Channel state information (CSI) processing unit (PU) usage for data collection

Abstract

In some example embodiments, there may be provided a method comprising transmitting at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are selected from a number of CPU resources which are available for measuring CSI-Reference Signals (RS), wherein the number of CPU resources for measuring CSI-RS are included in or part of a same CPU pool which is shared for data collection and for measuring CSI-RS, receiving at least one data collection configuration for the number of dedicated CPU resources for data collection, and conducting data collection based at least in part on the data collection configuration by using at least one of the number of dedicated CPU resources.

Description

[0001] CHANNEL STATE INFORMATION (CSI) PROCESSING UNIT (PU) USAGE FOR DATA COLLECTION FIELD

[0002] Various example embodiments relate to the field of wireless communication and, in particular to Channel State Information (CSI) Processing Unit (PU) usage for data collection.

[0003] BACKGROUND

[0004] In the field of wireless communication, for data collection, such as Artificial Intelligence (Al) related data collection and / or Machine Learning (ML) related data collection, the use of the Channel State Information (CSI) framework is being discussed. However, there are several challenges and / or ambiguities when using CSI Processing Units (CPUs) resources for data collection. In particular, since the CPU resources for data collection can be dedicated to measuring CSI-Reference Signals (RS), the CPU usage for data logging may be taken into account, identified and / or defined. Furthermore, a CPU duration occupancy that may, e.g. indicate how many CPUs are occupied and / or how long the CPUs may be occupied, may be taken into account, identified and / or defined.

[0005] It may be beneficial, among other things, to at least address and / or overcome one or more of these challenges, in particular to provide and / or ensure, e.g. proper and / or sufficient, handling and / or operation of CPU resources for data collection.

[0006] SUMMARY

[0007] According to some aspects, there is provided the subject matter of the independent claims. Some further aspects, advantages and / or features are defined in the dependent claims, the description and / or the accompanying drawings.

[0008] In a first example aspect, there may be provided an apparatus, comprising:

[0009] at least one processor; and

[0010] at least one memory storing instructions that, when executed by the at least one processor, cause the apparatus at least to perform:transmitting at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are available in addition to a number of CPU resources for measuring CSI- Reference Signals (RS), in particular wherein the number of dedicated CPU resources for data collection are independent of the number of CPU resources for measuring CSI-RS,

[0011] receiving at least one data collection configuration for the number of dedicated CPU resources for data collection, and

[0012] conducting data collection based at least in part on the data collection configuration by using at least one of the number of dedicated CPU resources.

[0013] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect described herein, wherein the number of dedicated CPU resources for data collection are complementary to the number of CPU resources for measuring CSI-RS.

[0014] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the number of dedicated CPU resources are part of a first CPU pool for data collection which is different from or provided separately to a second CPU pool for measuring CSI-RS for at least one of: an Artificial Intelligence (Al) model, or a Machine Learning (ML) model, wherein the second CPU pool includes the number of CPU resources for measuring CSI-RS.

[0015] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the number of dedicated CPU resources for data collection corresponds to or comprises a maximum available number of dedicated CPU resources for data collection.

[0016] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates that the apparatus supports simultaneous measurements for multiple data collection configurations.In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates a limit on the number of dedicated CPU resources available for simultaneous or concurrent logging of collected data.

[0017] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration differs from a CSI measurement configuration with respect to how collected data is to be reported.

[0018] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration is provided in dependence of the indicated number of dedicated CPU resources for data collection.

[0019] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates at least one type of logging that is to be conducted for collected data.

[0020] In an example embodiment, which may be referred to as a first example embodiment of the first example aspect, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates at least one data collection quantity

[0021] In an example embodiment, which may be referred to as a second example embodiment of the first example aspect, there is provided an apparatus, in particular according to the first example embodiment of the first example aspect, wherein the data collection quantity includes at least one of:

[0022] - at least one beam prediction in a spatial domain,

[0023] - at least one beam prediction in a time domain,

[0024] - at least one CSI prediction, or- at least one CSI compression.

[0025] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates a periodic logging of collected data.

[0026] In an example embodiment, there is provided an apparatus, in particular according to any one from the first example aspect to the second example embodiment of the first example aspect and / or any example embodiment in between, wherein the at least one data collection configuration indicates an event-based logging of collected data.

[0027] In an example embodiment, which may be referred to as a third example embodiment of the first example aspect, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the apparatus is further caused to perform:

[0028] initializing a count for the number of dedicated CPU resources for data collection that is occupied to be zero.

[0029] In an example embodiment, there is provided an apparatus, in particular according to the third example embodiment of the first example aspect, wherein the initializing is based at least in part on acknowledging the at least one data collection configuration.

[0030] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the apparatus is further caused to perform:

[0031] starting to count how many of the number of dedicated CPU resources for data collection are occupied in association with conducting the data collection.

[0032] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication is provided when an attach procedure between the apparatus and the network device is completed.In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication is provided as part of a capability exchange procedure between the apparatus and the network device.

[0033] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the data collection is conducted to collect at least one of: Artificial Intelligence (Al) related data, or Machine Learning (ML) related data.

[0034] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the data collection is conducted to collect beam management related data.

[0035] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the apparatus is a User equipment (UE) or wherein the apparatus is comprised in a UE.

[0036] Furthermore, in particular further according to the first example aspect, there may be provided an apparatus, comprising means for:

[0037] transmitting at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are available in addition to a number of CPU resources for measuring CSI- Reference Signals (RS), in particular wherein the number of dedicated CPU resources for data collection are independent of the number of CPU resources for measuring CSI-RS,

[0038] receiving at least one data collection configuration for the number of dedicated CPU resources for data collection, and

[0039] conducting data collection based at least in part on the data collection configuration by using at least one of the number of dedicated CPU resources.

[0040] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect described herein, wherein the number of dedicated CPU resourcesfor data collection are complementary to the number of CPU resources for measuring CSI-RS.

[0041] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the number of dedicated CPU resources are part of a first CPU pool for data collection which is different from or provided separately to a second CPU pool for measuring CSI-RS for at least one of: an Artificial Intelligence (Al) model, or a Machine Learning (ML) model, wherein the second CPU pool includes the number of CPU resources for measuring CSI-RS.

[0042] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the number of dedicated CPU resources for data collection corresponds to or comprises a maximum available number of dedicated CPU resources for data collection.

[0043] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates that the apparatus supports simultaneous measurements for multiple data collection configurations.

[0044] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates a limit on the number of dedicated CPU resources available for simultaneous or concurrent logging of collected data.

[0045] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration differs from a CSI measurement configuration with respect to how collected data is to be reported.

[0046] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments describedherein, wherein the at least one data collection configuration is provided in dependence of the indicated number of dedicated CPU resources for data collection.

[0047] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates at least one type of logging that is to be conducted for collected data.

[0048] In an example embodiment, which may be referred to as a first example embodiment of the first example aspect, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates at least one data collection quantity

[0049] In an example embodiment, which may be referred to as a second example embodiment of the first example aspect, there is provided an apparatus, in particular according to the first example embodiment of the first example aspect, wherein the data collection quantity includes at least one of:

[0050] - at least one beam prediction in a spatial domain,

[0051] - at least one beam prediction in a time domain,

[0052] - at least one CSI prediction, or

[0053] - at least one CSI compression.

[0054] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates a periodic logging of collected data.

