TR202100759A2 - DETECTION APPLICATION DEFINITION AND FORECAST - Google Patents

Abstract

Some applications in the present disclosure relate to detection application identification. In particular, a wireless signal is obtained for further processing. Next, the presence of a detection signal in the received signal is predicted, and said detection signal is a signal produced by a detection application. Then, based on the estimation of the detection signal, wireless reception, transmission or detection is performed. Detection application detection and/or identification can be performed by a trained module such as a machine learning-based module. Wireless reception, transmission or detection based on the forecast result may also include channel access, resource allocation, use of the detected detection signal for its own detection purposes, or the like.

Description

DESCRIPTION DETECTION APPLICATION DESCRIPTION AND PREDICTION The present description relates to the field of wireless communication and sensing. In particular, the disclosure provides devices and methods for wireless communication and sensing systems. BACKGROUND ART Wireless communications is a field that has been developing for several decades. Important standards organizations that serve as examples include the 3rd Generation Partnership Project (BGPP) and IEEE 802.11, commonly referred to as Wi-Fi. In practice, devices that support such standards or devices with proprietary communications and sensing applications may at least partially share the spectrum. Cognitive radio is one of the emerging technologies to utilize system spectrum. Cognitive radio devices are expected to dynamically use the best wireless channels in their environment to improve spectrum efficiency. To achieve this, spectrum occupancy information may be required to help model and predict spectrum availability for efficient dynamic spectrum access. Spectrum occupancy estimation may be based on using information on previous spectrum occupancy to predict future occupancy. Such a prediction is based on exploiting the inherent correlation between past and future occupancy. Some approaches use time domain correlation and thus use the spectrum estimate as a time series estimate. Some approaches also consider exploiting correlation along the frequency axis and thus exploit time-frequency correlation. Correlation may also exist in the spatial domain. Therefore, it may be desirable to take advantage of correlation in all mentioned areas. Some specific studies use traditional statistical methods for spectrum estimation, such as an autoregressive model-based method or Bayesian inference-based methods. There are some approaches that use machine learning (ML) methods, such as wavelet neural networks and shallow artificial neural networks. More recent works mention deep learning (DL)-based approaches to further improve performance. DL-based methods can use correlation across multiple domains. BRIEF DESCRIPTION Methods and techniques, as well as corresponding devices, that facilitate cognitive radio applications and other approaches to improve wireless spectrum utilization are described. In particular, it is recognized in the present disclosure that the identification of sensing applications can be used to further improve one or more aspects of wireless spectrum utilization, especially when the spectrum is possibly shared by different communication and/or sensing devices. The invention is defined by independent claims. Some example embodiments are provided by the dependent claims. In some exemplary embodiments, a method is developed for defining a sensing application: obtaining a received wireless signal; predicting in the received signal by a trained module the presence of a detection signal, which is a signal generated by a sensing application; It includes the steps of performing wireless reception, transmission or sensing operations based on the anticipated presence of said sensing application. Other exemplary embodiments relate to a computer-readable medium and program storing a program for executing said methods, as well as a training method, training apparatus, and an apparatus for determining a sensing application. These and other features and characteristics of the subject matter currently disclosed herein, as well as the methods of operation and functions of the relevant elements of the structures and the combination of parts and economies of production, will become clearer when considering the following description and the appended claims with reference to the accompanying drawings, all of which form part of this specification. It should be clearly understood, however, that the drawings are for illustrative and explanatory purposes only and are not intended as a description of the boundaries of the subject matter described. As used in the specification and claims, the singular form of "a" and "said" includes plural references unless the context clearly indicates otherwise. 4799/TR BRIEF DESCRIPTION OF THE DRAWINGS The structure and advantages of various applications can be understood by referring to the figures below. Figure 1 is a block diagram showing a basic communication system. Figure 2 is a flow diagram showing an example spectrum management method. Figure 3 is a schematic drawing showing an example pulse radar signal. Figure 4 is a schematic drawing showing various frame formats depending on whether the application creating the frame detects the JSC communication. Figure 5 is a block diagram of an example detection application estimate. Figure 6A is a schematic drawing showing simple detection scenarios in the absence of objects to be detected. Figure 68 is a schematic drawing showing simple detection scenarios in the presence of objects to be detected. Figure 7 is a schematic drawing showing a sensing scenario with multiple devices communicating and sensing over line of sight. Figure 8 is a schematic drawing showing a sensing scenario with multiple devices communicating and sensing without line of sight. Figure 9 is a flow diagram illustrating an exemplary method for detecting signal detection and/or identification. Figure 10 is a flow diagram illustrating an exemplary method for training a module for sensing signal detection and/or identification. Fig. 11 is a block diagram showing an exemplary apparatus for detecting signal and/or sensing identification. Fig. 12 is a block diagram showing an exemplary apparatus for training a module for detecting signal and/or sensing identification. DETAILED DESCRIPTION Hereinafter, for the purpose of explanation, "end, top", "bottom", "right, left", "vertical", "horizontal", "top" will be associated with the subject explained, as directed. However, unless expressly stated otherwise, it should be understood that the subject matter described may take various alternative variations and sequences of steps. It should be understood that certain devices and processes shown in the accompanying drawings and described in the following description are simply examples or aspects of the subject matter described. Therefore, specific dimensions and other physical characteristics relevant to embodiments or aspects described herein should not be considered limiting unless otherwise stated. Unless expressly identified as such, no aspect, component, element, structure, action, step, function, instruction and/or the like used should be constructed as critical or essential. Additionally, as used herein, "a" is intended to include one or more elements and may be used interchangeably with "one or more" and "at least one." Additionally, as used herein, "specific" is intended to include one or more elements (e.g., related elements, unrelated elements, a combination of related and unrelated elements, and/or the like) and does not include "one or more" or "at least one". Can be used interchangeably with . In cases where only one element is intended, the term "one piece" or similar language is used. Additionally, as used, the terms "having," "having," "having" or similar expressions are intended to be open-ended terms. Additionally, "based on" means "based at least in part" unless expressly stated otherwise. Although the techniques mentioned above may be successful in some cases, they do not provide complete information or awareness about the future spectrum or its use. Meanwhile, wireless sensing is gaining popularity in commercial devices for environmental monitoring, health monitoring, and many other applications. Military uses of wireless sensing, such as radar, have always been popular. The presence or absence of a signal or knowledge of its properties is generally not sufficient to describe or predict behavioral implementation, and therefore, this information is not very useful for sensing applications. In particular, sensing applications typically generate signals that may have a different pattern than some communications applications. For example, most wireless sensing applications produce periodic signal transmissions of varying periodicity. Effective planning and management of these transmissions will result in less wastage of spectrum and power. Therefore, if a device can determine which sensing application is using the spectrum, it can schedule its own transmissions to fall on empty slots or, if appropriate, use those transmissions for its own sensing. Other exemplary uses of the information are also possible, such as spectrum estimation, which can be provided to other devices in the field regardless of whether the spectrum is shared to ensure efficient resource access. In summary, it would probably be desirable to provide an approach to detect application identification and prediction to improve resource access and utilization without a single central resource management element. In this way, some of the multiple sensing and communication devices coexisting can operate more effectively. Figure 1 shows an example wireless system (WIS) where Tx represents a transmitter and Rx represents a receiver of the wireless signal. Said transmitter (Tx) may transmit a signal to the receiver (Rx) or a group of receivers or broadcast a signal over an interface (Itf). The interface can be any wireless interface. The interface may be specified through the resources available for transmission and reception by said transmitter (Tx) and receiver (Rx). Such resources may be defined in one or more (or all) of the time domain, frequency domain, code domain, and space domain. It should be noted that, in general, both "transmitter" and "receiver" can be integrated in the same device. In other words, the Tx and Rx devices in Figure 1 may also include the functionality of Rx and Tx respectively. The present disclosure is not limited to any particular transmitter (Tx), receiver (Rx) and/or interface (Itf) implementation. However, it can be easily applied to some existing communication systems and extensions of such systems or to new communication systems. Examples may be existing communications systems, such as 5G New Radio (NR) and/or in their current or future versions. The wireless signal does not necessarily need to be a communication signal, as it does not need to perform human or machine communication. In particular, it may be a detection signal, such as a radar signal, or any other type of wireless signal output from a sensing device, such as an audio signal or some signal reporting detection results to another device(s). For example, changes to the IEEE Sensing area may include support for wireless sensing in WLAN networks. Certain implementations can be used to improve the performance of devices conforming to this standard, for example to reduce the amount of redundant detection signals in an area or in a network. Fifth-generation (5G) New Radio (NR) standard or 6G standards may also implement wireless sensing as part of future cellular communications networks. Part of the current explanation may be that it helps estimate free spaces in licensed-exempt spectrum during opportunistic spectrum use, where most wireless sensing is expected to occur. The present description is also applicable to communication technologies under long-term development (LTE)/LTE-unlicensed (LTE-U). As mentioned above, spectrum awareness is an important part of cognitive radio (CR). Some embodiments of the present disclosure support the