Object identification for joint communication and sensing

Abstract

The present disclosure relates to a method for sensing target objects in a cellular network. The method provides a report including type information for a target object to a management node, enabling efficient use of network resources while providing valuable sensing capabilities. The type information is based on a pattern analysis of measurement results from a sensing measurement that employs sensing signals multiplexed with communication signaling. This approach allows for concurrent transmission of sensing and communication signals over the same channel, increasing overall system capacity and reducing costs. The disclosure achieves improved performance in various applications, such as smart cities, autonomous vehicles, or environmental monitoring systems, by enabling accurate and efficient target object sensing identification.

Description

[0001] OBJECT IDENTIFICATION FOR JOINT COMMUNICATION AND SENSING

[0002] TECHNICAL FIELD

[0003] Various examples of the disclosure generally pertain to joint communication and sensing. Various examples specifically pertain to object detection and object identification in joint communication and sensing.

[0004] BACKGROUND

[0005] Sensing has been identified as a potential new feature for the sixth generation (6G) of cellular mobile communications. It is designed to operate jointly with communication, and is therefore commonly referred to as Joint Communication and Sensing (JCAS) and Integrated Sensing and Communication (ISAC). JCAS is a framework in which communication and sensing operations are combined and optimized to share the same spectrum, hardware and signal processing capabilities, thereby increasing the efficiency and utility of 6G networks. Similarly, ISAC integrates sensing capabilities into communication systems, enabling the simultaneous transmission of communication and sensing signals, leading to new functionalities and use cases in 6G.

[0006] According to the International Telecommunication Union's (ITU) 6G vision described in IMT-2030 (International Mobile Telecommunications), sensing has been listed as a new capability for 6G. The ITU's IMT-2030 vision outlines the framework and capabilities expected in 6G networks, focusing on advanced capabilities such as high-speed connectivity, low latency, and integrated sensing.

[0007] URLLC, eMBB, and mMTC are the known usage scenarios from the 5G timeframe. URLLC (Ultra-Reliable Low Latency Communications) is designed for applications that require extremely reliable and low-latency communications, such as autonomous driving and industrial automation. eMBB (Enhanced Mobile Broadband) targets high-data-rate services, including extended Reality (XR) applications, such as augmented reality (AR), virtual reality (VR), and cloud gaming (CG). mMTC (Massive Machine Type Communications) supports large-scale loT (Internet of Things) deployments with a large number of low-complexity devices.

[0008] For 6G, ISAC is listed as one of the new usage scenarios. The operation of ISAC is expected to enable new use cases / services based on the 6G system. These services include intelligent transportation, aviation, enterprise, smart city, smart home, factories, consumer applications, telepresence, healthcare, and the public sector. For example, URLLC can benefit from ISAC / JCAS through enhanced situational awareness and real-time environmental sensing. For example, in autonomous driving, ISAC can provide critical sensing information about nearby objects and road conditions, improving safety and decision making. eMBB can leverage ISAC / JCAS to enhance user experiences in AR and VR by integrating real-world sensing data to provide more immersive and interactive environments. mMTC may benefit from ISAC / JCAS by enabling devices to not only communicate but also sense their environment. This can enhance applications in smart cities, smart homes, healthcare monitoring, and industrial loT by providing real-time sensing data for better resource management and automation. Incorporating sensing into a radio network infrastructure, such as a base station and user equipment (UE), can pose many new challenges, such as the transmission of reference signals for sensing purposes, sensing measurement, and coordination between nodes.

[0009] The 3GPP specifications define three types of sensing operations. Mono-static sensing refers to an operation where the sensing transmitter and receiver are co-located, typically within the same device. Bi-static sensing is an operation where the sensing transmitter and receiver are in different locations, allowing for wider coverage and more flexible deployment scenarios. They can be the same or different types of devices. Multi-static sensing involves multiple transmitters and receivers distributed across a network, providing enhanced sensing accuracy and robustness. These operations may be implemented by UEs and / or base stations, e.g. gNBs. A gNB can contain one or more Transmission-Reception Points (TRPs). Based on these operations, six sensing modes or topologies have been considered, i.e. TRP-TRP bistatic, TRP mono-static, TRP-LIE bi-static, LIE-TRP bi-static, UE-llE bi-static, UE mono-static.

[0010] The following sensing entities are defined within the ISAC framework:

[0011] UE (User Equipment): Capable of acting as either a SeRS (Sensing Reference Signal) transmitter or receiver. The SeRS is a specially designed signal used for sensing purposes that is transmitted by either the gNB or the UE and is critical for accurate sensing measurements. gNB (base station): Acts as a sensing transmitter or receiver. A gNB may include one or several Transmission and Reception Point(s), TRP(s).

[0012] Sensing target: Refers to (passive) objects (e.g., vehicles, unmanned arial vehicles (UAVs), people, or more generally: target objects), typically without network connectivity, whose detection or characteristics are of interest to the network.

[0013] Environment object (EO): Refers to objects outside the network's scope of interest.

[0014] SeMF (Sensing Management Function): Provides support for sensing measurements, post-processing of sensing data, and efficient computation of sensing results (e.g., sensing estimation). SeMF can be deployed as a new node in the core network or as a new function residing in another core network node. It manages sensing tasks, processes sensing data, and derives insights from the collected sensing measurements, playing a central role in the ISAC framework.

[0015] These sensing entities operate within a network, enabling various sensing use cases, including object detection and identification. In a typical downlink (DL)-like bi-static sensing ISAC deployment, the gNB serves as the SeRS transmitter, broadcasting SeRS signals to its environment. These signals propagate through the air and divide into different rays. Each ray will bounce off different objects, either the target or the environment object. These interactions change the characteristics of the reflected rays.

[0016] In this deployment, the UE acts as a SeRS receiver, with the goal of capturing the reflected rays and estimating the sensing channel. Sensing channel estimation is the process of evaluating the characteristics of SeRS signals, including the changes in the characteristics, as they reflect off objects, which helps determine the presence, location, velocity, and other characteristics of the sensing target. This estimation allows the UE to derive sensing measurements associated with the target, which are then reported to SeMF for further processing. The SeMF is used to perform various sensing tasks such as object detection and localization using the received measurement report from UE. In other sensing modes / topologies, the measurement report may also come from the gNB.

[0017] Object detection and identification may rely on estimating delay and Doppler, which can be further translated to range and velocity information related to the target. To capture Doppler information, the SeRS signal may be transmitted periodically over time.

[0018] The prior art techniques may face certain restrictions and drawbacks. For instance, it has been found that accurate object identification may be difficult based on such measurement report. Also estimating more detailed information such as the object’s type, shape and size may be difficult.

[0019] SUMMARY

[0020] Accordingly, advanced techniques of joint communication and sensing are required.

[0021] This need is met by the features of the independent claims. The features of the dependent claims define embodiments.

[0022] One aspect of the present disclosure relates to a method for use in a radio node of a cellular network. The method comprises providing, to a management node of the cellular network, a report including a type information for a target object. The type information is based on a pattern analysis of measurement results of a sensing measurement. The sensing measurement employs sensing signals for sensing the target object. The sensing signals are multiplexed with communication signaling of the cellular network.

[0023] A radio node may be understood as a device that transmits and receives radio signals, such as a base station or user equipment. A management node may be understood as a component that oversees and controls the operation of the cellular network. A target object refers to an entity whose characteristics are being sensed, such as a vehicle or a person.

[0024] A pattern analysis of measurement results may involve applying algorithms to identify patterns in the data collected during the sensing measurement. This can help to determine the type information for the target object. Sensing signals may be understood as specialized signals used for sensing purposes, which can be transmitted by the radio node or another radio node and received by the radio node. Multiplexing with communication signaling means that the sensing signals are combined with regular communication data for transmission over the cellular network.

[0025] An effect of this disclosure may be reliable type determination based on radio sensing techniques. By using pattern analysis of measurement results, the method can accurately identify the target object's characteristics, which is particularly useful in various applications such as smart cities or industrial automation.

[0026] In an example, the radio node may obtain, from the management node or another node of the cellular network, a request for sensing the target object. The type information may be provided upon obtaining the request.

[0027] A request for sensing the target object may be understood as a command or instruction from the management node to initiate the sensing process. This can help to control when and how the sensing measurement is performed. Obtaining the request may involve receiving a message or signal from the management node, which triggers the radio node to start the sensing process.

[0028] In this arrangement, an effect may be that the management node controls when the measurement sensing is initiated, allowing for more efficient use of network resources and coordination with other nodes in the cellular network. This can also help to ensure that the sensing measurement is performed only when necessary, reducing unnecessary data collection and processing.

[0029] In some examples, the type information comprises an indicator indicative of an object type and optionally an associated probability value.

[0030] The indicator indicative of an object type may be understood as a label or identifier that categorizes the target object into a specific class or category. This can help to provide more detailed information about the target object. The associated probability value may represent the confidence level or likelihood that the target object belongs to the indicated object type.

[0031] In an example, the type information comprises an indicator indicative of multiple object types and associated probability values indicating, for each of the multiple object types, a probability that the target object is of the corresponding object type.

[0032] The associated probability values may represent the confidence level or likelihood that the target object belongs to each of the indicated object types.

[0033] By providing probability values for each object type, this arrangement allows the system to reason about the uncertainty of the target object's identity, enabling it to make more informed decisions and adapt its behavior accordingly.

[0034] In an example, the pattern analysis is based on pattern reference information discriminating between multiple object types.

[0035] The pattern reference information may be understood as a set of predefined patterns that are used to distinguish between different object types. This can help to improve the accuracy and robustness of the object type identification process. The pattern analysis may involve comparing the measurement results against the pattern reference information to determine which object type best matches the target object.

[0036] In this arrangement, an effect may be that the system can effectively discriminate between multiple object types, even in cases where the objects have similar characteristics or features. This can lead to improved performance in various applications such as object tracking, classification, and recognition.

[0037] In some examples, the pattern reference information comprises at least one reference pattern. Each reference pattern of the at least one reference pattern may be assigned to an object type.

[0038] A reference pattern may be understood as a predefined set of characteristics or features that are specific to a particular object type. This can help to improve the accuracy and robustness of the object type identification process. The assignment of each reference pattern to an object type may involve mapping the pattern to a specific class or category of objects. In this arrangement, an effect may be that the system can efficiently map measurement results to corresponding object types using the pre-assigned reference patterns. This can lead to faster and more accurate object type identification, as well as improved performance in various applications such as object tracking, classification, and recognition.

[0039] The pattern analysis may comprise a cluster pattern matching. The at least one reference pattern may comprise at least one reference cluster pattern.

[0040] A cluster pattern matching may be understood as a process of comparing measurement results against predefined clusters or groups of patterns to identify the closest match. This can help to improve the accuracy and robustness of the object type identification process. A reference cluster pattern may be a specific instance of a cluster pattern that is associated with an object type.

[0041] By using cluster patterns, this arrangement allows the system to capture complex relationships between features and characteristics of objects, enabling it to make more informed decisions and adapt its behavior accordingly. Additionally, this may also enable the system to reduce the dimensionality of the feature space, as it can focus on identifying the closest match within a predefined set of cluster patterns rather than analyzing individual features separately. This can lead to faster and more accurate object type identification, as well as improved performance in various applications such as object tracking, classification, and recognition.

[0042] In various examples, the pattern reference information comprises a machine-learned model.

[0043] A machine-learned model may be understood as a computational model that has been trained on data to learn patterns and relationships between features. This can help to improve the accuracy and robustness of the object type identification process. An effect may be that the system can identify objects from different views, in different contexts and distances, by leveraging the machine-learned model's ability to capture complex patterns and relationships in the data. By using a machine-learned model, this arrangement allows the system to adapt to new situations and environments, enabling it to make more informed decisions and improve its performance over time. Additionally, this arrangement may also enable the system to identify objects even when they are partially occluded or viewed from unusual angles, as the machine- learned model can learn to recognize patterns in the data that are not easily visible to humans. The method may further comprise obtaining, from the management node or another node of the cellular network, at least one of the pattern reference information and / or a configuration associated with the pattern reference information. The pattern reference information may comprise reference patterns, wherein each reference pattern is assigned to an object type.

[0044] A configuration associated with the pattern reference information may be understood as a set of parameters or settings that define how to use the pattern reference information. This can help to improve the efficiency and effectiveness of the object type identification process.

[0045] In an example, the method may include selecting the pattern reference information based on an indication of at least one object type to be sensed. The indication of the at least one object type to be sensed may be a hint or suggestion provided by the management node or another node in the cellular network about what types of objects are likely to be present.

[0046] The method may include comparing the measurement results against reference patterns comprised in the pattern reference information and assigned to the at least one object type to be sensed. This can help to improve the accuracy and robustness of the object type identification process by focusing on a specific set of object types that are likely to be present. Thus, the system can efficiently analyze only the expected object types, reducing the computational complexity and improving the speed of the analysis. By selecting pattern reference information based on an indication of at least one object type to be sensed, this arrangement allows the system to adapt to changing environments and contexts, enabling it to make more informed decisions and improve its performance over time. Additionally, this arrangement may also enable the system to conserve resources by avoiding unnecessary computations and focusing only on relevant data.

[0047] In an example, the configuration associated with the pattern reference information may comprise an indication of multiple reference patterns assigned to the at least one object type of the indication.

[0048] Indicating multiple reference patterns allows to capture a more comprehensive view of the target object, enabling it to make more informed decisions and improve its performance over time. Additionally, this arrangement may also enable the system to adapt to changing environments and contexts by adjusting the weight or importance of each reference pattern in real-time. This can lead to improved accuracy and robustness in object type identification, especially in cases where the target object has multiple distinct features or characteristics.