[0055] In an example embodiment, there is provided an apparatus, in particular according to any one from the first example aspect to the second example embodiment of the first example aspect and / or any example embodiment in between, wherein the at least one data collection configuration indicates an event-based logging of collected data.In an example embodiment, which may be referred to as a third example embodiment of the first example aspect, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the means are configured for:

[0056] initializing a count for the number of dedicated CPU resources for data collection that is occupied to be zero.

[0057] In an example embodiment, there is provided an apparatus, in particular according to the third example embodiment of the first example aspect, wherein the initializing is based at least in part on acknowledging the at least one data collection configuration.

[0058] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the means are configured for:

[0059] starting to count how many of the number of dedicated CPU resources for data collection are occupied in association with conducting the data collection.

[0060] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication is provided when an attach procedure between the apparatus and the network device is completed.

[0061] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication is provided as part of a capability exchange procedure between the apparatus and the network device.

[0062] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the data collection is conducted to collect at least one of: Artificial Intelligence (Al) related data, or Machine Learning (ML) related data.In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the data collection is conducted to collect beam management related data.

[0063] In an example embodiment, there is provided an apparatus, in particular according to the first example aspect and / or any one of the associated example embodiments described herein, wherein the apparatus is a User equipment (UE) or wherein the apparatus is comprised in a UE.

[0064] In a second example aspect, there may be provided a method, comprising:

[0065] transmitting at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are available in addition to a number of CPU resources for measuring CSI- Reference Signals (RS), in particular wherein the number of dedicated CPU resources for data collection are independent of the number of CPU resources for measuring CSI-RS,

[0066] receiving at least one data collection configuration for the number of dedicated CPU resources for data collection, and

[0067] conducting data collection based at least in part on the data collection configuration by using at least one of the number of dedicated CPU resources.

[0068] In an example embodiment, there is provided a method, in particular according to the second example aspect described herein, wherein the number of dedicated CPU resources for data collection are complementary to the number of CPU resources for measuring CSI-RS.

[0069] In an example embodiment, there is provided a method, in particular according to the second example aspect and / or any one of the associated example embodiments described herein, wherein the number of dedicated CPU resources are part of a first CPU pool for data collection which is different from or provided separately to a second CPU pool for measuring CSI-RS for at least one of: an Artificial Intelligence (Al) model, or a Machine Learning (ML) model, wherein the second CPU pool includes the number of CPU resources for measuring CSI-RS.In an example embodiment, there is provided a method, in particular according to the second example aspect and / or any one of the associated example embodiments described herein, wherein the number of dedicated CPU resources for data collection corresponds to or comprises a maximum available number of dedicated CPU resources for data collection.

[0070] In an example embodiment, there is provided a method, in particular according to the second example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates that the apparatus supports simultaneous measurements for multiple data collection configurations.

[0071] In an example embodiment, there is provided a method, in particular according to the second example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates a limit on the number of dedicated CPU resources available for simultaneous or concurrent logging of collected data.

[0072] In an example embodiment, there is provided a method, in particular according to the second example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration differs from a CSI measurement configuration with respect to how collected data is to be reported.

[0073] In an example embodiment, there is provided a method, in particular according to the second example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration is provided in dependence of the indicated number of dedicated CPU resources for data collection.

[0074] In an example embodiment, there is provided a method, in particular according to the second example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates at least one type of logging that is to be conducted for collected data.

[0075] In an example embodiment, which may be referred to as a first example embodiment of the second example aspect, there is provided a method, in particular according to thesecond example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates at least one data collection quantity

[0076] In an example embodiment, which may be referred to as a second example embodiment of the second example aspect, there is provided a method, in particular according to the first example embodiment of the second example aspect, wherein the data collection quantity includes at least one of:

[0077] - at least one beam prediction in a spatial domain,

[0078] - at least one beam prediction in a time domain,

[0079] - at least one CSI prediction, or

[0080] - at least one CSI compression.

[0081] In an example embodiment, there is provided a method, in particular according to the second example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates a periodic logging of collected data.

[0082] In an example embodiment, there is provided a method, in particular according to any one from the second example aspect to the second example embodiment of the second example aspect and / or any example embodiment in between, wherein the at least one data collection configuration indicates an event-based logging of collected data.

[0083] In an example embodiment, which may be referred to as a third example embodiment of the second example aspect, there is provided a method, in particular according to the second example aspect and / or any one of the associated example embodiments described herein, wherein the method comprises:

[0084] initializing a count for the number of dedicated CPU resources for data collection that is occupied to be zero.

[0085] In an example embodiment, there is provided a method, in particular according to the third example embodiment of the second example aspect, wherein the initializing is based at least in part on acknowledging the at least one data collection configuration.In an example embodiment, there is provided a method, in particular according to the second example aspect and / or any one of the associated example embodiments described herein, wherein the method comprises:

[0086] starting to count how many of the number of dedicated CPU resources for data collection are occupied in association with conducting the data collection.

[0087] In an example embodiment, there is provided a method, in particular according to the second example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication is provided when an attach procedure between the apparatus and the network device is completed.

[0088] In an example embodiment, there is provided a method, in particular according to the second example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication is provided as part of a capability exchange procedure between the apparatus and the network device.

[0089] In an example embodiment, there is provided a method, in particular according to the second example aspect and / or any one of the associated example embodiments described herein, wherein the data collection is conducted to collect at least one of: Artificial Intelligence (Al) related data, or Machine Learning (ML) related data.

[0090] In an example embodiment, there is provided a method, in particular according to the second example aspect and / or any one of the associated example embodiments described herein, wherein the data collection is conducted to collect beam management related data.

[0091] In an example embodiment, there is provided a method, in particular according to the second example aspect and / or any one of the associated example embodiments described herein, wherein method is performed and / or performable by an apparatus, wherein the apparatus is a User equipment (UE) or wherein the apparatus is comprised in a UE.

[0092] In a third example aspect, there may be provided a non-transitory computer readable storage medium comprising program instructions that, when executed by an apparatus,in particular according to the first example aspect and / or any one of the associated example embodiments described herein, causes operations comprising:

[0093] transmitting at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are available in addition to a number of CPU resources for measuring CSI- Reference Signals (RS), in particular wherein the number of dedicated CPU resources for data collection are independent of the number of CPU resources for measuring CSI-RS,

[0094] receiving at least one data collection configuration for the number of dedicated CPU resources for data collection,

[0095] conducting data collection based at least in part on the data collection configuration by using at least one of the number of dedicated CPU resources.

[0096] The third example aspect may further comprise one or more steps and / or features of the method and / or any one of the associated example embodiments recited in the second example aspect.

[0097] In a fourth example aspect, there may be provided an apparatus, comprising:

[0098] at least one processor; and

[0099] at least one memory storing instructions that, when executed by the at least one processor, cause the apparatus at least to perform:

[0100] transmitting at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are selected from a number of CPU resources which are available for measuring CSI-Reference Signals (RS), in particular wherein the number of CPU resources for measuring CSI-RS are included in or part of a same CPU pool which is shared for data collection and for measuring CSI-RS,

[0101] receiving at least one data collection configuration for the number of dedicated CPU resources for data collection, and

[0102] conducting data collection based at least in part on the data collection configuration by using at least one of the number of dedicated CPU resources.In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect, wherein the CPU pool is further for measuring CSI-RS for at least one of: an Artificial Intelligence (Al) model, ora Machine Learning (ML) model.

[0103] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the number of dedicated CPU resources for data collection corresponds to at least a subgroup of the number of CPU resources for measuring CSI-RS.

[0104] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates a different limit for the number of dedicated CPU resources for data collection and for the number of CPU resources which are available for measuring CSI-RS.