identification of sensing transmissions and thus may benefit from the present disclosure. The current description can also be applied to low-power wide area network (LPWAN) technologies as it helps to improve power efficiency by reducing the number of redundancy detection transmissions. Therefore, Wize is related to LPWAN standards such as ZigBee, NarrowBand IoT and LoRaWAN. In general, some applications may use high frequencies or millimeter waves (mm waves) because spectrum availability and propagation characteristics are suitable for high-resolution wireless sensing. Can be used to manage resources for wireless sensing. The Tx and Rx devices given in Figure 2 may also include the functionality of Rx and Tx respectively. The said transmitter (Tx) and receiver (RX) can be implemented in any device such as a base station (eNB, AP) or terminal (UE, STA) or any other entity of the wireless system (WiS). A device such as a base station, access point, or terminal can implement both RX and Tx. The present disclosure is not limited to any particular transmitter (Tx), receiver (Rx) and or interface 200 implementation. However, it can be easily applied to some existing communication systems and extensions of such systems or to new communication systems. Examples of existing communications systems may be IEEE 802.11-based systems, such as 5G New Radio (NR) in current or future versions, or the recently reviewed IEEE 802.11be or similar. Sensing applications signals may also be embedded within resources provided by one or more or known systems, such as some IEEE 802.11 standards or possible specific extensions thereof, to support sensing applications. One of the applications of defining the sensing application is spectrum occupancy. In general, spectrum occupancy can be based on the use of previous occupancies arranged on time-frequency grids as well as spatial correlations of such occupancies to predict future spectrum occupancy states. This can be achieved, for example, by training some trainable models 4799/TR to predict occupancy probability for time-frequency occupancy grids. It is stated that machine learning or deep learning and classifier models can be trained on real-world data obtained through measurements and then easily used for spectrum estimation. Deep learning refers to trainable models with multiple layers. Machine learning refers to trainable models with any structure, including layer structure that implements one or more layers. Models can be embedded in functional and/or physical modules. Accordingly, when a trainable module is referred to here, we mean a functional module that implements a trainable model such as machine learning or deep learning or other types of models. Such models, when used, are typically pre-trained (in a training step) and are not modified (updated) during operation. However, the present disclosure is not limited to any particular spectrum occupancy estimate. It can also be accomplished without using trained modules using statistical models or other models. One non-limiting application of spectrum occupancy estimation is for radio resource management and/or spectrum allocation purposes in some wireless networks, such as mobile communications networks (e.g., 4G, 5G, 6G, or similar) or wireless local area networks (e.g., WLAN). In general, spectrum occupancy estimation can be used in cognitive radio systems. Figure 2 shows an example use of spectrum sensing in cognitive radio. The radio environment 210 may be formed by a plurality of BSs and UEs and/or other elements, such as shown in Figure 1 . One or more elements may then perform spectrum sensing 220 . Spectrum sensing is a procedure in which any device, before transmitting a signal, lists (detects) whether another device is already transmitting. Such an approach can be particularly useful in portions of spectrum that can be used by different systems, such as WLANs as well as portions of unincorporated spectrum that can be used by cellular systems or other systems. If the transmission band is determined to be usable in step 220, it can be decided to use the spectrum for transmission in step 240. If the spectrum is not available in step 220, then a different spectrum can be selected for use in step 230. Thus, the spectrum can be shared efficiently 250 and the spectrum usage information can also be evaluated by the radio environment 210. Here, spectrum occupancy estimation can be performed by the radio medium to select the spectrum to be sensed based on the result of the estimation. Spectrum occupancy estimation can use neural networks such as CNN or LSTM (recurrent networks in general) or similar. 4799/EN Future wireless devices will use sensors that can detect to support communications applications and/or enable a wide variety of other applications such as fully immersive extended reality, enabling smart environments, enhancing health-related applications through non-invasive testing and vital sign monitoring, and much more, improving quality of life. It is expected to increase or sometimes only wireless sensors. Wireless sensing applications may require periodic or continuous sensing transmissions. However, allowing all sensing devices to transmit their own sensing transmissions may reduce spectral efficiency and degrade the performance of networks operating in license-exempt bands. Additionally, because sensing transmissions are periodic, they are more likely to cause interference with communications transmissions if they are opportunistically scheduled or have opportunistic channel access mechanisms. This problem can be solved by including detection-aware channel access and detection coordination protocols in the standards. However, these will only enable communication and coordination for detection and devices within the same network. At the same time, this will increase control signaling overhead and complexity. Wireless communications trends are moving towards decentralized and minimally coordinated networks, with more wireless networks coexisting in the same area. Therefore, methods to identify and predict future sensing transmissions, or the act of identifying and predicting them before transmission, are required in standards. This allows devices to better allocate their resources and plan their transmissions. Wireless sensing is the process of obtaining information or awareness about the environment through measurements on received (e.g. reflected or directly received) electromagnetic signals. In this definition, processes such as spectrum/channel sensing, radar, common radar and communication, WLAN sensing and other methods can be considered wireless sensing methods. Most wireless sensing methods have periodic or continuous transmission patterns. An example would be a radar in which a continuous signal or periodic pulses are transmitted. Another example would be channel state information (CSI)-based WLAN detection, where packets are transmitted at specific intervals. Sensing transmissions may have a specific frame design or transmission mechanism that is specific to some sensing applications and may vary based on sensing application requirements and environmental conditions. Periodicity is an important condition for the detection of applications, since interruptions in the periodicity of transmitted/received signals due to interference from other devices, failure to schedule transmissions, or inability to access the channel using 4799/TR channel access protocols can cause interruptions in measurements. This can result in false alarms, missed detections, reduced resolution of detected information, and overall performance degradation of the detection application. Depending on the application, this can have serious financial consequences or life risks. Signal identification, devices, bandwidth, spectrogram image, etc. Wireless technology (LTE, 5G, etc.) based on some signal characteristics such as waveform, modulation type, etc. It allows defining some signal properties such as There are some applications where certain properties are defined for synchronization or authentication purposes. Also known is spectrum sensing, which is used to determine the spectrum occupancy status of primary users. However, it may require continuous spectrum detection. Alternatively, spectrum estimation techniques can be used to save time, energy, and computational overheads required by spectrum sensing. Although the techniques mentioned above have been successful in numerous applications, they do not provide complete information about future and current sensing applications. These applications within a network can be predicted using upper layer information. However, the problem arises when the spectrum is shared between different networks or devices. This is because the access point (AP) I base station (BS) / other devices will not be aware of these transmissions and the transmissions may not be suitable for the detection method of the AP / BS / other device or may cause interference. In this case, spectrum occupancy estimation techniques can be used to estimate spectrum occupancy. In future wireless communications, there may be several wireless sensing applications. In general, sensing applications use continuous signals (at certain intervals) which can greatly degrade spectral efficiency. This is especially the case when multiple sensing applications/devices are used simultaneously and/or in the same area. Certain embodiments and exemplary embodiments provide methods and apparatus for identifying and predicting sensing applications based on certain characteristics of received signals. If the sensing application is known, other devices can schedule their transmissions accordingly or use existing sensing transmissions for their own sensing applications without interfering with other applications. Thus, spectrum, power and other resources can be used efficiently. 4799/EN These are the two aspects of the current disclosure: detection application identification and detection application prediction. The first step in sensing application definition is to determine whether the signal is a communication signal or a detection signal. To describe this, periodicity and frame structure, along with other properties, can be used. These are explained in detail below. Detecting application type: Periodicity Detection signals can be sent at a certain rate or periodicity. Communication signals, on the other hand, are typically more random and do not have any significant periodicity (due to the nature of programming they may be periodic for 2-3 packets, but will not be periodic in the long run). An example detection signal is shown in Figure 3. There are a variety of radar waveforms available. The most popular are pulsed radar and frequency modulated continuous wave (FMCW) civil radar. In pulsed radar, as shown in Figure 3, the signal is "on" (or transmitting) for a period of the total frame time and is "off" (not transmitting) the remainder of the time. The duration of the signal "on" or "off" (duty cycle) provides a balance between range and range resolution. Longer "on" periods (wider pulse width) provide better range but poor resolution and vice versa. FMCW pulses are continuous (have no off period) and can also detect relative speed (pulsed radars cannot do this without some modifications). In order to decide whether the signal is a detection application signal or a communication application signal, it may be sufficient to determine, for example, whether the signal is continuous and has a definite waveform or whether the signal is a periodic signal, without determining the period of the signal. However, it should be noted that the present description is not limited to this and the period can also be defined, and the definition can contribute to the reliability of the detection process and can also be used to further define the detection application. Determining the application type: Frame structure 4799/EN Based on the detection application requirements, different frame structures can be used for wireless sensing. Sensing transmissions may have a frame structure known in some wireless communications standards. This frame structure can be similar to a data transmission frame structure, except that it has more suitable pilots or training arrays that can be used for sensing and can also support data