[0049] In an example, at least one of the pattern reference information or the configuration associated with the pattern reference information may be obtained once for multiple providing of the report and / or for each providing of the report.

[0050] Obtaining the pattern reference information or configuration once for multiple providing of the report means that the system can reuse the obtained information across multiple reporting instances. This can help to reduce the overhead of obtaining new information for each report, making the process more efficient. In an example, this arrangement may also enable the system to cache the obtained information and use it for future reports, reducing the need for repeated computations or queries. By reusing the obtained information, this arrangement allows the system to improve its performance over time and reduce the latency associated with obtaining new information.

[0051] Additionally, obtaining the pattern reference information or configuration once for each providing of the report means that the system can adapt to changing conditions and environments on a per-report basis. This can help to ensure that the system is always using the most up-to-date and relevant information for each report, making the process more accurate and reliable.

[0052] The configuration associated with the pattern reference information may comprise one or more parameters of the pattern analysis.

[0053] The one or more parameters may be understood as adjustable settings that control how the pattern analysis is performed. The one or more parameters may comprise an evaluation threshold for identifying a certain object type. The evaluation threshold may be a minimum score or value that must be met in order for the system to consider a particular object type as a match. Additionally, the one or more parameters may also comprise a timing correlation constraint for discriminating between multiple different objects. This can help to avoid erroneous detection of objects that are close together by ensuring that the system only considers objects that are distinct, at least for a certain level. This can help to reduce false positives and improve the overall accuracy of the object type identification process. Furthermore, the one or more parameters may also comprise a pattern matching metric for comparing two scattering patterns. The pattern matching metric may be a mathematical formula or algorithm that is used to determine how similar two patterns are.

[0054] In various examples, the configuration associated with the pattern reference information may comprise one or more timing constraints for the sensing measurement.

[0055] For example, the use of a timing constraint may enable the system to adjust its sampling rate in response to changing conditions. An adapted sampling rate may increase accuracy and efficiency by ensuring that the system is collecting data at the optimal frequency for the specific application or environment. Additionally, the one or more timing constraints may also help to reduce power consumption and improve overall system performance.

[0056] In some examples, the pattern analysis may comprise a comparison between the measurement results and at least two reference patterns.

[0057] The at least two reference patterns may differ in at least one of angles of scattering of the sensing signals, surface material of the target object, or measurement distances. This can help to improve the robustness of the detection process by providing multiple perspectives on the target object. For example, the use of multiple reference patterns with different angles of scattering may enable the system to detect objects from various perspectives, reducing the impact of occlusions or other environmental factors. Similarly, the use of multiple reference patterns with different surface materials may enable the system to detect objects with varying textures or reflective surfaces, improving its ability to adapt to changing environments. Additionally, this may also reduce false positives and improve overall detection accuracy. In some examples, the measurement results may comprise one or more observables associated with each scattering point of multiple scattering points of the target object sensed with the sensing signals.

[0058] The one or more observables may be understood as measurable quantities that provide information about the target object. By using one or more observables for each scattering point, this arrangement allows the system to capture detailed information on the target object. For example, the use of multiple scattering points may enable the system to detect objects from various perspectives, reducing the impact of occlusions or other environmental factors. Similarly, the use of one or more observables for each scattering point may enable the system to capture a more nuanced view of the target object's surface material, shape, and size. Additionally, this arrangement may also enable the system to adapt to changing environments and contexts by adjusting the number of scattering points and observables used for each measurement. In an example, the method may further comprise providing a set of observables from the sensing measurement on the basis of which the pattern analysis determined the type information.

[0059] The set of observables may include at least one of delay information, Doppler shift information, power information, and phase information.

[0060] Based on this, the management node may generate additional reference patterns and reconstruct pattern analysis in the radio node thus improve configuration of pattern analysis in the radio node. By leveraging the set of observables provided from the sensing measurement, this arrangement allows the system to refine its understanding of the target object and adapt its behavior accordingly.

[0061] For example, the use of delay information may enable the management node to generate additional reference patterns that account for variations in the timing of the sensing signals. Similarly, the use of Doppler shift information may enable the management node to reconstruct pattern analysis in radio node to capture changes in the velocity or direction of the target object. Additionally, this arrangement may also enable the management node to optimize the configuration of pattern analysis in radio node by adjusting the weight or importance of each observable.

[0062] In an example, the report may comprise a target identifier assigned to the target object.

[0063] The target identifier may be understood as a unique label or code that is associated with the target object. This can help to improve the accuracy and robustness of the object type identification process by providing a clear and unambiguous way to identify the target object. The object may be traced over several measurements / detections. By assigning a unique target identifier to the target object, this arrangement allows the system to maintain a record of the object's location and status over time.

[0064] In various examples, the report comprises a measurement quality value related to the sensing measurement.

[0065] The measurement quality value may be understood as a numerical representation of the accuracy and reliability of the sensing measurement. In this arrangement, the measurement quality value may indicate a peak-to-noise ratio of a peak power detected for the target object and a noise power.

[0066] For example, the peak power may be understood as the maximum power level detected for the target object during the sensing measurement. The noise power, on the other hand, may be understood as the background noise or interference present in the sensing environment.

[0067] By including the measurement quality value in the report, the management node may determine a reliability of the provided object type determination. An effect of this may be that the management node can assess the confidence level of the object type identification result. By analyzing the peak-to-noise ratio indicated by the measurement quality value, the management node may determine whether the sensing measurement was reliable and accurate enough to support the identified object type.

[0068] In an example, if the measurement quality value indicates a high peak-to-noise ratio, the management node may infer that the sensing measurement was of high quality and the identified object type is likely to be correct. Conversely, if the measurement quality value indicates a low peak-to-noise ratio, the management node may infer that the sensing measurement was of poor quality and the identified object type may not be reliable.

[0069] The report may further include an object presence detection quality value indicating a probability that the target object is present.

[0070] The object presence detection quality value may be understood as a numerical representation of the confidence level that the target object is actually present in the sensing environment. This value may indicate a probability that the target object is present, ranging from 0 to 1. In this arrangement, the report may provide an indication of whether the detected signal or measurement is likely to correspond to the actual presence of the target object, rather than noise or false signals. An effect of this arrangement is that it helps avoid false positive detections and identifications.

[0071] In an example, the method may further comprise providing a capability indication to the management node indicating a capability to provide type information for a target object.

[0072] The capability indication may be understood as a signal or message that informs the management node about the ability of the radio node to provide type information for a target object. This indication may include information about the capabilities and limitations of the radio node in terms of object identification.

[0073] An effect of this arrangement is that it allows the management node to make informed decisions about how to proceed with object identification. By considering the capability indication from the radio node, the management node can determine whether to perform object type identification in the radio node or in the management node itself. In an example, if the capability indication indicates that the radio node has limited capabilities for object identification, the management node may decide to perform the identification itself, using its own resources and algorithms. On the other hand, if the capability indication indicates that the radio node is fully capable of accurate and reliable object identification, the management node may delegate this task to the radio node. This type of capability is typically supported when the radio node has high computation capabilities and can perform complex computations. Hence, the radio node is capable of performing various computation tasks, including a complex sensing measurement. This arrangement further provides a technical advantage by enabling adequate configuration of object identification in the radio node. By taking into account the capabilities and limitations of the radio node, the system can optimize its performance and ensure that object identification is performed accurately and efficiently.

[0074] One aspect of the present disclosure relates to a method for use in a management node of a cellular network. The method comprises obtaining, from a radio node of the cellular network, e.g., the above described radio node, a report including type information for a target object. The type information is based on a pattern analysis of measurement results of a sensing measurement. The sensing measurement employs sensing signals for sensing the target object, and these sensing signals are multiplexed with communication signaling of the cellular network.

[0075] An effect may be that the method provides accurate type information about target objects, which is essential for various applications such as object tracking, surveillance, and environmental monitoring. The pattern analysis employed by the method allows for sophisticated identification and characterization of target objects based on their measured properties.

[0076] The method can be inter-related to the above described method for use in the radio node.

[0077] For example, the method may further comprise providing, to the radio node, a request for sensing the target object. The type information may be obtained upon providing the request.

[0078] The pattern analysis may be based on pattern reference information discriminating between multiple object types. The method may further comprise providing, to the radio node, at least one of the pattern reference information or a configuration associated with the pattern reference information. The at least one of the pattern reference information or the configuration associated with the pattern reference information may be provided once for multiple consecutive obtaining of the report and / or for each obtaining of the report.

[0079] The method may further comprise obtaining, from the radio node, a capability indication, indicating a capability to provide type information for a target object.

[0080] Effects described above in connection with the method for use in the radio node may similarly apply to the method for use in the inter-related management node and will therefore not be repeated.

[0081] One aspect of the present disclosure relates to a method for use in a management node of a cellular network. The method comprises obtaining, from a radio node of the cellular network, measurement results of a sensing measurement. The sensing measurement employs sensing signals for sensing a target object. The sensing signals are multiplexed with communication signaling of the cellular network. The method further comprises determining a type information for the target object based on a pattern analysis of the measurement results.

[0082] In other words, according to this method, the pattern analysis is performed in the management node based on sensing measurement results obtained from the radio node.

[0083] An effect may be that the method enables efficient use of network resources for sensing and communication purposes. By multiplexing sensing signals with communication signaling, the method allows both types of data to share the same transmission channel, the same hardware / equipment in both transmitter and receiver, reducing the need for separate channels or infrastructure. This can lead to improved spectral efficiency, reduced costs, and enhanced overall performance of the cellular network.

[0084] Additionally, the method may provide accurate type information about target objects, which is essential for various applications such as object tracking, surveillance, and environmental monitoring. The pattern analysis employed by the method allows for sophisticated identification and characterization of target objects based on their measured properties, even if the radio node does not provide sufficient capabilities for pattern analysis.

[0085] In an example, the method may further comprise providing, to the radio node, a request for providing the measurement results.

[0086] A request for providing the measurement results may be understood as a command or instruction sent from the management node to the radio node, prompting the radio node to perform sensing measurement and transmit the measurement results of the sensing measurement.

[0087] Providing this request may trigger the radio node to send the measurement results to the management node, which can then determine the type information for the target object based on a pattern analysis of the received measurement results. An effect may be that this arrangement enables more efficient and reliable exchange of measurement data between the radio node and the management node. By explicitly requesting the measurement results, the method can ensure that the data is transmitted only when needed, reducing unnecessary data transmission and minimizing network congestion. This can lead to improved overall performance, reduced latency, and enhanced reliability of the cellular network. The type information may comprise an indicator indicative of an object type and optionally an associated probability value.

[0088] An indicator indicative of an object type may be understood as a representation or classification of the target object based on its characteristics, features, or attributes. This indicator can provide a clear identification of the object type, enabling further processing or decision-making. The optional associated probability value may represent a confidence level or likelihood that the identified object type is correct. This information may be provided to an application relying on the object type identification.

[0089] The type information may comprise an indicator indicative of multiple object types and associated probability values indicating, for each of the multiple object types, a probability that the target object is of the corresponding object type.

[0090] The associated probability values may represent confidence levels or likelihoods that the identified object types are correct. These probability values can provide additional context and reliability to the object classification, allowing an application relying the identification for more informed decisions or actions.

[0091] In an example, this arrangement may enable a probabilistic approach to object classification, where the target object is assigned a probability of belonging to each of the multiple object types. This approach can account for uncertainty or ambiguity in the sensing data, leading to more accurate and reliable results and may allow for improved handling of ambiguous or noisy sensing data.

[0092] In an example, the method may further comprise providing, to the radio node, an indication of at least one object type to be sensed. The sensing measurement is to be performed based on the at least one object type to be sensed.

[0093] An indication of at least one object type to be sensed may be understood as a specification or instruction provided to the radio node, indicating which specific object types are to be targeted / expected during the sensing process. This indication can enable the radio node to focus its sensing efforts and tailor its measurement parameters accordingly. The sensing measurement is to be performed based on the at least one object type to be sensed may mean that the radio node adjusts its measurement settings, such as frequency range or transmission power, sampling rate, to optimize detection of the specified object types. This targeted approach can lead to more efficient use of resources and improved accuracy in detecting the desired object types.

[0094] In an example, the pattern analysis may include a cluster pattern matching.

[0095] A cluster pattern matching may be understood as a process of identifying and grouping similar patterns or features within the sensing data, based on their spatial distribution and relationships. This approach can enable the identification of complex patterns and anomalies that may not be apparent through individual feature analysis. The pattern analysis may include a comparison between the measurement results and at least two reference patterns. The at least two reference patterns may differ in at least one of: angles of scattering of the sensing signals, surface material of the target object, and measurement distances. By considering multiple reference patterns with varying parameters, the method can account for different object instances / arrangements of the same type and improve its accuracy. For instance, the angles of scattering of the sensing signals may refer to the way in which the sensing signals interact with the target object, including the angle at which they are scattered or reflected. The surface material of the target object may affect how the sensing signals are absorbed, reflected, or transmitted, while the measurement distances can influence the strength and characteristics of the received signals. This may enable more accurate and robust object classification across varying conditions.

[0096] The pattern analysis may be based on pattern reference information discriminating between multiple object types. A configuration associated with the pattern reference information may include one or more parameters of the pattern analysis. These parameters may be used to customize and refine the pattern analysis process. The one or more parameters may include an evaluation threshold for identifying a certain object type, a timing correlation constraint for discriminating between multiple different objects, and a pattern matching metric for comparing two scattering patterns.