[0105] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates a maximum available number of dedicated CPU resources for data collection from the number of CPU resources for measuring CSI-RS.

[0106] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates the number of dedicated CPU resources for data collection to be at least initially zero.

[0107] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates that the apparatus supports simultaneous measurements for multiple data collection configurations.

[0108] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments describedherein, wherein the at least one indication indicates a limit on the number of dedicated CPU resources available for simultaneous or concurrent logging of collected data.

[0109] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration corresponds to or includes a CSI measurement configuration.

[0110] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration is provided in dependence of the indicated number of dedicated CPU resources for data collection.

[0111] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates at least one type of logging that is to be conducted for collected data.

[0112] In an example embodiment, which may be referred to as a first example embodiment of the fourth example aspect, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates at least one data collection quantity

[0113] In an example embodiment, there is provided an apparatus, in particular according to the first example embodiment of the fourth example aspect, wherein the data collection quantity includes at least one of:

[0114] - at least one beam prediction in a spatial domain,

[0115] - at least one beam prediction in a time domain,

[0116] - at least one CSI prediction, or

[0117] - at least one CSI compression.

[0118] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments describedherein, wherein the at least one data collection configuration indicates a periodic logging of collected data.

[0119] In an example embodiment, there is provided an apparatus, in particular according to any one from the fourth example aspect to the first example embodiment of the fourth example aspect and / or any example embodiment in between, wherein the at least one data collection configuration indicates an event-based logging of collected data.

[0120] In an example embodiment, which may be referred to as a second example embodiment of the fourth example aspect, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the apparatus is further caused to perform:

[0121] initializing a count for the number of dedicated CPU resources for data collection that is occupied to be zero.

[0122] In an example embodiment, there is provided an apparatus, in particular according to the second example embodiment of the fourth example aspect, wherein the initializing is based at least in part on acknowledging the at least one data collection configuration.

[0123] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the apparatus is further caused to perform:

[0124] starting to count how many of the number of dedicated CPU resources for data collection are occupied in association with conducting the data collection.

[0125] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication is provided when an attach procedure between the apparatus and the network device is completed.

[0126] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication is provided as part of a capability exchange procedure between the apparatus and the network device.In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the data collection is conducted to collect at least one of: Artificial Intelligence (Al) related data, or Machine Learning (ML) related data.

[0127] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the data collection is conducted to collect beam management related data.

[0128] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the apparatus is a User equipment (UE) or wherein the apparatus is comprised in a UE.

[0129] Furthermore, in particular further according to the fourth example aspect, there may be provided an apparatus, comprising means for:

[0130] transmitting at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are selected from a number of CPU resources which are available for measuring CSI-Reference Signals (RS), in particular wherein the number of CPU resources for measuring CSI-RS are included in or part of a same CPU pool which is shared for data collection and for measuring CSI-RS,

[0131] receiving at least one data collection configuration for the number of dedicated CPU resources for data collection, and

[0132] conducting data collection based at least in part on the data collection configuration by using at least one of the number of dedicated CPU resources.

[0133] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect, wherein the CPU pool is further for measuring CSI-RS for at least one of: an Artificial Intelligence (Al) model, ora Machine Learning (ML) model.

[0134] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments describedherein, wherein the number of dedicated CPU resources for data collection corresponds to at least a subgroup of the number of CPU resources for measuring CSI-RS.

[0135] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates a different limit for the number of dedicated CPU resources for data collection and for the number of CPU resources which are available for measuring CSI-RS.

[0136] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates a maximum available number of dedicated CPU resources for data collection from the number of CPU resources for measuring CSI-RS.

[0137] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates the number of dedicated CPU resources for data collection to be at least initially zero.

[0138] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates that the apparatus supports simultaneous measurements for multiple data collection configurations.

[0139] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates a limit on the number of dedicated CPU resources available for simultaneous or concurrent logging of collected data.

[0140] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration corresponds to or includes a CSI measurement configuration.In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration is provided in dependence of the indicated number of dedicated CPU resources for data collection.

[0141] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates at least one type of logging that is to be conducted for collected data.

[0142] In an example embodiment, which may be referred to as a first example embodiment of the fourth example aspect, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates at least one data collection quantity

[0143] In an example embodiment, there is provided an apparatus, in particular according to the first example embodiment of the fourth example aspect, wherein the data collection quantity includes at least one of:

[0144] - at least one beam prediction in a spatial domain,

[0145] - at least one beam prediction in a time domain,

[0146] - at least one CSI prediction, or

[0147] - at least one CSI compression.

[0148] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates a periodic logging of collected data.

[0149] In an example embodiment, there is provided an apparatus, in particular according to any one from the fourth example aspect to the first example embodiment of the fourth example aspect and / or any example embodiment in between, wherein the at least one data collection configuration indicates an event-based logging of collected data.In an example embodiment, which may be referred to as a second example embodiment of the fourth example aspect, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the means are configured for:

[0150] initializing a count for the number of dedicated CPU resources for data collection that is occupied to be zero.

[0151] In an example embodiment, there is provided an apparatus, in particular according to the second example embodiment of the fourth example aspect, wherein the initializing is based at least in part on acknowledging the at least one data collection configuration.

[0152] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the means are configured for:

[0153] starting to count how many of the number of dedicated CPU resources for data collection are occupied in association with conducting the data collection.

[0154] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication is provided when an attach procedure between the apparatus and the network device is completed.

[0155] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication is provided as part of a capability exchange procedure between the apparatus and the network device.

[0156] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the data collection is conducted to collect at least one of: Artificial Intelligence (Al) related data, or Machine Learning (ML) related data.In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the data collection is conducted to collect beam management related data.

[0157] In an example embodiment, there is provided an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, wherein the apparatus is a User equipment (UE) or wherein the apparatus is comprised in a UE.

[0158] In a fifth example aspect, there may be provided a method, comprising:

[0159] transmitting at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are selected from a number of CPU resources which are available for measuring CSI-Reference Signals (RS), in particular wherein the number of CPU resources for measuring CSI-RS are included in or part of a same CPU pool which is shared for data collection and for measuring CSI-RS,

[0160] receiving at least one data collection configuration for the number of dedicated CPU resources for data collection, and

[0161] conducting data collection based at least in part on the data collection configuration by using at least one of the number of dedicated CPU resources.

[0162] In an example embodiment, there is provided a method, in particular according to the fifth example aspect, wherein the CPU pool is further for measuring CSI-RS for at least one of: an Artificial Intelligence (Al) model, ora Machine Learning (ML) model.

[0163] In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the number of dedicated CPU resources for data collection corresponds to at least a subgroup of the number of CPU resources for measuring CSI-RS.

[0164] In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates a different limit for the number ofdedicated CPU resources for data collection and for the number of CPU resources which are available for measuring CSI-RS.

[0165] In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates a maximum available number of dedicated CPU resources for data collection from the number of CPU resources for measuring CSI-RS.

[0166] In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates the number of dedicated CPU resources for data collection to be at least initially zero.

[0167] In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates that the apparatus supports simultaneous measurements for multiple data collection configurations.

[0168] In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication indicates a limit on the number of dedicated CPU resources available for simultaneous or concurrent logging of collected data.

[0169] In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration corresponds to or includes a CSI measurement configuration.

[0170] In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration is provided in dependence of the indicated number of dedicated CPU resources for data collection.In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates at least one type of logging that is to be conducted for collected data.

