental information indicates the physical properties of one or more surfaces in the environment of the network (100), wherein the physical properties indicate the material and / or texture of the one or more surfaces; Selecting and / or modifying at least one of the radio frequency sensing modes to be performed by at least a portion of the network (100) as radio frequency sensing, wherein at least one radio frequency sensing mode is selected and / or modified based on the environmental information, and Control at least a portion of the network (100) to perform at least one selected and / or modified radio frequency sensing mode.

15. A computer program product for controlling radio frequency sensing of a network (100), wherein the computer program product includes program code means that causes the system (130) according to claim 1 to perform the method according to claim 14.

Images