[0097] An evaluation threshold may be understood as a criteria used to determine whether a detected pattern matches a known object type. By adjusting this threshold, the method can balance the trade-off between false positives and false negatives. The use of timing correlation constraints, in particular, can avoid erroneous detection of objects that are close together.

[0098] In an example, the method may further comprise determining a target identifier assigned to the target object.

[0099] A target identifier may be understood as a unique label or code associated with a specific object. This identifier can be used to track and distinguish the object from others. By determining the target identifier, the method can establish a link between the detected object and its corresponding identity. This information can then be used for various purposes, such as tracking the object's movement, monitoring its status, or triggering specific actions based on its presence. By assigning a unique identifier to each object, the method can create a record of the object's history, including its location, velocity, and interactions with other objects. This capability is particularly useful in applications where tracking individual objects is critical, such as in logistics, surveillance, or asset management.

[0100] The method may further comprise obtaining a measurement quality value related to the sensing measurement from the radio node. The measurement quality value may indicate a peak-to-noise ratio (PNR) of a peak power detected for the target object and a noise power. The PNR can be understood as a metric that quantifies the relative strength of the signal compared to the background noise.

[0101] By obtaining the measurement quality value, the method can assess the reliability and accuracy of the sensing measurement. A high PNR may indicate a strong signal with minimal noise interference, while a low PNR may suggest a weak signal or significant noise contamination. This may enable more accurate object detection and characterization. By considering the measurement quality value, the method can adapt its decision-making process to account for variations in sensing conditions. For instance, if the measurement quality value indicates a high level of noise interference, the method may choose to repeat the measurement to validate the results. The method may further comprise determining an object presence detection quality value.

[0102] The object presence detection quality value may indicate a probability that the target object is present. This probability can be understood as a measure of confidence in the detection result, ranging from 0 (indicating no confidence) to 1 (indicating high confidence). A high quality value may indicate a strong likelihood of the target object being present, while a low quality value may suggest uncertainty or ambiguity in the detection result. This may reduce false positives.

[0103] One aspect of the present disclosure relates to a method for use in a radio node of a cellular network. The radio node may provide, to a management node of the cellular network, measurement results of a sensing measurement. The sensing measurement may employ sensing signals for sensing a target object. The sensing signals may be multiplexed with communication signaling of the cellular network.

[0104] The method can be inter-related to the above described method for use in the management node to support object type determination in the management node based on the provided measurement results of the sensing measurement

[0105] For example, the method may further comprise obtaining, from the management node, a request for providing the measurement results.

[0106] Further, the method may comprise obtaining, from the management node, an indication of at least one object type to be sensed. The sensing measurement may be performed based on the at least one object type to be sensed.

[0107] The method may comprise obtaining, from the management node, an indication of at least one target object to be sensed. The sensing measurement may be performed based on the at least one target object to be sensed.

[0108] The method may also comprise providing, to the management node, a measurement quality value related to the sensing measurement, the measurement quality value indicating a peak-to-noise ratio of a peak power detected for the target object and a noise power.

[0109] The measurement results may comprise one or more observables. The one or more observables are associated with each scattering point of multiple scattering points of the target object sensed with the sensing signals.

[0110] Effects described above in connection with the method for use in the management node may similarly apply to the method for use in the inter-related radio node and will therefore not be repeated.

[0111] It is to be understood that the features mentioned above and those yet to be explained below may be used not only in the respective combinations indicated, but also in other combinations or in isolation without departing from the scope of the invention.

[0112] BRIEF DESCRIPTION OF THE DRAWINGS

[0113] FIG. 1 schematically illustrates a mono-static sensing topology of a sensing measurement according to various examples.

[0114] FIG. 2 schematically illustrates a bi-static sensing topology of a sensing measurement according to various examples. FIG. 3 schematically illustrates a multi-static sensing topology of a sensing measurement according to various examples.

[0115] FIG. 4 schematically illustrates a system for JCAS / ISAC according to various examples.

[0116] FIG. 5 schematically illustrates a multiple scattering point measurement simulation according to various examples.

[0117] FIG. 6 schematically illustrates a delay spectrum of a channel impulse response according to various examples.

[0118] FIG. 7 schematically illustrates an enlarged part of the delay spectrum of FIG. 6.

[0119] FIG. 8 schematically illustrates a cluster pattern matching according to various examples FIG. 9 schematically illustrates a further cluster pattern matching according to various examples.

[0120] FIG. 10 is a signaling diagram according to various examples.

[0121] FIG. 11 is a signaling diagram according to further examples.

[0122] FIG. 12 schematically illustrates an apparatus according to various examples.

[0123] FIG. 13 schematically illustrates an apparatus according to various examples.

[0124] FIG. 14 is a flowchart of a method according to various examples.

[0125] FIG. 15 is a flowchart of a method according to various examples.

[0126] FIG. 16 is a flowchart of a method according to various examples.

[0127] FIG. 17 is a flowchart of a method according to various examples.

[0128] DETAILED DESCRIPTION

[0129] Some examples of the present disclosure generally provide for a plurality of circuits or other electrical devices. All references to the circuits and other electrical devices and the functionality provided by each are not intended to be limited to encompass only that which is illustrated and described herein. While particular labels may be assigned to the various circuits or other electrical devices disclosed, such labels are not intended to limit the scope of operation for the circuits and the other electrical devices. Such circuits and other electrical devices may be combined with each other and / or separated in any manner based on the particular type of electrical implementation that is desired. It is recognized that any circuit or other electrical device disclosed herein may include any number of microcontrollers, a graphics processor unit (GPU), integrated circuits, memory devices (e.g., FLASH, random access memory (RAM), read only memory (ROM), electrically programmable read only memory (EPROM), electrically erasable programmable read only memory (EEPROM), or other suitable variants thereof), and software, which interact with each other to perform the operation(s) disclosed herein. In addition, any one or more of the electrical devices may be configured to execute a program code that is embodied in a non-transitory computer-readable medium programmed to perform any number of the functions as disclosed.

[0130] In the following, embodiments of the invention will be described in detail with reference to the accompanying drawings. It is to be understood that the following description of embodiments is not to be taken in a limiting sense. The scope of the invention is not intended to be limited by the embodiments described hereinafter or by the drawings, which are taken to be illustrative only. The drawings are to be regarded as being schematic representations and elements illustrated in the drawings are not necessarily shown to scale. Rather, the various elements are represented such that their function and general purpose become apparent to a person skilled in the art. Any connection or coupling between functional blocks, devices, components, or other physical or functional units shown in the drawings or described herein may also be implemented by an indirect connection or coupling. A coupling between components may also be established over a wireless connection. Functional blocks may be implemented in hardware, firmware, software, or a combination thereof.

[0131] Hereinafter, aspects related to JCAS are disclosed. JCAS corresponds to a communication system that additionally offers sensing functionality. JCAS communication systems re-use hardware for sensing, thereby saving resources if compared to a reference scenario in which two separate systems are used. The aspects disclosed herein can be similarly applied to ISAC.

[0132] While positioning refers to the estimation of the position of (active) radio nodes connected to the cellular network, e.g., wireless terminals (UEs), sensing, on the other hand, enables to additionally sense passive objects in a surrounding of radio nodes connected to the cellular network. A passive object does not actively transmit or receive the sensing signals. Sensing signals are reflected and / or scattered at the passive objects. Sensing includes a transmission of a sensing signal or a set of sensing signals and processing these sensing signals to extract sensing information.

[0133] A sensing signal may be a chirped signal, i.e., incorporating a frequency sweep over a certain allocated bandwidth. A sensing signal may be an Orthogonal Frequency Division Multiplexing signal. A sensing signal may be Code Division Multiplexed. A base signal is sinusoidal, but a spreading code is applied. The spreading code includes a sequence of chips (e.g., +1, -1, -1 , +1, +1, -1 , ...). For instance, aperiodic or random or pseudo-random sequence of chips can be used. The sequence of chips is then mapped to phase values which stay constant during the chirp duration. The sensing signal that has a bandwidth that is proportional to the inverse of the chirp duration. By choosing orthogonal spreading sequences, code division multiplexing (CDM) of multiple sensing signals can be achieved. Thereby, a receiver radio node (RX radio node) receiving the sensing signals can separate respective information. A sensing signal can be similar to the existing reference signal in 5G NR (New Radio), such as positioning reference signal (PRS), Channel State Information Reference Signal (CSI-RS), etc. A sensing signal can be associated to a certain frequency range. A sensing signal in frequency range 1 and / or frequency range 2 can be multiplexed using OFDM. Sensing signals in different frequency ranges may have different configurations, such as sub-carrier spacing (SCS) or bandwidth. A sensing signal may be in a different form if it is operated in sub tera Hertz frequencies in which non-OFDM is expected to be deployed.

[0134] The sensing measurements of joint communication and sensing employ sensing signals that are multiplexed with communication signaling of the cellular network. For instance, time multiplexing and / or frequency multiplexing can be used. It would be possible that timefrequency resources are centrally allocated by a scheduler that schedules communication signaling as well to the sensing signals. For instance, a time-frequency resource grid of a wireless link may include multiple resource elements and these resource elements may be allocated either to communication signaling or sensing signals of the sensing measurement by the scheduler. For instance, such scheduler may reside at a base station of a radio-access network of the cellular network. By multiplexing these sensing signals with the communication signaling of the cellular network, interference between the communication signaling and the sensing measurement can be mitigated. Furthermore, the sensing signals and communication signaling may be both transmitted and received by the same hardware. In some examples, sensing and communication may use the exact same signal, such as one signal for both sensing and communication instead of two independent sequences / signals multiplexed in different spectrum.

[0135] Sensing can be used to support various use cases, such as object detection (presence), object identification, object tracking, object mapping, object positioning, object ranging, object counting, object velocity, etc.

[0136] JCAS / ISAC can employ various sensing topologies. Among varies topologies, three commonly discussed configurations are mono-static, bi-static and multi-static sensing.

[0137] FIG. 1 schematically illustrates a mono-static sensing topology A user equipment (UE) 106 implements, both, a transmitter radio node (TX radio node) transmitting sensing signals 120; as well as a receiver radio node (RX radio node) detecting echoes 126 of the sensing signals 120 reflected and / or scattered at a physical (passive) target object 114 in the surrounding. For instance, the UE 106 may use the same antenna panel for transmitting the sensing signals 120 and for receiving the echoes of the sensing signals 126.

[0138] FIG. 2 schematically illustrates a bi-static sensing topology. Here, a base station (BS) 108, e.g. a gNB, implements the TX radio node and a UE 106 implements the RX radio node. In another example, a UE 106 implements the TX radio node and a BS 108 implements the RX radio node. More generally, in a bi-static sensing topology, two different radio nodes participate in the sensing measurement and cooperate to implement the sensing measurement.

[0139] FIG. 3 schematically illustrates a multi-static sensing topology. Generally, in multi-static sensing topologies a plurality of TX radio nodes and / or a plurality of RX radio nodes are involved. The example of FIG. 3 can be seen as a combination of the mono-static sensing topology of FIG. 1 and the bi-static sensing topology of FIG. 2. In FIG. 3, the BS 108 implements, both, a TX radio node as well as a RX radio node, i.e. , a TX / RX radio node. A further BS 118 implements a TX radio node transmitting a further sensing signal 130 causing an echo 132 at the target object 114. The UE 106 implements a RX radio node. Multiple echoes 126, 302, 132 of the sensing signals 120, 130 are detected.

[0140] There are further sensing topologies, beyond those illustrated in FIG. 1, FIG. 2, and FIG. 3. For example, in a multi-static sensing topology in which more than three different radio nodes participate in and cooperate to implement the sensing measurement. For instance, multiple UEs can participate in the sensing measurement. The RX radio node may be different than the TX radio node. In another example, multiple BSs may act as TX radio nodes and a UE may act as RX radio node, participating in sensing measurement by receiving the sensing signals from the multiple BSs. Using a sensing measurement based on a sensing signal, it is possible to determine object-related information for a passive object such as the object 114. This is done by investigating multiple multipath components of the radio channel. The multipath components stem from radio signals reaching the RX radio node through various spatial paths due to reflections, diffraction, and scattering caused by the objects (obstacles) in the environment, such as buildings, trees, terrain, vehicles, persons, etc.. Each spatial path is associated with a respective distance between TX radio node and RX radio node; typically, those distances are different for different spatial paths and thus a given sensing signal (or, more precisely, multiple echoes of the sensing signal) arrives at the RX radio node at slightly different times for the different multipath components. This temporal spread, known as delay spread, leads to each multipath component having its own delay, amplitude or power, angle-of-arrival, angle-of- departure, and phase shift. Also, different Doppler characteristics (e.g., Doppler shifts) can be observed for moving objects. For example, a first spatial path may be a line-of-sight path or, at least, a path within minimum distance of the electromagnetic waves between the TX radio node and the RX radio node. The first spatial path may have higher amplitude or power of the electromagnetic waves at the RX radio node if compared to higher-order spatial paths, e.g., second or third spatial paths. Higher order spatial paths are typically associated with reflections at physical objects.

[0141] Illustrated in FIG. 4 is a scenario in a cellular network 100 in which the BS 108 and the UE 106 participate in a sensing measurement. While the specific scenario FIG. 4 corresponds to the bi-static sensing topology of FIG. 2, as a general rule, various configurations of sensing measurements are conceivable, e.g. mono-static and multi-static sensing topologies.