[0171] In an example embodiment, which may be referred to as a first example embodiment of the fifth example aspect, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates at least one data collection quantity

[0172] In an example embodiment, there is provided a method, in particular according to the first example embodiment of the fifth example aspect, wherein the data collection quantity includes at least one of:

[0173] - at least one beam prediction in a spatial domain,

[0174] - at least one beam prediction in a time domain,

[0175] - at least one CSI prediction, or

[0176] - at least one CSI compression.

[0177] In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one data collection configuration indicates a periodic logging of collected data.

[0178] In an example embodiment, there is provided a method, in particular according to any one from the fifth example aspect to the first example embodiment of the fifth example aspect and / or any example embodiment in between, wherein the at least one data collection configuration indicates an event-based logging of collected data.

[0179] In an example embodiment, which may be referred to as a second example embodiment of the fifth example aspect, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the method comprises:initializing a count for the number of dedicated CPU resources for data collection that is occupied to be zero.

[0180] In an example embodiment, there is provided a method, in particular according to the second example embodiment of the fifth example aspect, wherein the initializing is based at least in part on acknowledging the at least one data collection configuration.

[0181] In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the method comprises:

[0182] starting to count how many of the number of dedicated CPU resources for data collection are occupied in association with conducting the data collection.

[0183] In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication is provided when an attach procedure between the apparatus and the network device is completed.

[0184] In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the at least one indication is provided as part of a capability exchange procedure between the apparatus and the network device.

[0185] In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the data collection is conducted to collect at least one of: Artificial Intelligence (Al) related data, or Machine Learning (ML) related data.

[0186] In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments described herein, wherein the data collection is conducted to collect beam management related data.

[0187] In an example embodiment, there is provided a method, in particular according to the fifth example aspect and / or any one of the associated example embodiments describedherein, wherein method is performed and / or performable by an apparatus, wherein the apparatus is a User equipment (UE) or wherein the apparatus is comprised in a UE.

[0188] In a sixth example aspect, there may be provided a non-transitory computer readable storage medium comprising program instructions that, when executed by an apparatus, in particular according to the fourth example aspect and / or any one of the associated example embodiments described herein, causes operations comprising:

[0189] transmitting at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are selected from a number of CPU resources which are available for measuring CSI-Reference Signals (RS), in particular wherein the number of CPU resources for measuring CSI-RS are included in or part of a same CPU pool which is shared for data collection and for measuring CSI-RS,

[0190] receiving at least one data collection configuration for the number of dedicated CPU resources for data collection, and

[0191] conducting data collection based at least in part on the data collection configuration by using at least one of the number of dedicated CPU resources.

[0192] The sixth example aspect may further comprise one or more steps and / or features of the method and / or any one of the associated example embodiments recited in the fifth example aspect.

[0193] BRIEF DESCRIPTION OF THE DRAWINGS

[0194] Some example embodiments will now be described with reference to the accompanying drawings.

[0195] A full and enabling disclosure to one of ordinary skill in the art is set forth more particularly in the remainder of the specification including reference to the accompanying drawings wherein:

[0196] FIG. 1 shows an example signaling according to subject-matter described herein;

[0197] FIG. 2A shows an example signaling according to subject-matter described herein;FIG. 2B shows an example signaling according to subject-matter described herein, in particular according to Fig. 2A;

[0198] FIG. 3A shows an example signaling according to subject-matter described herein;

[0199] Fig. 3B shows an example signaling according to subject-matter described herein, in particular according to Fig. 3A;

[0200] Fig. 4 shows a flow chart of an example aspect of the subject-matter described herein; and

[0201] Fig. 5 shows a flow chart of an example aspect of the subject-matter described herein.

[0202] DETAILED DESCRIPTION

[0203] Reference will now be made in detail to the various embodiments, one or more examples of which are illustrated in the figures. Within the following description of the drawings, the same reference numbers may refer to same components. Generally, only the differences with respect to individual embodiments may be described. Each example may be provided by way of explanation and is not meant as a limitation. Further, features illustrated or described as part of one embodiment can be used on or in conjunction with other embodiments to yield yet a further embodiment. It is intended that the description includes such modifications and variations.

[0204] The drawings are schematic drawings which are not drawn to scale. Some elements in the drawings may have dimensions which are exaggerated for the purpose of highlighting aspects of the present disclosure and / or for the sake of clarity of presentation.

[0205] For Artificial Intelligence (Al) and / or Machine Learning (ML) enhancements related to beam management, two sub-use cases may be identified: i) beam prediction in the spatial domain (BM-Case1) and ii) beam prediction in the time domain (BM-Case2). It may bebeneficial, among other things, to support a reduced, e.g. Reference Signal (RS), overhead and / or lower beam measurements and / or reporting latency.

[0206] The following may provide some non-limiting background for Network (NW)-side model data collection, e.g. as in Rel-19, for which BM-Case1 and BM-case2 may have been investigated for offline training of Artificial Intelligence (Al) and / or Machine Learning (ML) models. For example, different base stations, e.g. gNBs, may collect data, in particular based on Layer 1 (L1) - Reference Signal Received Power (RSRP) measurement reports, e.g. corresponding to different beams, for example from different apparatuses, e.g. User Equipment's. It may be assumed that a base station, e.g. gNB, from the same vendor may be able to transmit reported measurements of all the cells to a server, for instance, data server, e.g. operated by the network vendor(s), so that this server may aggregate received data, e.g. from multiple base stations, e.g. gNBs, to form at least one training dataset. With a larger dataset, the base station, e.g. gNB, may be able to train a more general model, in particular to cover a large geographical area or multiple sites / locations. The data may be pre-processed, for example using normalization, scaling, encoding etc., in particular before the model training.

[0207] For NW-sided beam prediction inference operation, the following non-limiting background is provided:

[0208] L1 RSRP and / or beam ID can be collected to support the beam management data collection, e.g. for offline training.

[0209] the apparatus, e.g. the UE, may store the logged training data at Access Stratum (AS) layer.

[0210] Reporting of multiple instances of logged L1 measurement result, e.g. via a RRC messages from the apparatus, e.g. the UE, to the base station, e.g. the gNB.

[0211] Measurements and / or logging can be controlled, e.g. based on power state of the apparatus, e.g. the UE.

[0212] Periodic and / or event-triggered data logging may be supported, Support the use of Layer 3 (L3) measurements, e.g. event-triggered, On-demand reporting may be supported,When the apparatus, e.g. the UE, reaches its buffer limitation, the apparatus, e.g. the UE, may stop measurement for data collection purposes and / or logging.

[0213] The following may provide some non-limiting background on how Layer 1 (L1) measurement can be logged, e.g. based on Layer 3 (L3) events:

[0214] Layer 3 (L3) events can be configured to the apparatus, e.g. the UE, which may allow the apparatus, e.g. the UE, to trigger Layer 3 (L3) measurement(s), e.g. based on the configured events (for example a A1 event, where cell quality is good). The apparatus, e.g. the UE, may also be configured with an indication, e.g. to report Layer 3 (L3) measurement(s), in particular where the configured event is being no longer hold.

[0215] Logging of Channel State Information Reference Signal(s) (CSI-RS) in the apparatus, e.g. the UE, can have two possibilities:

[0216] -- Scenario A.1 (explicit): The apparatus, e.g. the UE, may start logging Layer 1 (L1) measurement(s), e.g. when it receives an explicit indication from the network, in particular after the network receives Layer 3 (L3) measurement(s). It may be up to the network to decide whether it may activate the logging of Layer 1 (L1) measurements at the apparatus, e.g. the UE. The explicit indication can be either activating the configured resources or re-configuring Layer 1 (L1) measurements.