transmissions, as shown in the second and third frame structures in Figure 4. Specifically, Figure 4 shows an example communication containing, in the first line, a preamble section, a DATA section (where payload or higher layer signaling can be transmitted), and two training sequence (TRN SEQ) sections that can be used for channel estimation and/or synchronization or the like. Shows the transmission frame structure. The second line shows an example detection transmission frame structure. As can be seen, no data is transmitted in this frame structure. Instead of the mentioned training sequences (TRN SEQ), detection sequences (SENS SEO) are provided. The third line uses a joint sensing and communications (JSC) transmission frame structure, which is similar to the detection frames structure, but also includes the DATA section as shown in the communication frame structure. Note, however, that these three sample frame structures are only simplified examples, and that the actual frame structure has more parts (areas) that carry various preamble parts, training parts, detection parts, data parts, and possibly also signal parts or other parts (such as filler parts). It should be noted that it may contain Additionally, devices that do not comply with wireless communications standards may use various frame designs that may differ from the frame designs of data transmissions. Detecting the type of application: other features Any properties that are different for communication and detection signals, as well as periodicity and frame structure, can be used to describe them or contribute to the identification. For example, a waveform may also be characteristic for a variety of different standards and may also differ in some cases for detection and communication systems. In summary, according to some embodiments, as a first step, it can be determined whether a signal is a signal generated by a sensing application or a signal generated by a communication application. In some example 4799/TR applications, sensing may also distinguish between communication, JSC, and sensing applications. Such a distinction can help define the resource usage model and identify the resource application or standard or device. If the signal is a detection signal, the second step (after determining whether the signal is a communications application or a signal generated by a sensing application) is to use signal properties – periodicity, bandwidth, detection time, and/or others – to identify the sensing application to which the signal belongs. it could be. Some of these features are explained in detail below. Defining the application: Periodicity Different sensing applications may have different predefined or predetermined periodicities. As noted above, periodicity can include the time interval when the signal is transmitted (ON) and the time interval when the signal is not transmitted (OFF). Periodicity corresponds to the transmission rate of the detection transmission and corresponds to the time resolution of the measurements (sensing task). In general, quickly changing actions! Applications to detect movements/objects require a higher periodicity or packet rate and vice versa. A lower packet (wave packet, pulse) speed on the radar may be a factor increasing the maximum range. This may be because the radar is listening for reflections between packets. The higher the packet rate, the less time the receiver may have to listen for reflections in some applications. Any resulting reflection may be affected by noise due to transmitted signals. Reflections from more distant objects may take longer to reach the receiver. If another civil war or radar signal is transmitted at this point, it may be difficult to decide whether there is a reflected signal and which transmitted signal the reflected signal belongs to. Also, the maximum range can be attributed to the transmission power. For example, for Wi-Fi sensing applications using CSI or received signal strength (RSS) or received signal strength indicator (RSSI), 100 packets per second (pkts/s) can be used for gesture recognition, and 5-packets per second (pkts/s) can be used for sleep monitoring (location and breathing rate). 20 pkts/s can be used. 4799/EN For example, for gesture recognition, models for specific types of movements can be derived from these measurements and used to detect the same movements in new measurements. Alternatively, a learning algorithm can be applied to cluster or classify measurements into specific groups/actions. The packet ratio here can determine the measurement resolution. If the action to be detected is a rapidly changing action, higher resolution may be required. In summary, sensing application identification can be achieved based on signal periodicity alone or in combination with other features. Description of the application: Bandwidth High bandwidth may be required for better range resolution and fine-grained CSl information, which may be required for special sensing applications. For example, applications such as high-resolution wireless imaging or detecting small objects and movements require greater bandwidth. For some wireless sensing processes, bandwidth can also affect the number and placement of training/sensing array subcarriers. In some applications, different subcarrier frequencies are affected differently by different movements. Therefore, a larger bandwidth can allow more information to be collected. Conversely, if the detection method is based on another measurement such as RSS/RSSI, the device may only take a measurement on the carrier frequency and bandwidth may not matter much. Intuitively, these sensing applications will use smaller bandwidth transmissions to save, for example, power and perhaps implementation costs. Therefore, the bandwidth occupied by a sensing application can distinguish between sensing applications for different object sizes as well as applications that do not detect objects at all, or the like. Application Identification: detecting duration and start/end times It may be unreasonable for Wi-Fi devices to constantly detect the environment if the application does not require it or has expired. For example, a fully immersive Video game may only last a few hours at a certain time of day. However, evloda monitoring applications will continue 24/7 or until the application is closed by the user. Similarly, sleep tracking apps will only work during night hours. Monitoring vital signs during exercise only takes a few hours and 4799/TR will most likely be planned. Therefore, the detection signal transmission time and start/end times can also indicate the detection application. Identifying a sensing application: other examples In summary, a sensing application can be defined based on one or more or all of the characteristics mentioned above. Other features such as waveform, time of day, day of the week, or day of the year may be used. When referring to the identification of a sensing application, what is meant may be to distinguish, for example, between a radar signal and medical/healthcare measurement reporting, or between other sensing applications. For the purposes of an example, the environment for sensing application definition may be local networks or area networks, and these may be home networks or networks spanning several buildings and a street in the neighborhood. However, the present disclosure is not limited to any particular scale. For example, the street is active and has autonomous vehicles passing by regularly. Autonomous vehicles perform radar detection with FMCW signals. Additionally, roadside units for smart transportation systems also detect roads for vehicle and pedestrian traffic. Buildings include private residences and offices, public/government buildings, schools, hotels, etc. it could be. Buildings, home/office monitoring sensors, gesture recognition sensors, sleep monitoring sensors, etc. It includes a large number of WLAN devices that communicate or otherwise detect various applications, as well as associated wireless environmental sensors such as Below, some home monitoring is explained in detail on the example of a 'private residence' with gesture recognition and sleep monitoring sensors. There are also many communication devices in all frequency bands. To summarize, the detection elements and sample parameters for the first scenario could be as desired below: (Mobile) Autonomous Vehicles: 0 Single transceiver FMCW radar waveform Frequency = 6 GHz Bandwidth (BW) = 40 MHz Periodicity = 100 pktsi's 4799/TR (Fixed) Roadside Unit 0 Single transceiver o FMCW radar waveform 0 Frequency = 60 GHz 0 Periodicity = 100 pkts/s Home Monitoring: 0 2 APs and 5 sensing/responsive nodes 0 Wi-Fi Physical Layer Protocol Data Unit (PPDU) frame format 0 Frequency = 2.4 GHz 0 Periodicity = 100 pktsi's Sleep Monitoring 0 1 transmitter and 1 receiver Wi-Fi PPDU Frequency = 60 GHz Periodicity = 10 pkts/s Gesture Recognition 1 AP and 5 sensing/responsive nodes FMCW radar waveform Frequency = 60 GHz Periodicity = 100 pktsi's As can be seen in the examples mentioned above, there are one or more features that allow discrimination between applications. It is noted that the above example is fictional and measurement values may change, which may make it easier or more difficult to make the distinction accurately. It should be noted that these five features (number of transmitters and receivers, frame structure/waveform, carrier frequency, bandwidth, and periodicity) are only examples here. In general, depending on the resolution desired for sensing application identification, one or more of these five features and/or any other feature that can distinguish between (or contribute to the distinction between) sensing applications may be used. In exemplary and non-limiting embodiments, features may be determined as noted above. Regarding the number of transmitters and/or receivers, these can be identified by RF fingerprinting or other identification techniques used in PHY authentication to identify transmitting devices or AP/STA. Below is an example where the detection application definition is used for its own detection purposes. For example, sensing and communication devices initially transmit their signals (for the sake of simplicity, without interference - in the presence of interference - which is normal, the results may be affected by the identification accuracy). The roadside unit is a critical detection application and is therefore constantly transmitting signals. Gesture recognition application is not critical and is an optional application. When the said gesture recognition application is launched, the detection device will first take snapshots of the spectrum in the required frequency band (60 GHz) and analyze them. You will see that there are two periodic broadcasts, one from the sleep monitoring device and one from the roadside unit. Other signals will be discarded and said gesture recognition device will further analyze these signals and compare their parameters with its own required transmission parameters. Of the two transmissions, the transmissions of the roadside unit are more suitable for gesture recognition application as they apply a similar waveform (FMCW radar type), frequency and so on. Therefore, instead of transmitting its own signals, the gesture recognition application can use the signals of the roadside unit. This is a possibility because some characteristics (waveform, frequency, periodicity) of both applications are similar. Only bandwidth requirements may differ. For example, the bandwidth of a roadside unit may typically be higher than with gesture recognition. Therefore, the bandwidth of the roadside unit signal will be sufficient for gesture recognition. As a result, spectrum can be used more efficiently. Alternatively, without comparing its required and available signal parameters, the gesture recognition application (module) in question may enter a training phase in which it uses all the sensing transmissions it can handle (according to hardware availability) and selects the one that results in the least errors or best meets the performance criteria of the sensing application. Now, in this example, to ensure smooth continuity of sensing execution, the gesture recognition device may need to ensure that sensing transmissions continue. This can be achieved