[0142] For instance, a node 110 implementing a Sensing Management Function (SeMF) can configure the BS 108 and / or the UE 106 or, more generally, one or more radio nodes, to participate in a sensing measurement. This can include configuration to transmit sensing signals and / or attempt to receive sensing signals and perform measurement and / or to provide measurement reports associated with the sensing signals. Different radio nodes engaging in the sensing measurement can be configured with specific types of measurement techniques, types of measurement reports or, more generally, specific reporting procedures. The measurement report may depend on the type of the radio node. For instance, mobile radio nodes may be configured with a different reporting procedure if compared to static radio nodes. Radio nodes dedicated to the sensing measurement can be configured with a specific reporting procedure. In this example, BS 108 transmits sensing signals 120 (e.g., Sensing Reference Signals, SeRSs) and UE 106 is expected to receive the reflected / scattered sensing signals 122, 124, 126. A communication 128 between the UE 106, BS 108 and the node 110 may be accomplished via wired or wireless communication of the cellular network 100, including e.g. LTE, 5G or 6G data communication or the Internet.

[0143] In detail, FIG. 4 illustrates a downlink(DL)-like bi-static sensing, i.e. the BS 108 acts as a SeRS transmitter, broadcasting SeRSs to its environment, and the UE 106 receives the reflected / scattered SeRSs. An SeRS may comprise an omnidirectional signal, or one or more spatially directed components. Here, each such SeRS component can be referred to as a 'ray'. Three rays 120 in different directions to objects 102, 104 and 114 are illustrated. Each ray then bounces off different passive objects, either the target object 114 or a surrounding object 102 or 104. These interactions change the properties of the reflected rays. The UE 106, in this deployment, works as a SeRS receiver aimed at capturing the reflected rays and estimating the sensing channel. This estimation allows the UE 106 to derive sensing measurements associated with the target object 114, which are then reported to the SeMF 110 for further processing. The SeMF 110 is used to perform various sensing tasks such as object detection, localization and identification using the received measurement report from UE 106. In other scenarios, if the UE 106 is capable and has also obtained all of the required parameters, the UE 106 may also perform various sensing tasks as mentioned above using the obtained measurement results. The UE 106 may use the output of the sensing tasks or report it to the SeMF 110. In other sensing schemes / topologies, the measurement report may also come from the gNB 108. In addition to the measurement processing, the SeMF 110 is also tasked to communicate with the UE 106 via e.g. the 3GPP layer primarily to 1) receive the sensing measurement reports and 2) provide assistance information to support the UE 106 in the sensing measurement. These two aspects will be described in more detail below.

[0144] One of the tasks of an ISAC system may be object identification. This task involves not only detecting the presence of the target object, but also estimating more detailed information such as the object’s type, shape and size. Single-point Radar Cross Section (RCS) measurements may not be sufficient for this task, as they may oversimplify the object’s scattering pattern. To accurately resolve an object's type, shape and size, an ISAC system may consider the object’s 3D properties, including a more complex scattering pattern with multiple scattering points.

[0145] FIG. 5 illustrates a ray tracing simulation of bi-static sensing. In this simulation, a vehicleshaped cube 114 is placed in an environment of a transmitter radio node 108 and a receiver radio node 106. The cube 114 serves as the target object. The transmitter radio node 108 and the receiver radio node 106 may be placed about 10 meters away from the object 114. As shown, multiple scattering points are distributed over the surface of the object 114. These scattering points reflect / scatter the incident rays 120 from the transmitter radio node 108 and reradiate them to the receiver radio node 106 as scattered rays 126. A further ray 502 may pass along a line of sight (LOS) from the transmitter radio node 108 to the receiver radio node 106.

[0146] In general, reflection and scattering are two related but distinct concepts in physics, particularly in the context of wave propagation. Both processes involve the interaction of waves with surfaces or objects, but they differ in how the waves behave after the interaction. Reflection occurs when a wave hits a surface and bounces back without changing direction significantly. The wave is said to be "reflected" by the surface. In reflection, the angle of incidence (the angle at which the wave approaches the surface) equals the angle of reflection (the angle at which the wave leaves the surface). Scattering occurs when a wave interacts with an object or surface and changes direction in multiple ways. Unlike reflection, scattering involves a more random and diffuse redistribution of the wave's energy. The scattered waves can travel in various directions.

[0147] Differences between reflection and scattering are directionality and energy distribution. Reflection involves a single bounce back direction from the surface, whereas scattering involves multiple changes in direction. In reflection, most of the energy is reflected back in a single direction, while in scattering, the energy is distributed over a larger angle.

[0148] For example, as illustrated in FIG. 5, when the transmitter radio node 108 transmits sensing signals 120 towards the target object 114, it may be assumed that each scattering point of the object independently reflects and scatters the incident sensing signals according to its own RCS. The signals reflected and scattered from all scattering points are aggregated, e.g., summing magnitude and phase contributions of each scatter. For example, groups of reflected / scattered signals may be formed based on a substantially the same delay, angle of arrival, and / or Doppler shift. For example, "substantially " the same angle of arrival may mean that the angles of arrival of the reflected / scattered signals do not differ by more than a predefined threshold, e.g., no more than 0.1°, 1°, or 5°. Similarly, "substantially " the same delay may mean that the delays of the reflected / scattered signals do not differ by more than a predefined threshold, e.g., no more than 0.5ns, 1ns, or 5ns. In other words, "substantially " the same may mean that the measurements are within an expected spread among scattering points. The different scattering points of the object may be assigned to the groups. Characteristics of each group, for example summing magnitude and phase, may thus be determined for each scattering point.

[0149] FIG. 6 and FIG. 7 depict a Channel Impulse Response (CIR) of a multiple scattering point measurement captured by the receiver radio node 106 of FIG. 5. The rays 126 from different scattering points are separated / resolved in a delay spectrum. Each peak / sample in the CIR represents a sum of rays sharing the same propagation delay. A sample may comprise a complex value including magnitude and phase of the received sensing signals. In FIG. 6 and FIG. 7 the absolute value of the complex value is plotted against the delay time. The plot in FIG. 6 shows the whole delay spectrum and the plot in FIG. 7 highlights a part of the delay spectrum of the reflection / scattering pattern.

[0150] A first peak 602 in the plot of FIG. 6 represents the direct LOS ray 502 from the transmitter radio node 108 to the receiver radio node 106. A second peak 604 indicates the specular reflection, as it has in this example the shortest propagation time in the group. Additional rays from different scattering points are separated / resolved in the delay spectrum. Following the specular reflection, there are numerous samples 606 representing the scattered rays from multiple scattering points at the object 114. Characteristics of the delay spectrum may be utilized to identify the object 114 as will be described in more detail below in connection with FIG. 8 and FIG. 9.

[0151] In accordance with the above-described ray-tracing simulation principles, it is apparent that multiple SeRSs interact with a target surface at diverse points, resulting in diffuse distributions of times of arrival, angles of incidence, and Doppler shifts within a collection of signals that may form the CIR as illustrated above. Based on the CIR, a group of such signals associated with a single object may be collectively referred to as a "cluster". A process for pattern matching based on CIR measurement is exemplified below, wherein the cluster characteristics are utilized to facilitate identification and analysis.

[0152] To facilitate object identification, the user equipment (UE) or gNB must possess knowledge regarding cluster patterns and establish associations between these patterns and corresponding object types. In various examples, the UE has access to vehicle cluster pattern data and human cluster pattern data. The objective is to estimate the likelihood of similarity between a given cluster pattern and the obtained measurement result, such as a Power Delay Profile (PDP), which may be derived from a Channel Impulse Response (CIR) measurement. By comparing the cluster pattern with the PDP, the receiver radio node, such as UE can then further estimate the probability that the target object corresponds to a vehicle or a human.

[0153] Alternatively, or in addition, the comparison process may also be performed in the network, such as by the SeMF or other network entity. In this case, the PDP and / or cluster pattern data may be transmitted to the SeMF for processing, allowing the network to perform object identification. The results may be provided to a requesting application and / or a UE. This distributed approach can leverage the computing resources of the network to improve the accuracy and efficiency of object identification.

[0154] The comparison process itself may be implemented using various methods, including artificial intelligence (Al)-based techniques. However, as an example, a method involving convolution of the cluster pattern and PDP, followed by normalization operations, is described herein. The resulting outputs can be interpreted as a likelihood function indicating the probability of potential target type candidates at different ranges.

[0155] In Integrated Sensing and Communication (ISAC) systems, object detection and identification can be enhanced through the use of pattern analysis, such as Cluster Pattern Matching (CPM) techniques. A technical challenge in implementing CPM may involve acquiring knowledge of Standard Cluster Patterns (SCPs), which can be achieved by leveraging the capabilities of a Sensing Management Function (SeMF). This requires coordination between radio nodes and the SeMF, including configuration of sensing measurements.

[0156] At the radio node side, the process may involve analyzing the obtained measurement result, such as Channel Impulse Response (CIR) data to separate reflected rays and identifying corresponding cluster patterns associated with specific objects, such as vehicles or humans. By matching these measurements against known cluster patterns, the system can facilitate object detection and identification in a more efficient and accurate manner.

[0157] Cluster pattern matching (CPM) may be considered a specific approach to pattern matching in which patterns are grouped or clustered based on similarities or common features before attempting to match them to input data. The patterns can be organized or "clustered" so that the system can make more efficient matching decisions. This approach can reduce the search space, speed up the matching process, and improve accuracy by focusing on the patterns that are most relevant to a given input. Characteristics or "patterns" that represent a cluster of patterns can be referred to as standard cluster patterns or reference cluster patterns. Each reference cluster pattern can be associated with an object type. A specific object type may have multiple reference cluster patterns, for example, for the same specific object type viewed from different angles, at different distances, or with different surface materials.

[0158] FIGs. 8 and 9 illustrate a cluster pattern matching of a reference cluster pattern and a measurement result of a sensing measurement. The sensing measurement may be based on sensing reference signals scattered at a target object to be identified. To facilitate CPM, a reference cluster patterns library may be predefined and may be provided by the communications network, for example by the sensing management function (SeMF). The library may comprise pre-stored information on various target object types measured under diverse conditions. For example, the reference cluster patterns may be generated based on measurements taken at different incident and outgoing angles of scattering, varying measurement distances (including Transmitter-to-Target and Receiver-to- Target distances), and distinct materials of the target. This ensures that the reference cluster patterns capture a comprehensive representation of the object's characteristics under various scenarios.

[0159] Each reference cluster pattern may be normalized in both power and time domains. For example, the expected Power Delay Profile (PDP) of the Channel Impulse Response (CIR) of each reference cluster pattern may be normalized to facilitate efficient matching with measured CIRs.

[0160] The reference cluster patterns should contain only the reflection pattern of the target object, without any influence from background reflections or noise. To achieve this in practical applications, two measurements may be taken: a first measurement (CIR1) that includes both the target and background, and a second measurement (CIR2) that involves the same setup but with the removal of the target. The reference cluster pattern can be generated by comparing CIR2 and CIR1 , for example by subtracting CIR2 from CIR1, resulting in a pattern that represents the target solely without any influence from the background. This resultant pattern serves as the reference cluster pattern and is used for object identification through CPM.

[0161] The exemplary cluster pattern matching of FIGs. 8 and 9 involves convolution of reference cluster patterns and a PDP derived from a CIR measurement of a target object. The PDP may be normalized such that the signal strength of the scattered sensing signals, which may vary significantly depending on the distance, may be blanked. As a result, only the shape of the PDP of the target object influences the convolution result. As described above, also the reference cluster patterns may be provided in normalized form.

[0162] In FIG. 8, the reference cluster pattern 802 may represent a vehicle, and the PDP 804 may represent the measurement result of a target object which is a vehicle. The reference cluster pattern 802 and the PDP 804 are convoluted. The convolution result 806 may be represented in continuous or discrete form. In FIG. 8, the convolution result 806 is discrete in time. Each solid arrow represents a likelihood that the target object is a vehicle at different delay times of the PDP.

[0163] In FIG. 9, the reference cluster pattern 902 may represent a human. As in FIG. 8, the PDP 804 may represent the measurement result of a target object which is a vehicle. The reference cluster pattern 902 and the PDP 804 are convoluted. The convolution result 906 may be represented in continuous or discrete form. In FIG. 9, the convolution result 906 is discrete in time. Each solid arrow represents a likelihood that the target object is a human at different delay times of the PDP.

[0164] Although in FIGs. 8 and 9 the reference cluster patterns 802, 902 and the PDP 804 are shown as continuous graphs, they may be provided as discrete values for a discrete convolution. A resulting likelihood for each of the reference cluster patterns that the target object is a human or a vehicle may be derived from the convolution results 806 and 906, for example by summing the discrete likelihoods or integrating a continuous convolution result.

[0165] Further pattern matchings between the PDP 804 and further reference cluster patterns may be performed. The further reference cluster patterns may relate to quadrupeds, birds, trees, or vehicles from different views (e.g. front, side, back), for example. Corresponding resulting likelihoods may be derived.

[0166] Thus, for each object type represented by the reference cluster patterns a corresponding likelihood is determined that the measured target object is of that specific type. In the example of FIGs. 8 and 9, the likelihood that the measured target object represented by PDP 804 is a vehicle may be larger than the likelihood that the measured target object represented by PDP 804 is a human.

[0167] The convolution is only one example that can be involved in the pattern matching. Other examples may involve cross-correlation, artificial intelligence (Al) techniques such as neural networks and machine learning (ML) techniques.