[0217] -- Scenario A.2 (implicit): The apparatus, e.g. the UE, may start logging Layer 1 (L1) measurements, e.g. associated with previously configured Layer 1 (L1) measurements without having any explicit indication from network. Rather, the indication may be implicit, e.g. an implicit indication which is derivable.

[0218] When the event no longer holds, that is the apparatus, e.g. the UE, may be exiting the Layer 3 (L3) events, the apparatus, e.g. the UE, can send Layer 3 (L3) measurement(s) to the network, e.g. to indicate that the event is not (nolonger) valid anymore. The apparatus, e.g. the UE, may need to stop logging of Layer 1 (L1) measurements. The stopping of logging may have several directions / possibilities:

[0219] -- Scenario B.1 (explicit): The apparatus, e.g. the UE, may send Layer 3 (L3) reports to the network. Upon receiving the Layer 3 (L3) report(s), the network can explicitly send an indication (e.g. explicit indication) to stop logging. It may be up to the network, e.g. when to send this indication. The explicit indication can be either deactivating the configured resources or de-configuring Layer 1 (L1) measurement(s).

[0220] -- Scenario B.2 (implicit): The apparatus, e.g. the UE, can stop logging of the configured Layer 1 (L1) measurements without having any explicit indication from the network. Rather, the indication may be implicit, e.g. an implicit indication which is derivable. This could depend on the stopping criteria, for example. Examples of stopping criteria may include a timer expiring, a low power state detected etc.

[0221] However, using the Channel State Information (CSI) Processing Units (CPUs), in particular of the legacy CSI reporting framework, for data collection may be challenging. For example, for Artificial Intelligence (Al) and / or Machine Learning (ML) related data collection, e.g. for beam management, for example for a network-sided model, the Channel State Information (CSI) (reporting) framework, may be used. In particular the same or related concepts as described herein in relation to a network-side model, may also be applicable to apparatus-side, e.g. UE-side, model(s), in particular if logging is supported and the network is collecting the measurement(s). To do so, several aspects may be clarified, set and / or defined to reuse the legacy Channel State Information (CSI) framework, e.g. effectively. On the other hand, a new information element or the extension of Radio Resource Management (RRM) framework (e.g. Layer 3 (L3)), may require further explanation of the Channel State Information (CSI) resources, e.g. within the legacy Channel State Information (CSI) (reporting) framework.

[0222] Based on the legacy Channel State Information (CSI) framework, the apparatus, e.g. the UE, may assume a number of CSI processing units (CPUs) to be occupied (e.g., whichmay exemplarily be referred to as occupied CPUsor"O_CPU"), in particular for a Channel State Information (CSI) report, e.g. for certain time duration(s) and / or period(s). For example, TS 38.214 defines how the apparatus, e.g. the UE, may use a max limit on a number of CPUs (which may exemplarily be referred to as "N_CPUs"), e.g. when calculating Channel State Information (CSI), e.g. for Channel State Information (CSI) reporting. The number of CPUs (e.g. O_CPUs) may be defined considering various contexts (such as reporting quantities (e.g., Layer 1 Reference Signal Received Power (L1-RSPR), Precoding Matrix Indicator (PMI) etc.), configuration set-up(s) (e.g., Channel State Information (CSI)-Reference Signal port(s), etc.). For example, the time duration(s) for CPU occupation(s) may be clarified in a specification. An aperiodic Channel State Information (CSI) report may occupy CPU(s) from a first symbol after the Physical Downlink Control Channel (PDCCH) triggering the Channel State Information (CSI) report until a last symbol of the scheduled Physical Uplink Shared Channel (PUSCH), e.g. carrying the report. A periodic or semi-persistent Channel State Information (CSI) report may occupy CPU(s) from a first symbol of the earliest one of each Channel State Information (CSI) Reference Signal (CSI-RS) and / or Synchronization Signal Block (SSB) resource, respective latest CSI-RS / SSB occasion, e.g. no later than the corresponding Channel State Information (CSI) reference resource, until a last symbol of the, in particular configured PUSCH / PUCCH, e.g. carrying the report. However, in configurations enabling the data collection, e.g. in / forbeam management, the apparatus, e.g. the UE may support to collect L1-RSRP and / or beam ID within an Access Stratum (AS) layer buffer. The apparatus, e.g. the UE, may report multiple instances of e.g. L1 logged measurements, in particular via RRC message to the base station, e.g. gNB. From the apparatus', e.g. the UE's, perspective, e.g. during data collection stage, the apparatus, e.g. the UE, may not be expected to report Layer 1 (L1) measurements via Layer 1 (L1) reports. However, the apparatus, e.g. the UE, may be expected to measure L1-RSRPs, in particular for the beams and log them in the AS layer buffer. Therefore, similar to what was done for other types of CSI reports, if the CSI-ReportConfig may be used for the network-sided- and / or UE-sided- (e.g. if supported) data collection purposes, it may be important to define the O_CPU and N_CPU usage and / or related timelines.

[0223] In particular, the subject-matter described herein may address the following:

[0224] In particular, since the CPU resources for data collection can be dedicated to measuring CSI-Reference Signals (RS), the CPU usage for data logging may be takeninto account, identified and / or defined. This may be done, as outlined herein, based at least in part on periodic and / or event-triggered logging of measurements.

[0225] Furthermore, a CPU duration occupancy that may, e.g. indicate how many CPUs are occupied and / or how long the CPUs may be occupied, may be taken into account, identified and / or defined.

[0226] It may be beneficial, among other things, to at least address and / or overcome one or more of these challenges, in particular to provide and / or ensure, e.g. proper and / or sufficient, handling and / or operation of CPU resources for data collection.

[0227] According to subject-matter described herein, the apparatus, e.g. the UE, may transmit or provide at least one indication, e.g. to a network device. The indication may be at least one indication (or more), or the indication may correspond to a single indication having multiple parts. In other words, there may be a first indication or at least a first part of the indication indicating a first information, and a second indication or second part of the indication indicating a second information. In an example, the at least one indication may indicate a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are available in addition to a number of CPU resources for measuring CSI- Reference Signals (RS). For example, the at least one indication, e.g. a first part thereof or a first indication, may indicate e.g. that dedicated CPU resources for data collection are available in addition to a number of CPU resources for measuring CSI-RS and, e.g. a second part thereof or a second indication, may indicate the number of the dedicated CPU resources for data collection which are available. However, the indication may, in addition or alternatively, also indicate other / further features and / or elements. The indication may be an explicit indication, e.g. (directly) included in, embedded in or encoded in a message, e.g. directly readable by the apparatus, e.g. the UE. It may be conceivable to send / transmit the indication directly, e.g. without embedding it in another message or the like. The indication may be a flag or pointer which may be included or encoded in the message, e.g. using at least one Information Element (IE) and / or one or more bit(s). The indication may also be an implicit indication, e.g. derivable from the message itself, e.g. the apparatus, e.g. the UE, may conclude to the indication based at least in part on the message, "based at least in part on" the indication may, for example, mean based on one or more part(s) of the indication. In addition or alternatively thereto, it may mean, forexample, based on the indication, e.g. as a whole, (or parts thereof) but also based on additional information and / or additional information element(s), parameters or data.