by defining the 4799/TR detection application. From its dictionary of sensing signal parameters and applications, or through learning techniques, said gesture recognition device may know or determine that the sensing signals it uses belong to a critical sensing application, and therefore sensing transmissions will proceed. When referring to the glossary of detection signal parameters and applications, it means that the application description mentioned above and the frequency, bandwidth, waveform, etc. It refers to a pre-stored list in which the values of properties such as are captured. Identification of the detection application can be achieved through a statistical algorithm. Alternatively, a machine learning or deep learning algorithm can be used to take a signal as input and classify the sensing application that generated said signal as one of predefined sensing applications. Hybrid approaches can be considered in which feature values, rather than the signal directly, are fed into a trained (trainable) module to extract the sensing application identification (classify the sensing application that generated the signal). Additionally, similar applications and sample applications can be used for signal intelligence. In other words, enemies' practices can be detected and some precautions can be taken before they happen. For example, if the detection process learns that the enemies want to sense the movements of their soldiers or people in general, it is obvious that they are trying to spy on them. Thus, some precautions can be taken to prevent possible damage. The current description can also be used as a predictive control mechanism. For example, the AP can determine the detection application and check whether it matches the predicted signals. This type of prior prediction can be used, for example, to use spectrum more efficiently, resource allocation, etc. It may be available for any purpose on cognitive radio. To estimate the spectrum, the methods shown in Aygi'JI et al. can be used. On the other hand, if the sensing application is predicted, the periodicity of the signal can be known, which will provide information about the future spectrum. Thus, it can be checked whether the predicted spectrum (e.g. estimated by the Aygül et al. method) matches the predicted spectrum realized by the periodicity-based technique. As another example, in the environment described above, another sensing device or application may enter the area. First, it can describe and predict current sensing practices. If there are similar or identical applications, you can use these detection signals for the 4799/TR detection application and there is no need to generate new signals. As a result, the required power is reduced and interference can also be reduced. Additionally, spectrum can be used more efficiently. In summary, as shown above, the present description is in no way limited to the spectrum (occupancy) estimate mentioned at the beginning. However, spectrum estimation is one of the applications to which the current sensing application definition can be applied, as provided in an example application below. Sensing application estimation can be used for spectrum occupancy estimation (but also for other purposes such as surveillance, threat and espionage detection, or similar). Spectrum estimation Detection application estimation can be as important as detection application identification. It has various advantages which are explained in detail in the following section. There are two steps to detection application estimation. The first step is to estimate spectrum occupancy. For this, techniques used in "Aygül, M. A., Nazzal, M., Sağlam, M. I., da Costa, D. 8., Ates, H. F., and hereinafter referred to as Aygül et al.) can be used, but the applicable techniques are not limited to the specific techniques described here. .If the spectrum is empty (available), it means there will be no detection application in the future. If it is full (occupied), it is necessary to distinguish between the communication signal given above and the detection signal whether it is a detection signal or any other signal present in the spectrum. The features and techniques mentioned above can be used for this purpose. If it is a detection signal, its application (the application that creates the detection signal) is defined. To predict a detection application, the techniques and features defined for the detection application can be used. The block diagram for this is shown in Figure 5. Specifically, the input to the trained module 520 is a measurement of the spectrum 510. In this example plot the spectrum 510 is represented as a two-dimensional time-frequency grid as is usual for time-frequency 4799/TR source-based systems such as Orthogonal Frequency Division Modulation (or Multiplexing) or various other Frequency Division Multiplexing (FDM) approaches in general. is represented. The trained module 520 in this example embodiment is a learning module, such as a machine learning or deep learning module, trained for the specific task of spectrum occupancy detection. The result 530 of the trained module is an estimate of the spectrum occupancy for a given portion of the spectrum for a given time instance or for a number of subsequent time instances. Based on this result in evaluation step 540, if the spectrum is not full (in this example binary zero is represented by "0"), no further action is taken. If the spectrum is full (it is decided whether it is this or a detection signal). If the signal filling the spectrum is a communication signal in the evaluation step 550, no further processing is required, at least in the context of the present explanation, which focuses here on the definition of the sensing application. The signal filling the spectrum is If it is a detection signal in the evaluation step 550, the detection application that produces this signal is also defined. Instead of first finding the spectrum occupancy and then finding the detection application, both can be performed by a learning algorithm such as deep learning. The details of a common spectrum occupancy and detection application are given below. There are two stages: training and testing. In the training phase, the data set is collected first. This data set contains the binary occupancies of these measurements and the class of the detection application. The interference phase (also called the production phase) corresponds to the actual operation of the module trained for sensing application identification and/or prediction. It should be noted that the present disclosure is not limited to any particular network modularity. For example, recognition of the application type (detection/communication) can be considered to be accomplished in two steps as described above. This can be accomplished by two different and separately trained modules (for example, two different neural networks or Long Short Term Memory or similar). It can also be achieved by mixing a machine learning (trained) module and a statistical algorithm. However, it is also possible to have a module trained to perform both, i.e. to decide based on the input whether it is a communication signal or one of a set of predefined detection applications. Below the second approach will be discussed as an example. 4799/TR Before training, the training data set can be obtained in various ways. For example, binary spectrum occupancies can be obtained by thresholding power spectral density (PSD) measurements (or other power or amplitude-related measurements) according to a certain threshold. In some applications, the threshold is adjusted according to the strength of the thermal noise. The noise level can be determined by a measurement or estimated as known in the art. Additionally, a margin of 3 dB may be considered to determine the final threshold value to account for unforeseen effects and variations. If the PSD measured in a certain frequency band is above (exceeds) this threshold, it is reported that this frequency band is occupied. On the other hand, it is reported (determined) in the frequency band. After collecting the measurements, the input data set can be obtained in time, frequency and space as described in Aygül et al. Two different tasks can be learned for output. The first is spectrum occupancy estimation and the second is detection application. The spectrum occupancy estimate can be represented by binary values with a unit. In other words, for each part of the spectrum, a binary value indicates whether that part of the spectrum is occupied or not. For sensing application definition, several sensing applications will be possible, so the number of output units should be the same as the number of sensing applications to represent them all. It should be noted here that, assuming the spectrum is full (or if there is any prior information), only sensing applications can be used as an output dataset. Otherwise, spectrum occupancy can be used as a class of detection applications. Once a data set is obtained, the machine can be trained based on said input and output data set. Later, in the testing phase, spectrum occupancy estimation and its application can be found. It should be noted here that the spectrum can be estimated instead of the spectrum occupancy. In this case, the output unit for spectrum occupancy will be changed to spectrum estimation, for example, from binary to digital. Figure 6A and Figure BB show an illustration of a general sensing application. Here, the wireless sensor (WS) transmitter (Tx) 610 is the wireless sensor transmitter and the WS receiver (RX) 690 is the wireless sensor receiver. WS Tx and WS Rx are synchronized and can coordinate with each other via wireless signaling or wired connection. WS Tx (610) and WS Rx (690) are located in a limited area such as room (600). Concentric circle segments show the electromagnetic field (wireless detection signal) generated by said WS Tx 610. As can be seen in the figure, the WS Rx (690) is within the range of the 10 4799/TR wireless detection signal. As explained above, the wireless detection signal can be a detection pulse or a continuous signal such as radar or an audio signal. However, it may also be possible to report a regular measurement in package form or similar. Figure 6A is just an example configuration of wireless sensing hardware. Other configurations include multiple transmitters and/or multiple receivers, transceivers, etc. may contain. The term transceiver here refers to a receiver or a transmitter or a combination of both. Initially, the WS Tx 610 can transmit detection signals when there is nothing to detect (no object within range of the monitored object size), as shown in Figure 6A. The WS Rx (690) receives these detection signals and takes measurements, such as CSI or any signal strength indicating measurements. Then, when there is something to detect, shown as stick man 650 in Figure GB, changes in measurement indicate this. In other words, the presence of the object 650 changes the channel and therefore the received signal detectable at WS Rx 690. It is noted that this is an example with only the transmitter and receiver (WS) working together. However, the present disclosure is not limited to this, and instead of detecting the transmitted signal from a particular transmitter located at a location other than the location of the receiver, some sensing applications may rely on a transceiver containing transmitter and receiver positioned together, such as a pulse radar, in which case the detected signal is It is a signal reflected from the detected object. Figure 7 and Figure 8 show other different sample detection scenarios. There may be different devices sensing and communicating in the area, only sensing devices denoted as WS Tx and WS Rx. Alternatively or additionally, detection may be performed between APs (or other network controllers/coordinators) and stations (STAs) or between STAs and other STAs. At the same time, other network devices can communicate. Detection devices can be line-of-sight (LOS), as shown in Figure 7, or non-line-of-sight (nLOS), as shown in Figure 8. For example, in Figure 7, a WS Tx transmits a first detection signal (dotted line). An AP1 transmits another detection signal (solid line). STA1 can receive the detection signal from APl, while WS RX can receive