[0168] Generally, the pattern matching may be performed by the receiving radio node or UE 106 and / or by a management node of the cellular network, for example by a gNB or a SeMF. Examples for these options will be described in more detail in the following.

[0169] Figure 10 shows a signaling diagram for a scenario where the UE 106 performs CPM.

[0170] In the signaling diagram of FIG. 10, the UE 106 is the receiver radio node that receives the reflected and scattered sensing signals 126, and the gNB 108 is the transmitter radio node that transmits the sensing signals 120. However, this is only an example and in other examples, the gNB 108 may be the receiver radio node and another gNB may be the transmitter radio node, or in further examples, the UE 106 may be the transmitter radio node and another UE or gNB may be the receiver radio node.

[0171] In general, the sensing receiver, i.e. UE 106, may perform a sensing measurement on a target object 114 and a pattern analysis 1012, in particular a CPM, of measurement results of the sensing measurement 1011 to determine a type information regarding the target object 114. The sensing measurement 1011 may be a multiple scattering point measurement. A sensing measurement or sensing estimation report 1014 including a type information for the target object 114 is provided to the management node, i.e. the SeMF 110.

[0172] Thus, apart from transmitting the report 1014, all other signaling may be optional.

[0173] In more detail, the UE 106 may transmit a capability indication 1002 to the SeMF 110. The capability indication 1002 may indicate whether the UE 106 is able to provide type information for a target object, e.g. in view of computation complexity or sensing measurement functionalities (e.g. support of multiple scattering point measurement), object types which can be detected, a number of objects that can be analyzed within one measurement or a certain time duration, supported pattern matching techniques, e.g. convolution, cross-correlation, AI / ML enhanced pattern matching.

[0174] The sensing signal transmitter radio node, in this example the gNB (or TRP in a gNB) 108, may transmit its sensing reference signal transmission (SeRS Tx) configuration 1004 to the SeMF 110. The SeRS Tx configuration may indicate characteristics of the SeRS, for example frequency, power, modulation, timeslots, directions, identifiers etc. SeMF 110 may also receive a similar SeRS Tx configuration from other gNBs. Hence, SeMF 110 may be a central node collecting the SeRS Tx configuration from multiple gNBs.

[0175] Based on the capabilities of the UE 106 and the SeRS Tx configuration, the SeMF 110 may decide that the UE 106 shall perform the object detection and identification. The SeMF 110 may determine and transmit a sensing configuration 1006, including SeRS Tx configuration, to the UE 106, e.g. frequency, modulation, timeslots, identifiers etc.

[0176] When triggered by the SeMF 110, the UE 106 receives one or more measurement requests 1008 to perform sensing measurement and object identification. The measurement request 1008 may represent the trigger for the UE 106 to start performing the sensing measurement 1011. The sensing measurement 1011 may start immediately or according to a (pre-)defined schedule.

[0177] The SeMF 110 can provide a signaling 1010 including pattern reference information and a configuration associated with the pattern reference information to the UE 106. The signaling 1010 may include an indication of one or more designated target types to be detected, such as one or more of vehicle, human, and unmanned aerial vehicle (UAV). In the other examples, this can be specified as “an expected target type” included in the sensing configuration 1006 from the SeMF 110 to the UE 106. The signaling 1010 may further include for each designated target type a corresponding reference cluster pattern or a group of multiple reference cluster pattern variations. This group of multiple reference cluster pattern variations may be associated with the reference cluster pattern variations at different incident / outgoing angles of the scattering, or at different measurement distances. The signaling 1010 may include object type / presence determination criteria, such as a threshold for the estimated target presence probability. An estimated result above the threshold should be considered as target presence. The threshold may depend on characteristics of the sensing signals, such as bandwidth or frequency. In some examples, the signaling 1010 may include a minimum tolerated time gap between two successive matched patterns in CIR or PDP. For example, in some use cases, two targets may appear to be in close proximity in the CIR or PDP, but such an arrangement may be considered to be physically impossible. Therefore, a minimum distance may be set between two consecutive matches.

[0178] Including the above information in the signal 1010 is only an example and in other examples this information may be included in the measurement request 1008, or the sensing configuration 1006. Further, the signaling 1010 may comprise a plurality of messages each including parts of the information.

[0179] Some or all of the information of signaling 1010, in particular the reference cluster patterns or the groups of reference cluster pattern variations, may be either communicated per measurement and CPM occasion, or it may be pre-communicated by SeMF to facilitate multiple measurement and CPM occasions. Pre-communicating this information may be provided via LPP-like or NRPPa-like protocols.

[0180] The sensing measurement 1011 may then be performed by transmitting SeRSs 120 from the gNB 108, and receiving SeRS 126 scattered and reflected by the target object 114. Based on the received SeRS 126 the UE 106 estimates the Channel Impulse Response (CIR). This CIR represents the multipath characteristics of the environment including the target object 114.

[0181] The UE 106 executes the Cluster Pattern Matching (CPM) 1012 method to detect and identify objects within its sensing range. The CPM may involve the following.

[0182] The UE 106 may apply various algorithms to compute the similarity or likelihood between the estimated CIR and multiple reference cluster patterns and optionally variations thereof stored in the reference cluster patterns library. These algorithms may include computing convolution or correlation values between the CIR and reference cluster patterns, performing likelihood estimation techniques to determine the probability of a match, and / or employing Artificial Intelligence / Machine Learning (AI / ML) inference methods to enhance detection accuracy. As a result, type information indicating a type of the target object may be achieved. If multiple target objects are present and detected by the sensing measurement 1011, for each target object a corresponding type information indicating the type of the corresponding target object may be achieved.

[0183] As described above, information derived from the CIR may be used instead of directly using the CIR, for example PDP, delay discrete PDP or a Range-Doppler (R-D) profile. The R-D profile may include information regarding the position of the target object based on the delay as well as a speed information regarding the target object based on Doppler information derived from the received SeRS 126.

[0184] The UE 106 determines that a target object is present if a subset of samples in the estimated CIR successfully matches or exhibits strong likelihood with one of the multiple reference cluster patterns or variations thereof. This match indicates that the measured multipath characteristics correspond to those of a known object type, thereby enabling reliable object detection and identification.

[0185] For each potential target object detected by the received SeRS, the UE 106 may provide a sensing measurement or sensing estimation report 1014 including one or more of the following related to the type information.

[0186] One or more probabilities or likelihoods that the target object corresponds to specific object type(s), estimated through CPM. The UE 106 may utilize its knowledge of target-type- related patterns / filters in either the PDP or the R-D profile to estimate these likelihoods. For example, the reported probabilities may include for each detected target / a set like denoting likelihoods of the target / being vehicle, human, or UAV. In some examples, only the probability for the most likely object type is reported, such as an enumerated indicating the most likely object type and a probability that this object type is correct. In some examples, only the most likely object type is reported, such as an enumerated indicating that the most likely object type is either vehicle, human, or UAV.

[0187] The report 1014 may further include a target Index, i.e. an index or identifier associated with the detected target object, such as a target ID. The target ID may help to trace a specific target object in successive occasions of sensing measurement and CPM. The UE 106 may assign a target index to the detected target object based on other information associated with the target object, such as the one or more probabilities or likelihoods that the target object corresponds to specific object type(s), or a position or speed of the target object as determined based on the sensing measurement 1011.

[0188] A signaling 1016 may be provided which includes samples of the sensing measurement 1011. For example, signaling 1016 may include a group of delay, doppler, and power samples in the PDP or R-D profile, estimated through CPM. By matching the reference cluster pattern with the PDP or R-D profile, the UE 106 may detect a subset of samples that describe a single target object. This information may be used by the SeMF 110 to generate further reference cluster pattern variations or improve existing reference cluster patterns or reference cluster pattern variations.

[0189] Further, signaling 1016 may include an estimation of the probability that an object (target or environment) is present at specific {delay, doppler} positions. This report element can be a soft value between 0 and 1 or a binary value (0 or 1), indicating whether the object is present or not. Alternative variants of this probability might include false detection probability or missed detection probability.

[0190] Signaling 1016 may also include information regarding Peak-to-Noise Ratio (PNR) measurements, i.e. a measurement of the ratio between the power received from the target and the noise level, which can be used as an indicator to detect targets. This may be reported in the form of a Reference Signal Received Quality (RSRQ) measurement based on the received SeRS. To estimate the noise level, the UE 106 may roughly identify a signal arrival region in the R-D profile or PDP that encompasses the SeRS scattered from objects and averages the power outside this region as the noise level.

[0191] FIG. 11 shows a signaling diagram for a scenario where the SeMF 110 performs the CPM.

[0192] In the signaling diagram of FIG. 11, the UE 106 is the receiver radio node that receives the reflected and scattered sensing signals 126, and the gNB 108 is the transmitter radio node that transmits the sensing signals 120. However, this is only an example and in other examples, the gNB 108 may be the receiver radio node and another gNB may be the transmitter radio node, or in further examples, the UE 106 may be the transmitter radio node and another UE or gNB may be the receiver radio node.

[0193] In general, the sensing receiver, i.e. UE 106, may perform a sensing measurement on a target object 114 and provide measurement results of the sensing measurement 1011 in a measurement report 1114 to the SEMF 110 for the SeMF 110 to perform a pattern analysis 1112 of the measurement results to determine a type information regarding the target object 114. The sensing measurement 1011 may be a multiple scattering point measurement.

[0194] Thus, apart from transmitting the measurement report 1114, all other signaling may be optional.

[0195] In more detail, the UE 106 may transmit a capability indication 1102 to the SeMF 110. The capability indication 1102 may indicate whether the UE 106 is able to provide measurement results that may be used by the SeMF for identifying a type of a target object. For example the UE 106 may indicate whether the UE 106 can perform multiple scattering point measurements.

[0196] The sensing signal transmitter radio node, in this example the gNB (or TRP in a gNB) 108, may transmit its sensing reference signal transmission (SeRS Tx) configuration 1104 to the SeMF 110. The SeRS Tx configuration may indicate characteristics of the SeRS, for example frequency, power, modulation, timeslots, directions, identifiers etc. The SeMF 110 may also receive a similar SeRS Tx configuration from other gNBs. Hence, SeMF may be a central node collecting the SeRS Tx configuration from multiple gNBs.

[0197] Based on the capabilities of the UE 106 and the SeRS Tx configuration, the SeMF 110 may determine and transmit a sensing configuration 1106 to the UE 106, e.g. frequency, modulation, timeslots, identifiers etc.

[0198] When triggered by the SeMF 110, the UE 106 receives one or more measurement requests 1108 to perform sensing measurement. The measurement request 1108 may represent the trigger for the UE 106 to start performing the sensing measurement 1011. The sensing measurement 1011 may start immediately or according to a (pre-)defined schedule. The SeMF 110 can provide a signaling 1110 including a configuration associated with the sensing measurement to the UE 106, e.g. an indication of at least one object type to be sensed. The UE 106 may adapt its sensing measurement 1011 accordingly, for example a density of samples may be adapted based on the (expected) object type to be sensed. The configuration associated with measurement may include an expected distance to or an expected position of the target object. The sensing measurement 1011 may be adapted accordingly, for example a receive direction and / or receive sensitivity of a receiver of the UE 106 may be set accordingly. In some examples, configuration associated with sensing measurement 1110 may include a configuration from the UE 106 how to provide the measurement results, for example as CIR, PDP or R-D profile.

[0199] Including the above information in the configuration 1110 is only an example, and in other examples, this information may be included in the measurement request 1108, or the configuration 1110 may comprise a plurality of messages, each of which includes portions of the above information. The configuration 1110 may be communicated after the measurement request 1108, as shown in FIG. 11 , or before the measurement request 1108 is transmitted.

[0200] Some or all of the information of configuration 1110, in particular the object type(s) to be sensed, may either be communicated per sensing measurement 1011, or may be precommunicated by the SeMF 110 to facilitate multiple sensing measurements 1011.

[0201] The sensing measurement 1011 may then be performed by transmitting SeRSs 120 from the gNB 108, and receiving SeRS 126 scattered and reflected by the target object 114.

[0202] Based on the received SeRS 126 the UE 106 estimates the Channel Impulse Response (CIR). This CIR represents the multipath characteristics of the environment including the target object 114.

[0203] The UE 106 provides a measurement report 1114 including the requested measurement results, for example CIR, PDP or R-D profile.

[0204] Based on the measurement result, the SeMF 110 executes the Cluster Pattern Matching (CPM) 1112 method to detect and identify objects within the sensing range of UE 106. The CPM 1112 may involve the following.

[0205] The SeMF 110 may apply various algorithms to compute the similarity or likelihood between the received measurement results, for example CIR, and multiple reference cluster patterns and optionally variations thereof stored e.g. in the reference cluster patterns library. These algorithms may include computing convolution or correlation values between the CIR and reference cluster patterns, performing likelihood estimation techniques to determine the probability of a match, and / or employing Artificial Intelligence / Machine Learning (AI / ML) inference methods to enhance detection accuracy. As described above, information derived from the CIR may be used instead of directly using the CIR, for example PDP, delay discrete PDP or a Range-Doppler (R-D) profile.

[0206] The SeMF 110 determines that a target object is present if a subset of samples in the estimated CIR successfully matches or exhibits strong likelihood with one of the multiple reference cluster patterns or variations. This match indicates that the measured multipath characteristics correspond to those of a known object type, thereby enabling reliable object detection and identification.