[0228] As mentioned, the at least one indication may be located in, embedded in or included at least one message or sent separately and / or independently therefrom. The at least one indication may, for example, be transmitted or provided in connection with or as part of at least one attach procedure, e.g. a UE attach procedure, in particular between the apparatus, e.g. the UE, and the network device. For example, the indication may be transmitted when the attach procedure is completed. An attach procedure may be regarded as a process by which the apparatus, e.g. the UE connects (or establishes a connection) to the network device, e.g. to establish a session with the network device. In addition or alternatively thereto, the at least one indication may be transmitted or provided in connection with or as part of a capability exchange procedure, in particular between the apparatus, e.g. the UE, and the network device. A capability exchange process may ensure that the network device (e.g. the gNB) understands or is informed about the apparatus', e.g. the UE's, supported features. The capabilities may be signaled to the network device in, in particular regular or repeated, capability exchange procedure(s) between the apparatus and the network.

[0229] "Channel State Information" or CSI may refer to the information about a state, condition and / or status of a channel, in particular that an apparatus, e.g. a UE, may report, e.g. to the network or network device (e.g. gNB in 5G). CSI may provide insights into the quality of a channel, e.g. enabling the network to optimize data transmission (e.g., selecting the best modulation scheme, power control, beamforming, and / or MIMO techniques etc.). The CSI framework may include a configuration of CSI reporting, e.g. the types of CSI that can be reported, and / or the condition(s) under which these reports may be generated. The specifications may detail the coding and / or structure of the CSI reports. The specifications may also outline the interactions between the CSI reporting and / or the scheduling of resources. A CSI reporting framework may allow the UE to send channel quality feedback. CSI reporting framework may support multiple reporting modes, in particular including one or more of: a i) periodic reporting mode, e.g. sending reports at fixed and / or regular / continuous (time) intervals, e.g. with a defined periodicity ora certain (time) slots, e.g. allowing for substantially constant / recurring feedback, ii) a semi-persistent reporting mode, e.g. CSI reports may be sent at predefined intervals, but the reporting instancescan be activated or deactivated, e.g. by the network device, as needed, which may reduce unnecessary reports, and / or iii) an aperiodic reporting mode, e.g. CSI reports may be sent only when (explicitly) requested by the network device, e.g. it may dynamically trigger a CSI report whenever needed. The CSI reporting framework may allow the network to balance the trade-off between reporting overhead and / or the need for timely channel state information. Periodic reporting may be used for scenarios where the channel conditions are relatively stable, while aperiodic reporting may be triggered by specific events and / or conditions that may require, in particular immediate attention. The CSI reports can include a variety of types, (reporting) quantities and / or metrics such as: a CQI, which may represent or be associated with a Modulation Coding Scheme (MCS) that can be used; a PMI, which may suggest and / or be associated with precoding matrix, and Rl, which may indicate a number of data streams (or layers) that can be supported. The CSI reports may provide detailed insights into the channel characteristics and / or (optimal) transmission parameters. The framework may specify how CSI is to be reported and / or utilized, in particular to optimize a performance of a radio access network (RAN). It may include mechanisms for the apparatus, e.g. the UE, to report channel conditions, which may be useful for adaptive modulation and / or coding, as well as for beamforming techniques that improve signal quality and / or coverage.

[0230] A "Synchronization Signal Block", "synchronization signal / PBCH block" or "SSB" may be regarded as a component or signaling or, it may be at least part of a signaling, that may be used in wireless communication, such as for example 5G or NR, to facilitate a synchronization or synchronization process between the apparatus, e.g. the UE, and a base station, e.g. of a network. The SSB may provide signals, channels and / or information that enable the device to, in particular initially, connect and / or stay synchronized with the base station, allowing it, for example, to communicate effectively over the network. The at least one SSB may be part of a signaling in which a plurality of SSBs are transmitted, e.g. one after another, successively, consecutively and / or periodically. For example, the signaling may include at least a first SSB and at least a second SSB, which for example may be transmitted (e.g. timely) after the first SSB. This may be also referred to as a sequence or series of SSBs. It may be possible to have SSBs transmitted in different modes, e.g. an always-on mode, where the base station e.g. continuously transmits SSBs, e.g. at a periodic interval, or an on-demand mode, e.g. where the base station transmits SSBs only when needed.A "Channel State Information (CSI) Processing Unit (CPU) resource(s)" or short "CPU resource(s)" may be regarded as processing unit(s), hardware unit(s) and / or computational resource(s) / unit(s), e.g. including also logical resource(s), that may be included or provided in the apparatus, e.g. the UE. In general, a CPU may be responsible for processing Channel State Information (CSI) reports, measuring channel quality, e.g. using CSI-RS (Channel State Information Reference Signals), computing CSI metrics, such as PMI (Precoding Matrix Indicator), CQI (Channel Quality Indicator), and / or L1-RSRP (Reference Signal Received Power), and reporting CSI, e.g. to the network device (e.g. a gNB), in particular via PUSCH / PUCCH.

[0231] "Dedicated" CPU resource(s) may be regarded as CPU resource(s), e.g. a dedicated, processing unit or hardware, in particular separate, additional, independent and / or different to N_CPU. The dedicated CPU resource(s) may correspond to the CPU resources, but may be configured, adapted, allocated, defined, adjusted, set, assigned and / or provided for / to one or more specific or dedicated use case(s). The dedicated CPU resource(s) may also be included or provided in the apparatus, e.g. the UE. For example, the specific or dedicated use case may be allocated to the dedicated CPU in addition to other, standard, legacy and / or conventional CSI tasks and / or CSI use cases of the CPU resource, such as measuring CSI-RS, CSI reporting etc. or instead of other, standard, legacy and / or conventional CSI tasks and / or CSI use cases of the CPU resource. For example, the dedicated CPU resource may be allowed, approved and / or cleared, e.g. to be occupied for data collection purposes.

[0232] According to subject-matter described herein, the number of dedicated CPU resources for data collection may be independent of, separate from, individual to and / or different from a number of CPU resources for measuring CSI-RS. The number of dedicated CPU resources for data collection may be, e.g. provided, complementary to, supplementary to, extra to and / or further to the number of CPU resources for measuring CSI-RS. The number of dedicated CPU resources may be part of or included in a first CPU pool for data collection. The first CPU pool may be different from or provided separately, additionally and / or independently to a second CPU pool for measuring CSI-RS for at least one of: an Artificial Intelligence (Al) model, ora Machine Learning (ML) model. The second CPU pool may include the number of CPU resources for measuring CSI-RS. In other words, the CPUresources may be organized, divided, managed, categorized, arranged, combined, joined, bundled, pooled and / or gathered into one or more "pool(s)" which may be referred to as "CPU pool(s)". A pool or CPU pool may be regarded as a collection, set, group, or conglomeration / agglomeration of CPU resources, e.g. a first pool for which may include a number, a plurality, a quantity or multiple dedicated CPU resources CPU pool for data collection and / or a second pool which may include a number, a plurality, a quantity or multiple CPU resources for measuring CSI-RS. In a particular and non-limiting example, the dedicated CPU resource(s) may be used, e.g. by the apparatus, e.g. by the UE, to measure CSI-RS resources, in particular associated with data collection and / or to log L1-RSRP and / or beam ID(s). As such, there may be no or negligible impact on the number of CPU resource(s) for measuring CSI-RS, e.g. no or negligible impact on legacy CSI processing resources / capacity. The number of dedicated CPU resources for data collection may correspond to or comprises a maximum available number, e.g. a limit, of dedicated CPU resources for data collection. The maximum available number or limit of dedicated CPU resources may be a separate limit, e.g. reported by the apparatus, e.g. the UE, in particular using the at least one indication. The at least one indication may indicate that the apparatus supports simultaneous measurements for multiple data collection configurations. The at least one indication may indicate a limit, e.g. a maximum number, on the number of dedicated CPU resources available for simultaneous or concurrent logging of collected data. In a particular and non-limiting example a parameter, which may be referred to as "O_DCPU", may be considered. "O_DCPU" may define or indicate a number of hardware units (which may be dedicated CPU resources or also referred to as data collection processing units (DCPU)), e.g. for at least one data collection configuration and / or measurement configuration. The value of "O_DCPU" may be defined, for example based on and / or in association with / to the measurement(s) that the apparatus, e.g. the UE, may perform or conduct, e.g. corresponding to the data collection configuration.