the detection signal from WS Tx. However, STA1 can also use the detection signal from WS Tx and WS Rx can also receive the detection signal from AP1 in some scenarios. What is meant by receiving here is to perceive it as present and possibly do further processing, in other words, not just receive it as part of the noise. In addition, AP2 and STA2 are in communication with each other, in other words, they are exchanging communication signals. In this simplified example, all these devices are in LoS. Some or all of these devices may operate in the same or at least partially overlapping spectrum. As shown in the figures, said detection signals may have different periodicity (indicated by different intensity of concentric circle segments representing the detection signal). In this case, it may be useful to define detection applications without direct communication or coordination with another wireless device. In response to detection of a sensing signal, devices can use signals available to them rather than transmitting their own sensing signals. For example, when STA1 detects that the WS Tx is transmitting a detection signal, it may use said detection signal in addition or as an alternative to the detection signal from AP1. It is even possible that the AP1 detects the sensing signal from the WS Tx and stops transmitting its own sensing signal, since one sensing signal may be sufficient. Or vice versa, the WS Tx detects that AP1 is transmitting a detection signal and said detection signal stops its transmission. As will be apparent to those skilled in the art, various coordination and adaptation practices of the sensing environment can be achieved once sensing applications (signals from sensing applications) are detected in an area. Figure 8 shows a scenario where there is not necessarily a LOS between some or all of the devices. In the figure, none of the STA3, STA4 and STA5 devices actually in the detection area have a LoS to other devices. However, the detection signal can be received in detection signal receiving devices. For example, STA5 does not have LoS to STA3 and STA4, but can still receive detection signals. Therefore, STA5 may decide to turn off its own signal and use the detection signal of STA4 or STA3 or similar. Using the sensing application identification and prediction technique described above, devices can find out which sensing application the signals are for, predict their duration and future spectrum usage, and schedule their own signals to be free of interference (resource allocation) or use these signals for their own sensing applications. Some specific examples are given in the explanation above. However, the present disclosure is not limited to these examples. On the contrary, variations and modifications 4799/TR may be advantageous for some scenarios. For example, any properties that are different for sensing and communication signals, such as frame structure, periodicity, resolution, RSS/RSSI values, or some of the aspects of sensing that will be defined in future standards, such as periodic channel access mechanisms, fallback behavior, special sensing sequences or waveforms, or the like. can be used to distinguish features. RSS/RSSI can be used instead of PSD to detect spectrum occupancy and/or to determine whether the signal is a communications signal or a detection signal, or to identify the sensing application that generated the signal in question. For example, RSS/RSSI measurement can be used effortlessly on most communication devices and can give a rough indication of user activity, i.e. spectrum usage. The term "user" here refers more broadly to a specific application running on a device. Some embodiments of the present disclosure may be used for military defense purposes or for civilian applications such as home surveillance or home appliances or entertainment. For example, one can learn which detection applications are used by other parties. In addition, the location of the sensors can be found based on the PSD value and application requirements. For example, if the detection application is learned (this means that a detection application is known and the trained module is trained to distinguish it), it is possible to have knowledge of the RSS/RSSI value, because different applications require different RSS/RSSI values. RSS/RSSI depends on the distance between the transmitter and receiver. Therefore, it is possible to at least approximately estimate the location or relative position of the sensing device (e.g., transmitter). There are various deep learning (DL) techniques that can be used in the present disclosure for spectrum occupancy estimation and sensing application detection, identification and/or prediction. Examples include multilayer perceptron (lVILP), recurrent neural networks (RNN) such as long short-term memory (LSTM), or convolutional neural network (CNN) or the like. Statistical methods can also be used instead of DL techniques. Detection signal identification and estimation need not be done with ML and DL techniques alone. Sensing application identification and estimation can be done jointly or separately in various domains such as time, frequency, space or code domain. The advent of DL is to alleviate the need for human-based feature extraction/processing/engineering. Therefore, it is possible to apply the above method directly to real-world measurements (PSD or RSS/RSSI) without any feature extraction. Conversely, the “deep” nature of the machine learning used can be reduced at the expense of adding more feature extraction. In an extreme case, DL can be avoided altogether and ML can be used if it is possible to obtain different features. While deep generally means using more than one processing layer, machine learning is also possible with a single layer. Other types of augmentation may still be considered in training and designing the trained module, rather than amplifying the received signal in the form of a vector to be fed to the classifier. For example, they can be shaped into another two-dimensional (2D) grid (e.g., time-frequency grid). Some effects and advantages One of the advantages of the present disclosure may be the provision of improved spectral efficiency. For sensing applications, there may be no need to generate a new signal as a signal that can be used for sensing may already exist. Therefore, wastage of spectrum can be avoided and the recorded spectrum can be used for something else. Power efficiency can be another advantage of sensing application detection and/or identification. Since additional signals may not need to be transmitted, less power will be wasted. This can be critical for IoT and other power-limited wireless sensing devices. Resource allocation can be facilitated by some of the practices described above. In current WiFi standards, there is no control between networked APs. Therefore, coordination techniques between different APs are required to eliminate interference while using the spectrum efficiently. However, by reducing detection signal transmission based on detection, identification, and/or estimation of the detection signal, interference can also be reduced. Information about the periodicity of detection signals can be used to help estimate spectrum occupancy. If a signal is identified as a detection signal and its periodicity can be determined, future (near future) transmissions can be predicted and used for a more accurate spectrum occupancy estimate. Information about the periodicity of detection signals can be used to help estimate spectrum occupancy. If a signal is identified as a detection signal and its periodicity can be determined, future (near future) transmissions can be predicted and used for a more accurate spectrum occupancy estimate. 4799/TR Learning the detected information can also be achieved. If we know which signals are detection signals, we can process them and obtain information about the environment simply by using transmissions from other devices. Blind signal analysis can be useful for (e.g. military) defense or civilian applications: detection signal information can be used for military reconnaissance. For example, one can learn what other parties want to feel and what they feel. This can be done by analyzing detection signals using blind signal analysis (or other methods). In communication, unless a node is an illegitimate node (such as an eavesdropper), it may not be possible to use someone else's signal, since signals in the communication are typically allocated to individual users. However, in the sensing process, any signal that meets the requirements of the sensing application can be used, as the signals in sensing typically do not carry any private/confidential data or information. Blind signal analysis can also be used for commercial applications: Some applications of the present disclosure include CR and software-defined radio (SDR), public safety radio, spectral efficiency improvement, optimization of operation for various networks sharing spectrum, enabling testing and measurement of communication and sensing devices, or can be used to make it easier. This can be achieved by analyzing detection signals using blind signal analysis (or other methods). Any of the above advantages serve to facilitate the integration of multisensing and joint sensing and communication (JSC) devices into current and next-generation wireless networks. As explained above, a method is developed for defining a sensing application, generally speaking, with respect to an application. The flow diagram of the method is shown in Figure 9. Said method consists of obtaining a received wireless signal (910); and the step of estimating (920) by a trained module the presence of a (generally one or more) detection signals in the received signal, which is a signal generated by a sensing application. Acquisition in this context refers to the process of receiving via the wireless interface in some embodiments of the device, for example, devices that include the front side of the transceiver. However, the devices of the present disclosure may also be separated from the actual transceiver 4799/EN and have an interface solely for communication (to control and obtain/provide received signals to be transmitted). For example, the retrieval operation may be retrieving from such an interface or retrieving from the store/buffer or similar. The trained module can be an ML or a DL module or some other type of trained module. In general, the present disclosure provides that, based on conditions that distinguish between sensing applications according to, for example, some predefined characteristics (such as one or more of those mentioned above, including periodicity, carrier, bandwidth, waveform, etc.), said trained module is a type of It can be replaced with a module that implements a statistical algorithm or another decision algorithm. Combinations of educational modules and some decision-based modules may be considered. In other words, instead of or in combination with the trained (e.g. ML/DL) module, other types of methods can be used for prediction, such as statistical methods or deterministic methods. For example, if a signal is observed to repeat periodically, ML or DL methods may not be necessary to detect the presence of such detection signal. Whether detection has occurred or not can be determined deterministically. Similarly, given a set of predefined properties and their values for specific applications, the identification process can be performed for sensing application identification. Also, in cases where detection applications use specific header information to determine network types, this or other header information can be detected and used deterministically to identify an application. It may be inconvenient to define parameters (properties and values) and application sets for each environment. Therefore, depending on the deployment scenario, the mentioned trained modules may provide better results in more complex scenarios where, for example, deterministic or stochastic discrimination is difficult or complex. Further, the method provided in Figure 9 includes performing 930 wireless reception, transmission, or detection based on the estimated presence of said sensing application. Some examples of such actions taken in response to or based on the forecast include