[0207] A signaling 1116 from the UE 106 to the SeMF 110 may include a measurement quality value related to the sensing measurement 1011. The measurement quality value may include Peak-to-Noise ratio measurements, i.e. a measurement of the ratio between the power received from the target and the noise level, which can be used as an indicator to detect targets. This may be reported in the form of a Reference Signal Received Quality (RSRQ) measurement based on the received SeRS.

[0208] In the signaling diagrams of FIGs 10 and 11 , the CPM 1012 / 1112 is performed by the UE 106 or by the SeMF 110. However, the CPM 1012 / 1112 may be performed in both the UE 106 and the SeMF 110. For example, the UE 106 may perform the sensing measurement 1011 and may provide the results of the sensing measurement to the SeMF 110. In addition, the UE 106 may perform the CPM 1012 based on some reference cluster patterns and may provide the type information for the target object 11 to the SeMF 110. The SeMF 110 may perform CPM 1112 based on some other reference cluster patterns in parallel and finally determine the type information for the target object 114 based on CPM 1012 and CPM 1112. Thus, CPM may be performed more quickly in parallel at the UAE 106 and the SeMF 110.

[0209] FIG. 12 schematically illustrates an apparatus 1200, e.g., a node or a device. For instance, the apparatus 1200 can implement the SeMF 110. The apparatus 1200 includes a processor 1202 and a memory 1204. The processor 1202 and the memory 1204 form a compute circuitry. The apparatus 1200 also includes a communication interface 1206. The processor 1202 can communicate with other apparatuses via the communication interface 1206. The processor 1202 can load program code from the memory 1204 and execute the program code. The processor 1202 can perform techniques as disclosed herein upon loading and executing the program code. For instance, the processor 1202 can execute the method of FIG. 14 or FIG. 16.

[0210] FIG. 13 schematically illustrates an apparatus 1300, e.g., a node or a device. For instance, the apparatus 1300 can implement the UE 106 or the gNB 108. The apparatus 1300 includes a processor 1302 and a memory 1304. The processor 1302 and the memory 1304 form a compute circuitry. The apparatus 1300 also includes a communication interface 1306 that supports wireless communication via one or more antennas 1308. The processor 1302 can communicate with other apparatuses via the communication interface 1306. The processor 1302 can receive and / or transmit SeRS via the interface 1306 and the antenna(s) 1308. The processor 1302 can load program code from the memory 1304 and execute the program code. The processor 1302 can perform techniques as disclosed herein upon loading and executing the program code. For instance, the processor 1302 can execute the method of FIG. 15 or FIG. 17.

[0211] FIG. 14 is a flowchart 1400 of a method according to various examples. FIG. 14 generally relates to implementation of a sensing measurement and object identification. FIG. 14 specifically relates to the management of a sensing measurement executed by one or more nodes.

[0212] The method of FIG. 14 can be executed by a node of a cellular network. For example, the method can be executed by a compute circuitry of the node of the cellular network. For instance, the method can be executed by a processor upon loading and executing program code that is stored in a memory. For example, the method of FIG. 14 may be executed by a node that is located in a core network of the cellular network. The node may be the apparatus 1200 of FIG. 12. The node may be a management node, i.e. , execute a management function for managing a plurality of sensing measurements and object identifications at multiple radio nodes. For instance, such node may be labeled Sensing Management Function (SeMF). The SeMF can implement a collection of measurement reports, processing of measurement reports and object identification reports, e.g., for localization and / or control / coordination amongst various radio nodes executing sensing measurements. In another example, SeMF can be a new function of the legacy location management function (LMF) as in 5G core network.

[0213] At optional box 1402, the node obtains information indicative of a capability of one or more radio nodes, e.g., one or more UEs 106 and / or one or more base stations, such as gNB 108. The capability is associated with performing sensing measurements, object type identification and / or reporting thereof. For instance, box 1402 can include receiving a higher- layer control message such as sensing-protocol message, or a positioning protocol message (e.g., Third Generation Partnership, 3GPP, Long Term Evolution, LTE, Positioning Protocol, LPP, message or a message having a format related to the LPP message format). In other examples, the capabilities of the one or more radio nodes may be predefined in the network or otherwise provided by the one or more radio nodes, e.g., upon registration with the cellular network, and thus known to the node.

[0214] The capability can be indicative of whether the respective radio node is capable of executing a sensing measurement, e.g., using a certain topology, and an object type identification, e.g., based on patten analysis, such as performing CPM. The capability can be indicative of whether the respective radio node is capable of using a certain type of sensing signal, e.g., a chirped signal or a CDM signal. The capability can be indicative of whether the respective radio node is capable of performing a certain type of sensing measurement technique. Alternatively or additionally, the capability can be indicative of whether the respective radio node is capable of supporting a certain reporting procedure for reporting on a sensing measurement and type identification. For instance, the capability can be indicative of whether the respective radio node can provide a measurement report that includes certain information. The capability can be indicative of whether the respective radio node is capable of performing CPM and providing a report on estimated types for a specific target object, for example probability values indicating that the target object is of a specific object type.

[0215] At optional box 1404, the node provides information indicative of a configuration of the sensing measurement to each of one or more radio nodes. For instance, box 1404 can be responsive to obtaining the information indicative of the capability at box 1402. It would be possible that box 1404 is responsive to a respective sensing request obtained from an application. For instance, the application may implement an object counting, object tracking, object identification or object positioning use case. The application can then request execution of a respective sensing measurement; this can trigger providing the configuration to the one or more radio nodes. In another example, configuration information may contain the SeRS configuration in which the SeRS is to be used for sensing measurement at the sensing receiver. In some examples, the configuration information may be predefined in the network or otherwise provided to the one or more radio nodes, e.g., upon registration with the cellular network, and thus known to the one or more radio nodes.

[0216] The configuration can select between different sensing topologies. For example, the configuration message or the configuration messages can be indicative of a selected one of mono-static sensing topology (cf. FIG. 1), bi-static sensing topology (cf. FIG. 2), and multi-static sensing topology (cf. FIG. 3).

[0217] It would be possible that the configuration determines the radio resources and properties, such as a frequency bandwidth of a transmission of a sensing signal. For instance, it would be possible to specify a number of subcarriers that are to be employed for transmitting the sensing signal. Alternatively or additionally, it would be possible to indicate one or more bandwidth parts that are to be used for transmitting the sensing signal.

[0218] It would be possible that the configuration is indicative of the specific radio resources - e.g., timing information and / or timeslots and / or frequency resources and / or time-frequency resources such as resource blocks or resource elements - to be used for transmitting the sensing signals. As a general rule, the configuration can be determined in accordance with the capabilities of the one or more radio nodes (cf. box 1402).

[0219] At optional box 1406, a sensing measurement request (trigger) is provided to the radio nodes to start the sensing measurement. Along with the sensing measurement request at box 1406, pattern reference information and a configuration associated with the pattern reference information may be provided at optional box 1408. It would be possible to include information of at least one object type to be sensed at box 1408. Reference cluster patterns may also be provided at box 1408. Reference cluster patterns used by a CPM of the UE 106 (see below) may be selected based on the at least one object type to be sensed.

[0220] For instance, certain radio nodes may have limited sensing and identification capabilities. To give an example, certain radio nodes may not be able to process measurement data associated with the sensing measurement and CPM for all reference cluster patterns of a library. A set of object types that are expected to be in the environment of the node may be configured as of the associated configuration and the node performs the CPM for the set of object types only. The optional box 1408 can also be provided prior to box 1406. In some examples, the sensing measurement and object identification may be started autonomously by the radio nodes, e.g., according to a predefined schedule provided in the network or otherwise provided to the radio nodes, e.g., upon registration with the cellular network. Similarly, in various examples, the pattern reference information and the configuration associated with the pattern reference information may be predefined in the network or otherwise provided to the radio nodes, e.g., upon registration with the cellular network. The pattern reference information and the configuration associated with the pattern reference information may be provided once for any following object identifications, for a plurality of following object identifications or for each object identification separately.

[0221] At box 1410, one or more reports including a type information for a target object are obtained from the radio node. The report may include only the most probable type determined for the target object. In some examples, the report may include for each of a plurality of object types, for example the set of object types provided in the configuration associated with the pattern reference information (cf. box 1408), a corresponding likelihood value indicating that the target object is of that corresponding object type.

[0222] It would be possible that each target object is assigned an object identifier such that a specific target object can be easily traced in successive measurements.

[0223] It would be possible that the one or more reports are transparent to a radio-access network of the cellular network. An end-to-end protocol between the node (e.g., implementing a SeMF) and the reporting radio nodes may be established. For instance, higher-layer control messages can be signaled on logical links established between each of the radio nodes and the SeMF. Therefore, other nodes in the radio-access network do not need to process such measurement reports. The measurement reports are transparent to the radio-access network of the cellular network. For example, it can be in a form of sensing protocol message or LPP-like protocol message.

[0224] The one or reports may include timing information of a transmission or reception of the sensing signals. I.e., time stamps can be provided that are linked to the actual execution of a transmission of a sensing signal. This is, in particular, helpful for dynamic environments with frequently changing properties of passive target objects. A time resolution can thereby be increased.

[0225] At optional box 1412, one or more observables of the sensing measurement may be obtained, e.g., a group of the detected delay / doppler / power samples in the R-D profile, estimated by cluster pattern matching in the node. By matching the cluster pattern and the PDP / R-D profile, the node may detect a group of samples describing a single target object. This group of samples may only be a subset of the whole PDP / R-D profile and it may contain the same sample size as the cluster pattern. The obtained samples may be used by the management node to generate further reference cluster patterns or to refine an existing reference cluster pattern.

[0226] A measurement quality value relating to the sensing measurement may be obtained, for example in the report of box 1410 or together with the measurement observables in box 1412. The measurement quality value may include the ratio between power coming from the target object and the noise, i.e., a peak-to-noise ratio, as an indicator to detect a target. For example, this can be in a form of Reference Signal Received Quality (RSRQ) measurement based on the received SeRS. In some examples, the measurement quality value may be indicative of a quality or probability of object presence detection. For example, based on the power, delay and doppler information, the radio node can estimate the probability of an object (target / environment object) being present at specific position, e.g. based on delay and / or Doppler. This report element can be a soft value between 0 and 1 , or can also be a binary value, e.g., either 0 or 1 , indicating whether the object is present or not. Some other variants of probability may also be considered, such as false alarm probability or missed detection probability.

[0227] FIG. 15 is a flowchart of a method 1500 according to various examples. The method of FIG. 15 generally pertains to actions associated with a sensing measurement and pattern analysis such as CPM. FIG. 15 is for use in a radio node participating in a sensing measurement and performing CPM. The node may be the apparatus 1300 of FIG. 13. For instance, the method 1500 of FIG. 15 can be executed by a transmitter radio node or a receiver radio node. The method 1500 can be executed by a processor upon loading and executing program code from a memory. For instance, the method 1500 can be executed by the processor 1302 of the apparatus 1300 upon loading and executing program code from the memory 1304. The method 1500 can be executed by a UE such as the UE 106 or can be executed by a base station such as the gNB 108.

[0228] The method 1500 of FIG. 15 can be inter-related to the method 1400 of FIG. 14.

[0229] At optional box 1502, a capability associated with the sensing measurement is provided to a node of a cellular network to which the radio node is connected. For instance, the capability can be provided to a node implementing a SeMF. Aspects with respect to such signaling of the capability have been previously discussed in connection with FIG. 14 box 1402.

[0230] At optional box 1504, a configuration of the sensing measurement is obtained. For example, one or more configuration messages may be obtained. Box 1504 can be responsive to providing a capability at box 1502. Aspects with respect to such configuration have been previously discussed in connection with FIG. 14: box 1404.

[0231] At optional box 1506, a sensing measurement request may be obtained, for example from SeMF 110. In addition, at optional box 1508, pattern reference information and a configuration associated with the pattern reference information may be obtained. In another example, box 1508 can be performed prior to box 1506. In various examples, box 1508 can be performed each time the radio node obtains the measurement request, or as a just one-time configuration. Aspects with respect to such sensing measurement request and pattern reference information have been previously discussed in connection with FIG. 14: boxes 1406 and 1408, respectively.

[0232] At box 1510, the radio node performs a sensing measurement in accordance with the sensing information obtained in box 1504. For example, the node can perform multiple scattering point measurements.

[0233] At box 1512, the radio node performs a pattern analysis, for example CPM, considering the pattern reference information and associated configuration obtained in box 1508. At box 1514, one or more sensing measurement or sensing estimation reports are provided including a type information for a target object. Aspects with respect to such reports have been previously discussed in connection with FIG. 14: box 1410.

[0234] At optional box 1516, one or more observables of the sensing measurement may be provided. Aspects with respect to such messages have been previously discussed in connection with FIG. 14: box 1412.

[0235] Summarizing, above techniques provide a set of reference cluster patterns, for example in a library in the management node. Each reference cluster pattern is associated to a target type. A reference cluster pattern may be represented like an CIR measurement output, in particular as normalized PDP that may be obtained from a CIR measurement. Reference cluster patterns and additional configuration information regarding the reference cluster patterns may be communicated from the management node to the radio node. The management node may trigger the radio node to perform sensing measurements and pattern analysis. The target object detection and type identification based on the reference cluster patterns, for example by use of CPM optionally in connection with AI / ML, may be performed in the radio node. The radio node provides the result of the pattern analysis, i.e. information regarding the object type of the target object, and may additionally provide conditions of the measurements and quality metrics.