[0233] It should be noted that the apparatus, e.g. the UE, may indicate the separate, dedicated and / or independent capability for the dedicated CPU resources available for data collection. So, the network may know whether the apparatus, e.g. the UE, can be configured with dedicated CSI-RS resources or resource sets, in particular for Data Collection (DC) purposes. The network may transmit, e.g. pilot CSI-RS signal(s) based on and / or associated with dedicated DC. If the network is not or may not be aware of this dedicated capability, then it may, e.g. by default, consider a shared resource allocation, inparticular for DC, as well as legacy CSI report. Therefore, this would limit total allocation for legacy CSI reporting.

[0234] A "data collection configuration" may be regarded as setting(s), instruction(s) and / or a set of parameters that define how, when, and / or what types of data the apparatus, e.g. the UE, may collect and / or what, when and / or how to log and / or store this data. The data to be collected may be Artificial Intelligence (Al) -related and / or Al-based data, Machine Learning (ML)-related and / or ML-based data and / or beam management related and / or beam management based data and / or beam management related measurement(s). In a non-limiting example, a data collection configuration may determine, e.g., which signals or resources (e.g., CSI-RS-based, SSB-based; which may also be referred to a as measurement resource set) the apparatus, e.g. the UE, should measure or monitor; what measurement quantity- / ies (e.g., L1-RSRP, L1-RSRQ, beam ID; which may also be referred to a as measurement target) should be logged; how frequently the measurements should be collected; how long the apparatus, e.g. the UE, should store the collected data, e.g. before reporting (which may also be referred to a as logging duration and / or buffering); when the apparatus, e.g. the UE, should monitor and / or log (which may also be referred to a as a measurement and / or logging trigger; e.g. periodic logging and / or event-based logging); how the collected data may be transmitted and / or reported to the base station, e.g. the gNB, (e.g., RRC signaling, UElnformationResponse etc.; which may also be referred to a as reporting mechanism). The data collection configuration may in particular include and / or indicate at least one type of logging that is to be conducted, e.g. for collected data and / or for data to be collected. The at least one data collection configuration may include and / or indicate a periodic logging of collected data and / or for data to be collected. The at least one data collection configuration may include and / or indicate an event-based logging of collected data and / or for data to be collected. The at least one data collection configuration may include and / or indicate at least one data collection quantity. The at least one data collection quantity may include at least one of: at least one beam prediction in a spatial domain, at least one beam prediction in a time domain, at least one CSI prediction, or at least one CSI compression.

[0235] According to subject-matter described herein, the at least one data collection configuration may differ and / or may be different from a CSI measurement configuration with respect to, or in terms of how collected data and / or data to be collected is to be reported. The at leastone data collection configuration is provided, generated and / or created, e.g. by the network device, based at least in part on, in dependence of, with regard to, tailored to, adapted / adjusted to, customized to, matched to, coordinated with and / or taking into account the indicated number of dedicated CPU resources for data collection which are available in addition to a number of CPU resources for measuring CSI-RS, e.g. which may be transmitted via the at least one indication by the apparatus, e.g. the UE.

[0236] According to subject-matter described herein, the number of dedicated CPU resources for data collection may be included in, comprised in, incorporated in and / or be part of a, in particular same and / or shared, CPU pool. The CPU pool may be shared and / or commonly, jointly and / or combinedly used for, in particular both, data collection and for measuring CSI-RS. The CPU pool may be referred to as a shared CPU pool. The CPU pool may include a number of CPU resources and / or it may include a number of, in particular different, subgroups, subclusters and / or subsets, for example a first subgroup and a second subgroup etc. For example, the number dedicated CPU resources for data collection may correspond to, define and / or include at least one subgroup or subset of the number of CPU resources for measuring CSI-RS. In other words, the number dedicated CPU resources for data collection may correspond to, define and / or include at least one subgroup or subset of the, in particular shared, CPU pool. The CPU pool may be for measuring CSI-RS for at least one of: an Artificial Intelligence (Al) model, or a Machine Learning (ML) model. In such a scenario, for example, CSI-RS resource(s) which are to be measured for data collection purposes, may be used for other CSI reporting. This may be referred to as example Case 1. Thus, it may be clarified in the specification how an occupancy of the CPU usage is affected. For example, how many CPU resource(s) are occupied and / or a duration of the CPU resource(s) usage, e.g. for data collection. In one example, a shared hardware or a shared pool (e.g. N_CPU) may be used by the apparatus, e.g. the UE, to measure CSI-RS resource(s), in particular associated with data collection (e.g. to log L1-RSRP and / or beam I D(s)) and / or to calculate CSI for other types of CSI reports (e.g., CSI acquisition, beam reporting etc.). This may be referred to as example Case 2-1.

[0237] As mentioned, the at least one indication may indicate a number of, a quantity of and / or how many dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection may be selected, defined, assigned, allocated, determined, destined,specified and / or identified from the number of CPU resources for measuring CSI-Reference Signals (RS). In other words, the at least one indication may indicate a number of, a quantity of and / or how many dedicated CPU resources are available, provided, accessible, supplied and / or deployed, e.g. within the apparatus, e.g. the UE, e.g. using the same / shared CPU pool.

[0238] The at least one indication may indicate a different limit for the number of dedicated CPU resources for data collection and for the number of CPU resources which are available for measuring CSI-RS. For example, the at least one indication may indicate a first limit for the number dedicated CPU resources for data collection and a second limit for the number of CPU resources which are available for measuring CSI-RS, wherein the first limit and the second limit are different. In one example, as in example Case 1, the apparatus, e.g. the UE, may also consider separate hardware limits for CSI-ReportConfigs that enable data collection. This may be referred to as example Case 2-2. The at least one indication may indicate a maximum available number of dedicated CPU resources for data collection from the number of CPU resources for measuring CSI-RS. In another variant, as in Case 1 , the UE may also consider separate hardware limits for CSI-ReportConfigs that enable data collection. The at least one indication may indicate the number of dedicated CPU resources for data collection to be at least initially, at least at first and / or at least originally to be zero (0). In one example, the apparatus, e.g. the UE, may be defined to consider Zero CPU units (O_CPU=0) for CSI-ReportConfigs that enable data collection. This may be referred to as example Case 2-3.

[0239] According to subject-matter described herein, the at least one indication may indicate that the apparatus supports simultaneous measurements for multiple data collection configurations. The at least one indication may indicate a limit on the number of dedicated CPU resources available for simultaneous or concurrent logging of collected data. The at least one data collection configuration may correspond to, be substantially identical to, be substantially the same as, or may include a CSI measurement configuration. In other words, and in a non-limiting example, the apparatus, e.g. the UE, may be configured within the CSI-Report Config. The at least one data collection configuration may be provided, generated and / or created, e.g. by the network device, based at least in part on, in dependence of, with regard to, tailored to, adapted / adjusted to, customized to, matched to, coordinated with and / or taking into account the indicated number of dedicated CPUresources for data collection which are selected from a number of CPU resources which are available for measuring CSI-RS, e.g. which may be transmitted via the at least one indication by the apparatus, e.g. the UE.