the following sample applications. Based on the presence or absence of the detection signal in a plurality of measurement time samples, the periodicity of the detection signal is determined. This can be achieved by detecting the presence/absence of the detection signal in the received signal and then analyzing the past captured signal. It can also be accomplished by correlating the waveform of a particular sensing application with the received signal, which possibly contains noise and interference from other signals. It can be performed by a trained module. The periodicity of the signal can also be used to improve the spectrum occupancy estimate, for example as an additional input or feedback to the spectrum occupancy estimate. Based on the spectrum occupancy estimate, scheduling of the signal can be performed in a wireless device (STA, AP, WS or similar). For example, if the spectrum occupancy estimate (e.g., based on the periodicity of the identified sensing signal) indicates that a particular portion of a spectrum will be used by the identified sensing application, the wireless device may transmit signals (either communication or sensing, or both) in a different part of the spectrum and/or at a different time. It may decide to transmit (than expected for the detection application). It should be noted that the sensing application identification and/or spectrum occupancy estimation may, in some applications, be performed on the same device that uses the results of the identification. However, there may be applications in which a first (e.g., controller) device performs spectrum occupancy estimation and/or sensing application detection or identification. Such a device can then broadcast the results (or multicast or transmit only) to one or more devices in the field. They can then adapt their behavior (e.g. channel access, detecting signal transmission, or detecting signal reception and evaluation) accordingly. In other words, the detected sensing application can be used to improve spectrum occupancy estimation. Additionally, the detected sensing application features can be used in programming. A device with knowledge of the availability of a particular sensing application (and prediction for future availability in certain resources) can program its data or sensing signal to avoid using the same frequency and/or bandwidth and/or time. Alternatively or additionally, a device having an estimate of the detection signal sent by a particular transmitter may use said detection signal for its own detection, for example, for evaluating the gifts of objects in the field. In unencrypted applications, if the sensing application reports environmental conditions or channel conditions, the reports can be read by multiple devices that detect and predict the presence of such signals. Alternatively or additionally, transmitters of detection signals may stop transmitting detection signals or decide to reduce the periodicity of their own detection signals when they have information about a similar sensing application detected/identified in the same area (e.g. in proximity). 4799/EN In some exemplary embodiments, the sensing application is an application that produces said sensing signal with a pre-configured periodicity or as a continuous transmission signal. Such applications have features that make it possible to easily distinguish them from communication signals (repetitive character and/or possibly the extraction and identification of communication signals can be achieved with some reliability and can be used in any of the example scenarios mentioned above. Periodicity, e.g. by a person (e.g. in health monitoring Devices) ) or can be pre-configured at the sensing application transmitter by the application or similar. It should be noted here that in the case of continuity the continuous signal is continuous and subject to minimum time resolution, for example continuous signals can be a continuously transmitted on/off signal pattern or a sine pattern or similar. - this does not mean that the signal must be constant. The minimum time resolution may depend on the hardware capabilities. In other words, even continuous transmissions typically have a periodic character. Here, the term continuous is used as in the art: continuous means radar transmitted in one direction. Unlike civil signals, there is no discernible gap in the detection signal. The periodicity referred to above is additionally intended for cases where there is a predefined time interval between detection signals (wave packets). Of course, the predefined gap may also have some variations depending on defective equipment or wave propagation conditions (at the receiver). In an exemplary embodiment, the sensing application is one of radar sensing, wireless sensing, wireless local area sensing, channel status information sensing. Wireless sensing may include channel or spectrum sensing. Wireless local area detection may include beamforming detection. It should also be noted that the term wireless sensing can be considered more general and includes any type of sensing that uses a wireless signal to detect itself or for (possibly periodic) reporting of sensing results. This type of wireless sensing or wireless local area sensing may also include detection using Wi-Fi technology. This usage includes equipment (using STA or AP), frequency bands, frame designs, CSl, CIR, CFR, RSS, RSSl, etc. It may be in terms of communication-related measurements such as. Some examples of wireless sensing and wireless local area 4799/TR sensing, such as Yongsen Ma, Gang Zhou, and Shuangquan Wang. 2019. WiFi Sensing with Channel State Information: ASurvey.ACM Comput. Surv.52, 3, Article 46 In some exemplary embodiments based on any of the above-mentioned applications or examples, the prediction process further includes identifying the sensing application as one of a plurality of predefined sensing applications (e.g., identifying the application or application type). As mentioned above, detecting the presence of a sensing application and more accurately identifying such an application (or type of application) can be performed together or separately. Here, the expression "identification of application" should be understood broadly. For example, in some applications this may mean distinguishing between a set of predefined application types, such as health and fitness/sleep monitoring on the one hand, and object presence/motion/speed on the other. However, in some applications, a finer distinction can be achieved, such as that exemplified in the applications detailed above. It may be possible to further distinguish gesture detection signals from health or sleep monitoring or reporting signals. In an exemplary embodiment, the sensing application is an application that generates the sensing signal with a preconfigured periodicity. Next, the process of determining the sensing implementation includes estimating the preconfigured periodicity based on the presence of the predicted sensing signal within the received wireless signal; and using pre-configured estimated periodicity to identify the sensing application from among a plurality of pre-defined sensing applications. The identification process can be performed by a trained module (e.g. ML, DL). Detection of presence (and perhaps periodicity) can be accomplished with a deterministic or statistical approach. However, as already mentioned above, other configurations are also possible, including joint detection and identification by one trained module, two separate trained modules, or a pure statistical I deterministic approach. For example, the determination of sensing applications is accomplished based on one or more of the following characteristics: frame structure, bandwidth, sensing time, sensing start time, sensing end time, and waveform. One or more features mentioned differ for at least two of a plurality of predefined sensing applications. It is then possible to make some 4799/TR predictions regarding the identification of the current application. Properties refer to features as exemplified in the above embodiments, such as measurable parameters whose values are different for at least two embodiments. As mentioned above, presence detection and identification can be performed sequentially. In such a case, in an embodiment, said determination of the sensing application as one among a plurality of predefined sensing applications is performed if: - the presence of the sensing signal in the received wireless signal is confirmed by the prediction step; and - not performed if the presence of the detection signal in the received wireless signal is not confirmed by the prediction step. This has already been explained with reference to Figure 5. Specifically, in an exemplary embodiment, obtaining the received wireless signal includes measuring one or more of the power density spectrum, received signal strength, or received signal strength indicator. Said method further includes determining the presence or absence of a communication or sensing signal in the received wireless signal by the trained model. If it is determined that a communication signal or a detection signal is present in the received wireless signal, a prediction is made whether the signal is a detection signal or not. If the signal is a detection signal, the detection application is determined as one of a plurality of predefined detection applications. Although the example explained with reference to Figure 5 shows binary occupancy, it should be noted that a soft value can be used to estimate occupancy determination i' in general. This type of software value can express the reliability of the occupancy estimate. If it is determined that a detection signal does not exist, said method may include omitting the prediction (identification) step. If the signal is a communication signal, the method may further include bypassing the determination of the application identification. Additionally or alternatively, if the signal is a communications signal, said method may include identifying the communications signal according to any method known in the art. 4799/TR As mentioned above, in some applications, the determination of the detection application and the presence estimation are carried out together by the trained module. In some example embodiments, future occupancy of resources may be predicted by the sensing application based on the result of the prediction step. The prediction process can be performed by a trained module, or by deterministically assuming certain periodicity, or by a mixture of both. The estimator can be modified depending on the situation. If there are many sensing applications performing the forwarding process, it may be better to use machine learning. Otherwise, parameter-based estimation may perform better in simpler scenarios with well-distinguishable detection implementations. In all of the above embodiments and examples, performing wireless reception, transmission, or sensing may involve scheduling wireless reception, transmission, or sensing on resources that are anticipated to be unoccupied by the sensing application. Additionally or alternatively, performing wireless reception, transmission or sensing includes deciding whether to transmit a detection signal based on the results of the prediction step (presence/absence and/or identification) and whether to perform detection based on the decision step. Therefore, self-use of the detection signal can be realized based on the detection definition. This type of self-use case uses information about the existence and types of sensing application signals to decide whether to transmit its own sensing signals or use existing signals, as already explained above. Within a network this can be used to reduce sensing traffic by coordinating with other devices to use a reduced or even minimum number of sensing transmissions to serve sensing applications. In addition to the above-mentioned approaches to detecting and identifying the detection application, training methods and devices are being developed for training the module used for detection identification. For example, as shown in Figure 10, a method for training a module to identify a sensing application is being developed. Said method includes obtaining a received wireless signal in step 1010 and obtaining the