[0236] FIG. 16 is a flowchart 1600 of a method according to various examples. FIG. 16 generally relates to an implementation of a sensing measurement and object identification. FIG. 16 specifically relates to the management of a sensing measurement executed by one or more nodes.

[0237] The method of FIG. 16 can be executed by a node of a cellular network. For example, the method can be executed by a compute circuitry of the node of the cellular network. For instance, the method can be executed by a processor upon loading and executing program code that is stored in a memory. For example, the method of FIG. 16 may be executed by a node that is located in a core network of the cellular network. The node may be the apparatus 1200 of FIG. 12. The node may be a management node, i.e., execute a management function for managing a plurality of sensing measurements and object identifications at multiple radio nodes. For instance, such node may be labeled Sensing Management Function (SeMF). The SeMF can implement a collection of measurement reports, processing of measurement reports and object identification reports, e.g., for localization, and / or control / coordination amongst various radio nodes executing sensing measurements. In another example, SeMF can be a new function of the legacy location management function (LMF) as in 5G core network.

[0238] At optional box 1602, the node obtains information indicative of a capability of one or more radio nodes, e.g., one or more UEs 106 and / or one or more base stations, such as gNB 108. The capability is associated with performing sensing measurements, object type identification and / or reporting thereof. Aspects with respect to such capabilities have been previously discussed in connection with FIG. 14: box 1402.

[0239] In the following, it is assumed the management node comes to the conclusion that the pattern analysis, in particular CPM, is performed at the management node and the radio node provides the sensing measurement. For example, the management node may decide to not perform the pattern analysis in the radio node based on the capability obtained from the radio node. However, in other examples, the management node may decide to not perform the pattern analysis in the radio node based on other criteria, for example network regulations, availabilities of computing resources or power considerations, for example to save energy at the radio node in case of a wireless battery powered radio node.

[0240] At optional box 1604, the node provides information indicative of a configuration of the sensing measurement to each of one or more radio nodes. For instance, box 1604 can be responsive to obtaining the information indicative of the capability at box 1602. Aspects with respect to such configurations have been previously discussed in connection with FIG. 14: box 1404.

[0241] At optional box 1606, a sensing measurement request (trigger) is provided to the radio nodes to start the sensing measurement. Along with the sensing measurement request at box 1606, measurement associated information may be provided in optional box 1608. It would be possible to include information of at least one object type to be sensed, e.g., expected object types, in box 1608. The UE 106 (see below) may be adapt its sensing measurement properties based on the at least one object type to be sensed. In further examples, the measurement associated information may include a distance or direction of an object to be sensed. The UE 106 may be adapt its sensing measurement properties based on distance or direction.

[0242] The optional box 1608 can also be provided prior to box 1606. In some examples, the sensing measurement may be started autonomously by the radio nodes, e.g., according to a predefined schedule provided in the network or otherwise provided to the radio nodes, e.g., upon registration with the cellular network. Similarly, in various examples, the measurement associated information may be predefined in the network or otherwise provided to the radio nodes, e.g., upon registration with the cellular network. The measurement associated information may be provided once for any following sensing measurements, for a plurality of following sensing measurements or for each sensing measurement separately.

[0243] At box 1610, one or more measurement reports of the sensing measurement are obtained from at least one of the radio nodes participating in the sensing measurement. It would be possible that all receiving radio nodes provide a respective measurement report. More than one receiver radio node can be configured to provide a measurement report at a given time and / or within certain time window.

[0244] It would be possible that the one or more measurement reports are transparent to a radio-access network of the cellular network. An end-to-end protocol between the node (e.g., implementing a SeMF) and the reporting radio nodes may be established. For instance, higher- layer control messages can be signaled on logical links established between each of the radio nodes and the sensing management function. Therefore, other nodes in the radio-access network do not need to process such measurement reports. The measurement reports are transparent to the radio-access network of the cellular network. For example, it can be in a form of sensing protocol message or LPP-like protocol message.

[0245] The one or measurement reports may include timing information of a transmission or reception of the sensing signal. I.e., time stamps can be provided that are linked to the actual execution of a transmission of a sensing signal. This is, in particular, helpful for dynamic environments with frequently changing properties of passive target objects. A time resolution can thereby be increased.

[0246] At least one of the one or more measurement reports may include one or more observables for each of multiple multipath components of a radio channel of the sensing signals. For instance, a measurement report can include multiple observables for each of the multiple multipath components. In case of a multiple scattering point measurement, the measurement report may comprise one or more observables associated which each scattering point of the multiple scattering points of the target object sensed by the sensing signals. For example, vectors of corresponding observables associated with each scattering point of the multiple scattering points may be included. In particular, a CIR can be obtained in the measurement report. The CIR may be obtained as a list of observables, for example a list of complex values including amplitude and phase of the multiple scattering point measurement. The list of observables may be specifically adapted by the radio node according to the measurement associated information provided in 1606. Based on the measurement associated information, for example the expected object types, the radio node may provide observables that provide specific characteristics in the CIR to identify the expected object types. The list of specifically adapted observables may be a subset of the whole CIR, i.e. samples specifically describing a specific expected object type.

[0247] It would be possible to obtain a PDP or an R-D profile instead of the CIR.

[0248] Generally, the multiple observables can be one or more of the following: amplitude of the sensing signal at the receiver radio node; phase of the sensing signal at the receiver radio node; power of the sensing signal at the receiver radio node; delay of the sensing signal at the receiver radio node; angle-of arrival of the sensing signal at the receiver radio node; angle-of- departure of the sensing signal at the transmitter radio node; and / or Doppler frequency shift of the sensing signal at the receiver radio node.

[0249] The at least one measurement report can, accordingly, resolve the multiple multipath components of the multiple scattering points. This means that the at least one measurement report does not only provide information on a single one of the multipath components, e.g., the primary path of the channel providing the lowest propagation delay (often associated with line- of-sight propagation); but rather provides information on multiple paths of the channel.

[0250] This enables accurate post-processing at the node. For instance, complex postprocessing algorithms can be executed that detect, locate, track, and in particular identify objects reliably based on such full channel information.

[0251] At optional box 1612, the node may obtain information on a measurement quality of the data in the obtained measurement report. The measurement quality may include Peak-to-Noise Ratio Measurements, i.e. a measurement of the ratio between the power received from the target and the noise level, which can be used as an indicator to detect targets.

[0252] At optional box 1614, the node can perform a pattern analysis on the information of the measurement report. For example, the node can perform a CPM using reference cluster patterns from the library, optionally in connection with AI / ML techniques. The information on the measurement quality obtained at box 1612 may be used to facilitate pattern analysis. FIG. 17 is a flowchart of a method 1700 according to various examples. The method of FIG. 17 generally pertains to actions associated with a sensing measurement and pattern analysis such as CPM. FIG. 17 is for use in a radio node participating in a sensing measurement and providing a measurement reported to a management node for performing CPM. The node may be the apparatus 1300 of FIG. 13. For instance, the method 1700 of FIG. 17 can be executed by a transmitter radio node or a receiver radio node. The method 1700 can be executed by a processor upon loading and executing program code from a memory. For instance, the method 1700 can be executed by the processor 1302 of the apparatus 1300 upon loading and executing program code from the memory 1304. The method 1700 can be executed by a UE such as the UE 106 or can be executed by a base station such as the gNB 108.

[0253] The method 1700 of FIG. 17 can be inter-related to the method 1600 of FIG. 16.

[0254] At optional box 1702, a capability associated with the sensing measurement is provided to a node of a cellular network to which the radio node is connected. For instance, the capability can be provided to a node implementing a SeMF 110. Aspects with respect to such signaling of the capability have been previously discussed in connection with FIG. 16 box 1602.

[0255] At optional box 1704, a configuration of the sensing measurement is obtained. For example, one or more configuration messages may be obtained. Box 1704 can be responsive to providing a capability at box 1702. Aspects with respect to such configuration have been previously discussed in connection with FIG. 16: box 1604.

[0256] At optional box 1706, a sensing measurement request may be obtained, for example from SeMF 110. In addition, at optional box 1708, measurement associated information may be obtained. In another example, box 1708 can be performed prior to box 1706. Aspects with respect to such sensing measurement request and pattern reference information have been previously discussed in connection with FIG. 16: boxes 1606 and 1608, respectively.

[0257] At box 1710, the radio node performs a sensing measurement in accordance with the sensing information obtained in box 1704.

[0258] At box 1712, the radio node provides one or more measurement reports of the sensing measurement. Aspects with respect to such measurement reports have been previously discussed in connection with FIG. 16: box 1610.

[0259] At optional box 1714, the node may provide information on a measurement quality of the data in the provided measurement report. Aspects with respect to such information have been previously discussed in connection with FIG. 16: box 1612.

[0260] Summarizing, above techniques provide a set of reference cluster patterns, for example in a library in the management node. Each reference cluster pattern is associated to a target type. A reference cluster pattern may be represented like an CIR measurement output, in particular as normalized PDP that may be obtained from a CIR measurement. The management node may trigger the radio node to perform sensing measurement and to provide results of the sensing measurement to the management node. A target object detection and type identification based on the reference cluster patterns, for example by use of CPM optionally in connection with AI / ML, may be performed in the management node.

[0261] Summarizing, at least the following EXAMPLES have been disclosed: EXAMPLE 1. A method for use in a radio node (106) of a cellular network (100), the method (1500) comprising: providing (1514), to a management node (110) of the cellular network (100), a report (1014) including a type information for a target object (114), the type information being based on a pattern analysis (1012) of measurement results of a sensing measurement (1011), the sensing measurement (1011) employing sensing signals (120, 126) for sensing the target object (114), the sensing signals (120, 126) being multiplexed with communication signaling of the cellular network (100).

[0262] EXAMPLE 2. The method of EXAMPLE 1, obtaining (1506), from the management node (110) or another node of the cellular network (100), a request (1008) for sensing the target object (114), wherein the type information is provided upon obtaining the request (1008). EXAMPLE 3. The method of EXAMPLE 1 or EXAMPLE 2, wherein the type information comprises an indicator indicative of an object type and optionally an associated probability value.

[0263] EXAMPLE 4. The method of any one of the preceding EXAMPLES, wherein the type information comprises an indicator indicative of multiple object types and associated probability values indicating, for each of the multiple object types, a probability that the target object (114) is of the corresponding object type.

[0264] EXAMPLE 5. The method of any one of the preceding EXAMPLES, wherein the pattern analysis (1012) is based on pattern reference information discriminating between multiple object types.

[0265] EXAMPLE 6. The method of EXAMPLE 5, wherein the pattern reference information comprises at least one reference pattern, wherein each reference pattern of the at least one reference pattern is assigned to an object type.

[0266] EXAMPLE 7. The method of EXAMPLE 6, wherein the pattern analysis (1012) comprises a cluster pattern matching, wherein the at least one reference pattern comprises at least one reference cluster pattern.

[0267] EXAMPLE 8. The method of any one of EXAMPLES 5 to 7, wherein the pattern reference information comprises a machine-learned model.

[0268] EXAMPLE 9. The method of any one of EXAMPLES 5 to 8, further comprising: obtaining (1508), from the management node (110) or another node of the cellular network (100), at least one of the pattern reference information or a configuration associated with the pattern reference information (1010).

[0269] EXAMPLE 10. The method of EXAMPLE 9, wherein the configuration associated with the pattern reference information (1010) comprises an indication of at least one object type to be sensed, wherein the pattern reference information (1010) is selected based on the indication of the at least one object type to be sensed.

[0270] EXAMPLE 11. The method of EXAMPLE 10, wherein the pattern analysis (1012) comprises a comparison between the measurement results and reference patterns comprised in the pattern reference information and assigned to the at least one object type to be sensed.

[0271] EXAMPLE 12. The method of any one of EXAMPLES 9 to 11 , wherein the configuration associated with the pattern reference information (1010) comprises an indication of multiple reference patterns assigned to the at least one object type of the indication.

[0272] EXAMPLE 13. The method of any one of EXAMPLES 9 to 12, wherein the at least one of the pattern reference information or the configuration associated with the pattern reference information (1010) is obtained once for multiple providing of the report (1014) and / or for each providing of the report (1014).

[0273] EXAMPLE 14. The method of any one of EXAMPLES 9 to 13, wherein the configuration associated with the pattern reference information (1010) comprises one or more parameters of the pattern analysis (1012).

[0274] EXAMPLE 15. The method of EXAMPLE 14, wherein the one or more parameters of the pattern analysis (1012) comprise one or more of the following:

[0275] - an evaluation threshold for identifying a certain object type;

[0276] - a timing correlation constraint for discriminating between multiple different objects; and

[0277] - a pattern matching metric for comparing two scattering patterns.

[0278] EXAMPLE 16. The method of any one of EXAMPLES 9 to 15, wherein the configuration associated with the pattern reference information (1010) comprises one or more timing constraints for the sensing measurement (1011).

[0279] EXAMPLE 17. The method of any one of the preceding EXAMPLES, wherein the pattern analysis (1012) comprises a comparison between the measurement results and at least two reference patterns, wherein the at least two reference patterns differ in at least one of: angles of scattering of the sensing signals, surface material of the target object (114), and measurement distances.

[0280] EXAMPLE 18. The method of any one of the preceding EXAMPLES, wherein the measurement results comprise one or more observables, the one or more observables being associated with each scattering point of multiple scattering points of the target object (114) sensed with the sensing signals (120, 126).

[0281] EXAMPLE 19. The method of any one of the preceding EXAMPLES, further comprising: providing (1516) a set of observables from the sensing measurement (1011), on the basis of which the pattern analysis (1012) determined the type information.