precise reference in step 1020. The definitive reference in this 4799/TR application can be the decision on the presence or absence of a detection signal when a module is provided to detect the presence or absence. The exact reference in another application may be the identification of the application that generated the detection signal when a module is trained for identification (or for both presence detection and identification). Said method further comprises modulating: one or more of the resulting representation of a received wireless signal (e.g., RSSI or other measurements representative of the signal or its properties), and a desired indication of the presence or absence of a detection signal in the representation, and a desired indication of a sensing application producing the sensing signal. It involves entering both. After entering the training pair, the method includes modifying (1030) at least one parameter of the module according to the input. This type of approach can be repeated for large numbers (possibly large amounts) of training data (training pairs in supervised learning). It is stated that this training method is only an example and that other approaches such as unsupervised learning can also be applied. Although embodiments and examples of the present disclosure are provided in terms of a method above, it should be noted that the corresponding device providing the functionality described by the methods is also provided. For example, a Device is provided to identify a sensing application. Said device may include a processing circuit configured to perform a step according to any of the above-mentioned methods. For example, the processing circuitry may be configured to obtain a received wireless signal and predict the presence of a detection signal in the received signal, where the detection signal is a signal generated by a sensing application. The device may further include a transceiver to perform wireless reception, transmission or sensing based on the estimated presence of said sensing application. As an alternative to said transceiver, the processing circuitry may control an external transceiver to perform wireless reception, transmission or sensing based on the estimated presence of said sensing application. Figure 11 shows an exemplary device 1100 that may implement some embodiments of the present disclosure. For example, the device could be the device that detects application detection or identification. Such a device may include memory 1110 , processing circuitry 1120 , a wireless transceiver 1140 , and possibly a user interface 1130 . The device in question may be, for example, a base station (part) or a terminal / STA or another device as mentioned above. Memory 1110 may store the program executable by processing circuit 1120 to perform the steps of any of the above-mentioned methods. The processing circuitry may include one or more processors and/or other specialized or programmable hardware. The wireless transceiver 1140 may be configured to receive and/or transmit wireless signals. The transceiver 1140 may also include baseband processing that may detect, decode, and interpret data according to some standard or predefined rules. However, this process is not necessary and devices with sensing-only implementations can only implement the underlying one or two protocol layers. For example, the transceiver can be used to make measurements, communicate with other devices such as base stations or terminals. The device 1100 may also include a user interface 1130 for displaying the device's messages or status, or the like, and/or receiving input from a user. A data bus 1101 connects the memory, processing circuitry, wireless transceiver, and user interface. Figure 12 illustrates a module 1160 for obtaining the received signal, a module 1170 for the above-mentioned prediction, and said transceiver for adapting wireless reception, transmission, or detection based on the estimated presence of said sensing application. These modules 1160-1180 can be retrieved from memory and executed by said processing circuit 1120. The examples provided above are not intended to limit the present disclosure. There are many modifications and configurations that can be used additionally or alternatively, as will be briefly explained below. Similarly, a device for training a module to identify a sensing application is being developed, said device providing the module with: the resulting representation of a received wireless signal and a desired indication of the presence or absence of a sensing signal in the representation and the desired indication of a sensing application generating the sensing signal. will enter one or both of its indicators; and a processing circuit configured to modify at least one parameter of the module based on the input. 4799/EN This current description can be used on any device used for wireless sensing. For example, health monitoring, activity classification, gesture recognition, people counting, wall detection, emotion recognition, attention monitoring, keystroke recognition, air drawing, imaging, step counting, speed estimation, sleep detection, traffic monitoring, smoke detection, metal detection, Sign language recognition, humidity estimation, wheat moisture detection, fruit ripeness detection, sneeze detection, etc. Besides these embodiments, embodiments of the present disclosure can be used in JSC technologies. This description can also be used for sensing applications to support communication applications such as obstacle monitoring for beam management. Therefore, devices that can benefit from this invention are electrical kitchen appliances, television sets, smart bus stops, smart office equipment (printers, etc.), lighting systems, WLAN and WiFi devices, etc. such as smart homes/offices/cities/factories/etc. may have devices. Other devices, heart rate monitors, motion detectors, smart watches, etc. There may be independent wireless sensors such as In addition to these applications, the invention can also be used for military services such as enemy sensors, the presence of enemy devices, and what they detect can be learned and some precautions can be taken. This statement is especially useful in network controllers and CR, reconfigurable radio systems, etc. APs, BSs, edge nodes, advanced nodes, etc. for technologies such as It can be used in management devices such as In some embodiments, the detection signal is a continuous or periodic radar signal. In some embodiments, the sensing signal is transmitted by a sensing application that supports wireless sensing, wireless local area sensing, and/or non-invasive medical sensing. In general, wireless sensing is accomplished by measuring some characteristics of a received signal. On the other hand, communication is carried out by detecting the received signal information encoded in the transmitter in question. In communication, some features of the received signal (such as demodulation and decoding) are used to perform the detection process. There are numerous frame designs, waveforms and transmission schemes used for wireless sensing and JSC. Their use depends on the wireless detection method used. For example, wireless detection can be done in one of the following ways: 4799/TR Wireless fingerprinting: Measurements are taken in different locations or with a fixed sensor while performing different actions. These measurements can be preprocessed and stored in a lookup table along with their respective positions/actions. Radar-based detection: Detection signals are transmitted with a certain speed and duty cycle, and the reflections of these signals depend on the object's range, size, relative speed, material, etc. It is processed to learn information such as. In radar applications, waveforms or frame designs with high auto-correlation properties and low peak to average power ratio (PAPR) are preferred due to their high transmit power and very low receive power. Common waveforms are pulsed waveforms or frequency modulated continuous waveforms. Factors based on frame design and transmission mechanism that may affect radar performance; transmit power, duty cycle, pulse repetition frequency, frequency modulation, carrier frequency, auto-correlation properties, bandwidth, signal period, separation angle, beam width, beam scanning speed or the like. Environmental factors that can affect radar performance are the amount of clutter, physical properties of the object (size, material, etc.), location of the object, atmospheric conditions (e.g. humidity), or the like. Spectrum conditions that may affect radar performance are occupancy, number of users, interference, channel conditions (time/frequency selective) or similar. Pattern-based detection: This is similar to wireless fingerprinting except that measurements are not pre-stored in a lookup table, but instead a pattern is extracted from the measurements that represents the action/object to be detected, and said pattern is used to detect the action/object in future measurements. Example frame designs and waveforms are physical protocol data unit (PPDU) packet and orthogonal time frequency domain (OTFS) waveforms in Wi-Fi and OFDM. Factors related to frame design and transmission mechanism that may affect detection performance; detection/training/pilot sequence, placement of these sequences (on which subcarriers), bandwidth, carrier frequency, packet length, packet repetition frequency, separation angle, beam width, beam scanning speed, RF disturbances or the like. Environment-related factors that may affect detection performance; a set of moving objects in the area (ambient stationarity), a set of users/objects to be detected, the nature of the movement/action/activity to be detected (large/small displacement, slow/fast movement or similar), atmospheric conditions, the nature of the object to be detected /user's physical characteristics or similar. Spectrum conditions that may affect radar performance are occupancy, number of users, interference, channel conditions (time/frequency selective) or similar. In general, the present description is not limited to the three types of detection mentioned above. Characteristics of the received signal measured by wireless sensing include, but are not limited to, time-of-flight, RSSI, CSI, or the like. Transmitted signal parameters that affect the performance of wireless sensing include, but are not limited to, correlation capabilities, pilots, beam angle and width (if beamforming occurs), duty cycle, transmission rate, or the like. Wireless sensing, communications, and JSC devices must operate/coexist seamlessly in the same (or at least partially overlapping) frequency bands with maximum efficiency, in terms of spectrum usage, power, or the like, and in terms of sensing and communications performance, efficiency, reliability, detection accuracy, or the like. is desired. This can be facilitated by detecting and/or identifying detection signals as described herein. Applications in software and hardware The methodologies described here can be implemented in a variety of ways, depending on the application. For example, these methodologies can be applied across hardware, operating system, firmware, software, or any combination of both or all of these. For a hardware implementation, any processing circuit that can contain one or more processors can be used. For example, hardware includes application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs). It may include one or more of the processors, controllers, any electronic devices or other electronic circuit units or elements designed to perform the functions described above. 4799/EN When implemented as program code, functions performed by the transmitting apparatus (device) may be stored as one or more instructions or codes in a non-volatile computer-readable storage medium, such as memory 610 or any other type of storage. Computer readable media includes physical computer storage media, which can be any usable media accessible by the computer or, in general, by the processing circuit 620. Such computer-readable media may include RAM, ROM, EEPROM, optical disk storage, magnetic disk storage, semiconductor storage, or other storage devices. Some specific and non-limiting examples include compact disc (CD), CD-ROM, laser disc, optical disc, digital versatile disc (DVD), BIu-ray (BD) disc, or the like. Combinations of different storage media are also possible – in other words, distributed and heterogeneous storage can be used. The above-mentioned embodiments and exemplary embodiments illustrate some non-limiting examples. It is understood that various modifications can be made without departing from the claimed subject. For example, modifications can be made to adapt the examples to new systems and scenarios without departing from the central concept described here. Selected embodiments and examples In summary, some embodiments of the present disclosure relate to the identification of the sensing application. In particular, a wireless signal is obtained for further processing. The presence of a detection signal is then predicted in the received signal, and said detection signal is a signal generated by a detection application. Then, based on the estimation of the detection signal, wireless reception, transmission or detection is performed. Sensing application detection and/or identification can be performed by a trained module, such as a machine learning-based module. The wireless reception, transmission or detection performed based on the prediction result may also include channel access, resource allocation, use of the detected detection signal for own detection purposes, or the like. According to one embodiment, a method is developed for defining a sensing embodiment: obtaining a received wireless signal; predicting, by a trained 4799/TR module, the presence of a detection signal in the received signal, which is a signal generated by a sensing application; It includes the steps of performing wireless reception, transmission or sensing operations based on the anticipated presence of said sensing application. For example, a sensing application is an application that generates the sensing signal with a preconfigured periodicity or as a continuous transmission signal. For example, said sensing application is one of radar sensing, wireless sensing, wireless local area sensing, channel status information sensing. In some exemplary embodiments, the prediction process further includes determining the sensing application as one of a plurality of predefined sensing applications. According to an exemplary embodiment, the sensing application is an application that generates the sensing signal with a pre-configured periodicity, identifying the sensing application, estimating the pre-configured periodicity based on the presence of the predicted sensing signal within the received wireless signal, and identifying the sensing application from among a plurality of pre-defined sensing applications. It involves using pre-configured estimated periodicity for For example, determining detection applications is based on one or more of the following: frame structure, bandwidth, detection time, detection start time, detection end time, and waveform characteristics; and said one or more features, at least two of a plurality of predefined sensing applications. In some embodiments, said determination of the sensing application as one of a plurality of predefined sensing applications is achieved if: the presence of the detection signal in the received wireless signal is confirmed by the prediction step; and is not performed if the presence of the detection signal in the received wireless signal is not confirmed by the prediction step. In some exemplary embodiments, obtaining the received wireless signal includes measuring one or more of the power density spectrum, received signal strength, or received signal strength indicator; determining by the trained model the presence or absence of a communication or sensing signal in the received wireless signal; if it is determined that a communication signal or a detection signal is present in the received wireless signal, predicting whether the signal is a detection signal; and if the signal is a sensing signal, determining the sensing application as one of a plurality of predefined sensing applications. For example, determining the detection application and estimating the presence are jointly performed by the trained module. Said method may further include predicting the future occupancy of resources by the sensing application based on the result of the prediction step. For example, performing wireless reception, transmission, or detection involves scheduling wireless reception, transmission, or detection on resources that are predicted to be unoccupied by the sensing application. For example, performing wireless reception, transmission, or sensing may include deciding whether to transmit a sensing signal based on the results of the prediction step; and whether or not detection is performed based on the decision-making step. According to one embodiment, a method is developed for training a module to identify a sensing application, said method comprising: the resulting representation of a received wireless signal, and a desired indication of the presence or absence of a sensing signal in the representation, and a desired indication of the presence or absence of a sensing application generating the sensing signal. entering one or both indicators; and modifying at least one parameter of the module based on the input. According to one embodiment, a device for identifying a sensing application is developed, said device comprising a processing circuit configured to obtain a received wireless signal and to predict in the received signal the presence of a sensing signal, which is a signal generated by a sensing application; and a transceiver that enables wireless reception, transmission or detection based on the estimated presence of said 4799/TR sensing application. According to one embodiment, a device for training a module to identify a sensing application is developed, said device providing the module with: a representation of a received wireless signal, and a desired indication of the presence or absence of a sensing signal in the representation, and a desired indication of a sensing application generating the sensing signal. one or both will enter; and a processing circuit configured to modify at least one parameter of the module based on the input. Corresponding methods are also provided, including steps performed by any of the processing circuit implementations mentioned above. In addition, a computer program is developed that contains code instructions that, when implemented by a computer or a processing circuit, perform the steps of any of the methods mentioned above and are stored in a non-volatile medium. According to some embodiments, the processing circuitry and/or transceiver is located within an integrated circuit, IC. Although the subject matter described has been described in detail for illustrative purposes based on what are currently considered to be the most practical and preferred embodiments, such detail is for this purpose only and the subject matter disclosed is not limited to the embodiments disclosed, but rather modifications which are within the spirit and scope of the appended claims. It should be understood that it is intended to cover equivalent regulations. For example, it should be understood that the subject matter described herein is intended to include, to the extent possible, one or more features of any application may be combined with one or more features of any other application.TR TR TR TR TR

Claims (1)

1. CLAIMS A method of defining a sensing embodiment comprising: obtaining a received wireless signal; predicting in the received signal by a trained module the presence of a detection signal, which is a signal generated by a sensing application; It includes the steps of performing wireless reception, transmission or sensing operations based on the anticipated presence of said sensing application. The method according to claim 1, wherein the sensing application comprises a radar detection, wireless sensing, wireless local area detection is one of channel status information detection. The method according to any one of claims 1 to 3, wherein the prediction process further includes determining the detection application as one of a plurality of predefined detection applications. The method according to claim 4, wherein the sensing application is an application that generates said sensing signal with a pre-configured periodicity, determining the sensing application, estimating the pre-configured periodicity based on the presence of the predicted sensing signal within the received wireless signal, and pre-defined It involves using pre-configured estimated periodicity to identify the sensing application among a large number of sensing applications. The method according to any one of claims 4 to 5, wherein the determination of the detection applications is performed based on one or more of: frame structure, bandwidth, detection time, detection start time, detection end time and waveform characteristics; and said one or more features differ for at least two of a plurality of predefined sensing applications. The method according to any one of claims 4 to 6, wherein said determination of the sensing application as one among a plurality of predefined sensing applications: is carried out if the presence of the detection signal in the received wireless signal is confirmed by the prediction step; and is not performed if the presence of the detection signal in the received wireless signal is not confirmed by the prediction step. The method according to any one of claims 1 to 7, wherein: obtaining the received wireless signal includes measuring one or more of the power density spectrum, the received signal strength, or the received signal strength indicator; determining by the trained model the presence or absence of a communication or sensing signal in the received wireless signal; if it is determined that a communication signal or a detection signal is present in the received wireless signal, predicting whether the signal is a detection signal; if the signal is a detection signal, it includes specifying the detection application as one of a plurality of predefined detection applications. The method according to any one of claims 4 to 6, wherein determining the detection application and estimating the presence is performed jointly by the trained module. The method according to any one of claims 1 to 9, further comprising: predicting the future occupancy of the resources by the sensing application, based on the result of the prediction step. The method according to claim 10, wherein performing the wireless reception, transmission or sensing includes scheduling the wireless reception, transmission or sensing on resources that are predicted to be unoccupied by the sensing application. The method according to any one of claims 1 to 11, wherein performing the wireless reception, transmission or detection process includes deciding whether to transmit a detection signal based on the results of the prediction step; and whether or not detection is performed based on the decision-making step. A method of training a module to identify a sensing application, said method comprising one or both of: the module: a representation of a received wireless signal, and a desired indication of the presence or absence of a sensing signal in the representation, and a desired indication of a sensing application generating the sensing signal. to be entered; and modifying at least one parameter of the module based on the input. A device for defining a sensing application, said device comprising: a processing circuit configured to obtain a received wireless signal, and to predict in the received signal the presence of a sensing signal, which is a signal generated by a sensing application; and a transceiver that enables wireless reception, transmission, or detection based on the estimated presence of said sensing application. A device for training a module to identify a sensing application, said device providing the module with: an indication of a received wireless signal, and one or both of a desired indication of the presence or absence of a sensing signal in the representation and a desired indication of a sensing application generating the sensing signal will enter; and a processing circuit configured to modify at least one parameter of the module based on the input. TR TR TR TR TR