[0282] EXAMPLE 20. The method of EXAMPLE 19, wherein the observables comprise at least one of delay information, Doppler shift information, power information, and phase information.

[0283] EXAMPLE 21. The method of any one of the preceding EXAMPLES, wherein the report (1014) comprises a target identifier assigned to the target object (114).

[0284] EXAMPLE 22. The method of any one of the preceding EXAMPLES, wherein the report (1014) comprises a measurement quality value related to the sensing measurement (1011), the measurement quality value indicating a peak-to-noise ratio of a peak power detected for the target object (114) and a noise power.

[0285] EXAMPLE 23. The method of any one of the preceding EXAMPLES, wherein the report (1014) comprises an object presence detection quality value indicating a probability that the target object (114) is present.

[0286] EXAMPLE 24. The method of any one of the preceding EXAMPLES, further comprising: providing (1502), to the management node (110), a capability indication (1002), indicating a capability to provide type information for a target object (114).

[0287] EXAMPLE 25. A method for use in a management node (110) of a cellular network (100), the method (1600) comprising: obtaining (1610), from a radio node (106) of the cellular network (100), a report (1014) including a type information for a target object (114), the type information being based on a pattern analysis (1012) of measurement results of a sensing measurement (1011), the sensing measurement (1011 ) employing sensing signals (120, 126) for sensing the target object (114), the sensing signals (120, 126) being multiplexed with communication signaling of the cellular network (100).

[0288] EXAMPLE 26. The method of EXAMPLE 25, further comprising: providing (1406), to the radio node (106), a request for sensing the target object (114), wherein the type information is obtained upon providing the request.

[0289] EXAMPLE 27. The method of EXAMPLE 25 or EXAMPLE 26, wherein the pattern analysis (1012) is based on pattern reference information discriminating between multiple object types, wherein the method (1400) further comprises: providing (1408), to the radio node, at least one of the pattern reference information or a configuration associated with the pattern reference information (1010), wherein the at least one of the pattern reference information or the configuration associated with the pattern reference information (1010) is provided once for multiple consecutive obtaining (1410) of the report (1014) and / or for each obtaining (1410) of the report (1014).

[0290] EXAMPLE 28. The method of any one EXAMPLES 25 to 27, further comprising: obtaining (1402), from the radio node (106), a capability indication (1002), indicating a capability to provide type information for a target object (114).

[0291] EXAMPLE 29. A method for use in a management node (110) of a cellular network (100), the method (1600) comprising: obtaining (1610), from a radio node (106) of the cellular network (100), measurement results (1114) of a sensing measurement (1011), wherein the sensing measurement (1011) employs sensing signals (120, 126) for sensing a target object (114), the sensing signals (120, 126) being multiplexed with communication signaling of the cellular network (100), determining a type information for the target object (114) based on a pattern analysis (1112) of the measurement results.

[0292] EXAMPLE 30. The method of EXAMPLE 29, further comprising: providing (1606), to the radio node (106), a request (1108) for providing the measurement results.

[0293] EXAMPLE 31. The method of EXAMPLE 29 or EXAMPLE 30, wherein the type information comprises an indicator indicative of an object type and optionally an associated probability value.

[0294] EXAMPLE 32. The method of any one of EXAMPLES 29 to 31 , wherein the type information comprises an indicator indicative of multiple object types and associated probability values indicating, for each of the multiple object types, a probability that the target object (114) is of the corresponding object type.

[0295] EXAMPLE 33. The method of any one of EXAMPLES 29 to 32, further comprising: providing (1608), to the radio node (106), an indication of at least one object type to be sensed, wherein the sensing measurement (1011) is to be performed based on the at least one object type to be sensed.

[0296] EXAMPLE 34. The method of any one of EXAMPLES 29 to 33, wherein the pattern analysis (1112) comprises a cluster pattern matching. EXAMPLE 35. The method of any one of EXAMPLES 29 to 34, wherein the pattern analysis (1112) comprises a comparison between the measurement results and at least two reference patterns, wherein the at least two reference patterns differ in at least one of: angles of scattering of the sensing signals, surface material of the target object (114), and measurement distances.

[0297] EXAMPLE 36. The method of any one of EXAMPLES 29 to 35, wherein the pattern analysis (1112) is based on pattern reference information discriminating between multiple object types, wherein a configuration associated with the pattern reference information comprises one or more parameters of the pattern analysis (1112), wherein the one or more parameters comprise one or more of the following:

[0298] - an evaluation threshold for identifying a certain object type;

[0299] - a timing correlation constraint for discriminating between multiple different objects; and

[0300] - a pattern matching metric for comparing two scattering patterns.

[0301] EXAMPLE 37. The method of any one of EXAMPLES 29 to 36, further comprising: determining a target identifier assigned to the target object (114).

[0302] EXAMPLE 38. The method of any one of EXAMPLES 29 to 37, further comprising: obtaining (1612) a measurement quality value (1116) related to the sensing measurement (1011), the measurement quality value (1116) indicating a peak-to-noise ratio of a peak power detected for the target object (114) and a noise power.

[0303] EXAMPLE 39. The method of any one of EXAMPLES 29 to 38, further comprising: determining an object presence detection quality value indicating a probability that the target object (114) is present.

[0304] EXAMPLE 40. A method for use in a radio node (106) of a cellular network (100), the method (1700) comprising: providing (1712), to a management node (110) of the cellular network (100), measurement results (1114) of a sensing measurement (1011), wherein the sensing measurement (1011) employs sensing signals (120, 126) for sensing a target object (114), the sensing signals (120, 126) being multiplexed with communication signaling of the cellular network (100).

[0305] EXAMPLE 41. The method of EXAMPLE 40, further comprising: obtaining (1706), from the management node (110), a request (1108) for providing the measurement results (1114).

[0306] EXAMPLE 42. The method of EXAMPLE 40 or EXAMPLE 41 , further comprising: obtaining (1708), from the management node (110), an indication of at least one object type to be sensed, wherein the sensing measurement (1011) is performed based on the at least one object type to be sensed.

[0307] EXAMPLE 43. The method of any one of EXAMPLES 40 to 42, further comprising: obtaining (1706), from the management node (110), an indication of at least one target object (114) to be sensed, wherein the sensing measurement (1011) is performed based on the at least one target object (114) to be sensed.

[0308] EXAMPLE 44. The method of any one of EXAMPLES 40-43, further comprising: providing (1714), to the management node (110), a measurement quality value (1116) related to the sensing measurement (1011), the measurement quality value (1116) indicating a peak-to-noise ratio of a peak power detected for the target object (114) and a noise power.

[0309] EXAMPLE 45. The method of any one of EXAMPLES 40 to 44, wherein the measurement results (1114) comprise one or more observables, the one or more observables being associated with each scattering point of multiple scattering points of the target object (114) sensed with the sensing signals (120, 126).

[0310] EXAMPLE 46. A node configured for participating in a sensing measurement (1011) employing sensing signals (120, 126) for sensing a target object (114) , the sensing signals (120, 126) being multiplexed with communication signaling of a cellular network (100), the node (106) comprising compute circuitry (1302, 1304) configured to: provide (1514), to a management node (110) of the cellular network (100), a report (1014) including a type information for the target object (114), the type information being based on a pattern analysis (1012) of measurement results of the sensing measurement (1011). EXAMPLE 47. The node of EXAMPLE 46, wherein the compute circuitry (1302, 1304) is configured to execute the method (1500) of any one of EXAMPLES 1 to 24. EXAMPLE 48. A node of a cellular network (100), the node (110) comprising compute circuitry (1202, 1204) configured to: obtain (1410), from a radio node (106) of the cellular network (100), a report (1014) including a type information for a target object (114), the type information being based on a pattern analysis (1012) of measurement results of a sensing measurement (1011), the sensing measurement (1011) employing sensing signals (120, 126) for sensing the target object (114), the sensing signals (120, 126) being multiplexed with communication signaling of the cellular network (100).

[0311] EXAMPLE 49. The node of EXAMPLE 48, wherein the compute circuitry (1202, 1204) is configured to execute the method (1400) of any one of EXAMPLES 25 to 28.

[0312] EXAMPLE 50. The node of EXAMPLE 48 or EXAMPLE 49, wherein the node (110) is in a core network of the cellular network (100).

[0313] EXAMPLE 51. A node of a cellular network (100), the node (110) comprising compute circuitry (1202, 1204) configured to: obtain (1610), from a radio node (106) of the cellular network (110), measurement results (1114) of a sensing measurement (1011), wherein the sensing measurement (1011) employs sensing signals (120, 126) for sensing a target object (114), the sensing signals (120, 126) being multiplexed with communication signaling of the cellular network (100), and determine a type information for the target object (114) based on a pattern analysis (1112) of the measurement results (1114).

[0314] EXAMPLE 52. The node of EXAMPLE 51 , wherein the compute circuitry (1202, 1204) is configured to execute the method (1600) of any one of EXAMPLES 29 to 39.

[0315] EXAMPLE 53. The node of EXAMPLE 51 or EXAMPLE 52, wherein the node (110) is in a core network of the cellular network (100).

[0316] EXAMPLE 54. A node configured for participating in a sensing measurement (1011) employing sensing signals (120, 126) for sensing a target object (114), the sensing signals (120, 126) being multiplexed with communication signaling of a cellular network (100), the node (106) comprising compute circuitry (1302, 1304) configured to: provide (1712), to a management node (110) of the cellular network (100), measurement results (1114) of the sensing measurement (1011).

[0317] EXAMPLE 55. The node of EXAMPLE 54, wherein the compute circuitry (1302, 1304) is configured to execute the method (1700) of any one of EXAMPLES 40 to 45.

[0318] EXAMPLE 56. A system comprising the node of any one of EXAMPLES 46, 47, 54 and 55 and the node of any one of EXAMPLES 48 to 53.

Claims

CLAIMS1. A method for use in a radio node of a cellular network, the method comprising: providing, to a management node of the cellular network, a report including a type information for a target object, the type information being based on a pattern analysis of measurement results of a sensing measurement, the sensing measurement employing sensing signals for sensing the target object, the sensing signals being multiplexed with communication signaling of the cellular network.

2. The method of claim 1 , obtaining, from the management node or another node of the cellular network, a request for sensing the target object, wherein the type information is provided upon obtaining the request.

3. The method of claim 1 , wherein the type information comprises an indicator indicative of an object type and optionally an associated probability value.

4. The method of claim 1 , wherein the pattern analysis is based on pattern reference information discriminating between multiple object types, wherein the pattern reference information comprises at least one reference pattern, wherein each reference pattern of the at least one reference pattern is assigned to an object type.

5. The method of claim 4, wherein the pattern analysis comprises a cluster pattern matching, wherein the at least one reference pattern comprises at least one reference cluster pattern.

6. The method of claim 4, wherein the pattern reference information comprises a machine-learned model.

7. The method of claim 4, further comprising: obtaining, from the management node or another node of the cellular network, at least one of the pattern reference information or a configuration associated with the pattern reference information.

8. The method of claim 7, wherein the configuration associated with the pattern reference information comprises an indication of at least one object type to be sensed, wherein the pattern reference information is selected based on the indication of the at least one object type to be sensed.

9. The method of claim 7, wherein the configuration associated with the pattern reference information comprises an indication of multiple reference patterns assigned to the at least one object type of the indication.

10. The method of claim 7,wherein the at least one of the pattern reference information or the configuration associated with the pattern reference information is obtained once for multiple providing of the report and / or for each providing of the report.

11. The method of claim 7, wherein the configuration associated with the pattern reference information comprises one or more parameters of the pattern analysis.

12. A method for use in a management node of a cellular network, the method comprising: obtaining, from a radio node of the cellular network, a report including a type information for a target object, the type information being based on a pattern analysis of measurement results of a sensing measurement, the sensing measurement employing sensing signals for sensing the target object, the sensing signals being multiplexed with communication signaling of the cellular network.

13. A method for use in a management node of a cellular network, the method comprising: obtaining, from a radio node of the cellular network, measurement results of a sensing measurement, wherein the sensing measurement employs sensing signals for sensing a target object, the sensing signals being multiplexed with communication signaling of the cellular network, determining a type information for the target object based on a pattern analysis of the measurement results.

14. The method of claim 13, wherein the type information comprises an indicator indicative of an object type and optionally an associated probability value.

15. The method of claim 13, further comprising: providing, to the radio node, an indication of at least one object type to be sensed, wherein the sensing measurement is to be performed based on the at least one object type to be sensed.

16. The method of claim 13, wherein the pattern analysis comprises a cluster pattern matching.

17. The method of claim 13, wherein the pattern analysis comprises a comparison between the measurement results and at least two reference patterns, wherein the at least two reference patterns differ in at least one of: angles of scattering of the sensing signals, surface material of the target object, and measurement distances.

18. The method of claim 13, wherein the pattern analysis is based on pattern reference information discriminating between multiple object types, wherein a configuration associated with the pattern reference information comprises one or more parameters of the pattern analysis, wherein the one or more parameters comprise one or more of the following:- an evaluation threshold for identifying a certain object type;- a timing correlation constraint for discriminating between multiple different objects; and- a pattern matching metric for comparing two scattering patterns.

19. The method of claim 13, further comprising: determining a target identifier assigned to the target object.

20. The method of claim 13, further comprising: obtaining a measurement quality value related to the sensing measurement, the measurement quality value indicating a peak-to-noise ratio of a peak power detected for the target object and a noise power.

Images