[0240] The subject-matter descried herein may affect a technical standard specification, e.g. a 3rdGeneration Partnership Project (3GPP) standard specification in the following way, which is outlined as merely an example, e.g. all features, elements and / or steps which are described should be considered optional and, in particular non-limiting. The impact on the specification may be exemplarily outlined / reflected as follows:

[0241] I. Example of having dedicated CPU resources for data collection which are available in addition to a number of CPU resources for measuring CSI-RS

[0242] 5.X.X.X Data Logging Criteria

[0243] The UE indicates the number of supported simultaneous data logging NLogwith parameter simultaneousDataLoggingPerCC in a component carrier, simultaneousDataLoggingPerAIICC across all component carriers. If a UE supports NLogsimultaneous data logging it is said to have NLogdata collection processing units (DCPU) for performing data collection. If L DCPU are occupied for logging of data collection configurations in a given OFDM symbol, the UE has NLog~ unoccupied DCPUs. If N data collection configurations start occupying their respective DCPUs on the same OFDM symbol on which NLog- L DCPUs are unoccupied, where each data collection configuration n = 0, ...,N - 1 corresponds to O^pu, the UE is not reguired to log the IV - M reguested data collection configurations with lowest priority (defined separately), where 0 < M < N is the largest value such that n=o °DCPU -NLog ~ L holds.

[0244] Logging for a data collection configuration occupies a number of DCPUs for a number of symbols as follows:

[0245] ODCPU = 1 for a data collection configuration with higher layer parameter DataCollectionQuantity set to 'L1-RSRP'.0DCPu = for a data collection configuration with higher layer parameter DataCollectionQuantity set to ‘cri-RI-PMI-CQI1.

[0246] 0DCPu = Y for a data collection configuration with higher layer parameter DataCollectionQuantity set to ‘cri-RI-PMI-CQI1and with codebookType set to 'typel l-Doppler-r18'.

[0247] For a data collection configuration, the DCPU(s) are occupied for a number of OFDM symbols as follows:

[0248] occupies DCPU(s) from the first symbol of the earliestone of each transmission occasion of periodic or semi-persistent CSI-RS / SSB resource for data collection measurement, wherein the earliest transmission occasion is the first transmission occasion after activation of data collection, until the last symbol of the latest one of the CSI-RS / SSB resource for data collection measurement, wherein the latest transmission occasions if the last transmission occasion before deactivation of data collection.

[0249] II. Example of having dedicated CPU resources for data collection which are selected from a number of CPU resources which are available for measuring CSI-RS

[0250] 11.1 Example Case 1

[0251] The UE indicates the number of supported simultaneous CSI calculations NCPUwith parameter simultaneousCSI-ReportsPerCC or simultaneousCSI-SubReportsPerCC-r18 in a component carrier, and simultaneousCSI-ReportsAIICC or simultaneousCSI-SubReportsAIICC-r18 across all component carriers. If UE is configured with at least one CSI report setting with subconfiguration in a component carrier, UE shall use parameter simultaneousCSI-SubReportsPerCC-r18 in the component carrier; otherwise, UE shall use simultaneousCSI-ReportsPerCC in the component carrier. If UE is configured with at least one CSI reporting setting with sub-configuration in any component carrier,UE shall use simultaneousCSI-SubReportsAIICC-r18; otherwise, UE shall use simultaneousCSI-ReportsAIICC. If a UE supports NCPUsimultaneous CSI calculations it is said to have NCPUCSI processing units for processing CSI reports. If L CPUs are occupied for calculation of CSI reports in a given OFDM symbol, the UE has NCPU- L unoccupied CPUs. If N CSI reports start occupying their respective CPUs on the same OFDM symbol on which NCPU- L CPUs are unoccupied, where each CSI report n = 0, ...,N - 1 corresponds to 0^, the UE is not required to update the N - M requested CSI reports with lowest priority (according to Clause 5.2.5), where 0 < M < N is the largest value such that 2

[0252]

[0253] .^=0 ^CPU — ^cpu ~ holds.

[0254] A UE is not expected to be configured with an aperiodic CSI trigger state containing more than NCPUReporting Settings. Processing of a CSI report occupies a number of CPUs for a number of symbols as follows:

[0255] OCPU= 0 for a CSI report with CSI-ReportConfig with higher layer parameter reportQuantity set to 'none' and CSI-RS-ResourceSet with higher layer parameter trs-lnfo configured

[0256] OCPU= 0 for a CSI report with CSI-ReportConfig with higher layer parameter reportQuantity set to ‘none’ or 'datalogging'.

[0257] OCPU= 1 for a CSI report with Itm-CSI-ReportConfig or a CSI report with CSI-ReportConfig with higher layer parameter reportQuantity set to 'cri-RSRP, 'ssb-Index-RSRP, 'cri-SINR, 'ssb-lndex-SINR, 'cri-RSRP-lndex', 'ssb-lndex-RSRP-Index', ' cri-SINR- Index', 'ssb-lndex-SINR-lndex' or 'none' (and CSI-RS-ResourceSet with higher layer parameter trs-lnfo not configured)

[0258] OCPU = 1 for a CSI report with a CSI report with CSI-ReportConfig with higher layer parameter reportQuantity set to 'none' or ‘datalogging’ and CS / -RS-ResourceSet with higher layer parameter dedicatedDataLogging configuredOCpu = (Y + 1) • X, for a CSI report with CSI-ReportConfig with higher layer parameter reportQuantity set to 'tdcp' and with number of delays Y configured by higher layer parameter Y, where the value of X e {1, 2} is reported by UE capability.

[0259] For a CSI report with CSI-ReportConfig with higher layer parameter reportQuantity set to ‘none’ or 'datalogging' and NZP-CSI-RS-ResourceSet with higher layer parameter dedicatedDataLogging is configured for data collection, the CPU(s) are occupied for a number of OFDM symbols as follows:

[0260] - A periodic and semi-persistent CSI report occupies CPU(s) from the first symbol of the earliest one of each transmission of periodic or semi-persistent CSI-RS / SSB resource for channel measurements and data collection, wherein the earliest transmission occasion is the first transmission occasion after the PDCCH trigger until the last symbol of the scheduled PUSCH / PUCCH carrying the report and after activation of data collection, until the last symbol of the latest one of CSI-RS / SSB resource for channel measurement and data collection measurement, wherein the latest transmission occasions is the last transmission occasion before deactivation of data collection.

[0261] 11.2 Example Case 2

[0262] The UE indicates the number of supported simultaneous CSI calculations NCPUwith parameter simultaneousCSI-ReportsPerCC in a component carrier, and simultaneousCSI-ReportsAIICC across all component carriers. The UE indicates the number of supported simultaneous CSI calculations for data logging NLog< Ncpuwith parameter simultaneousCSI-ReportsLogPerCC in a component carrier, and simultaneousCSI-ReportsLogAIICC across all component carriers. If a UE supports NCPUsimultaneous CSI calculations it is said to have CPU CSI processing units for processing CSI reports. If a UE supports NLogsimultaneous CSI calculation for data logging it is said to have NLogdata collection processing units (DCPU) for performing data collection. If L CPUs are occupied for calculation of CSI reports in a given OFDM symbol, the UE has NCPU- L unoccupied CPUs. If N CSI reports start occupying their respective CPUs on the same OFDM symbolon which NCPU- L CPUs are unoccupied, where each CSI report n = 0, ...,N - 1 corresponds to 0^, the UE is not required to update the N - M requested CSI reports with lowest priority (according to Clause 5.2.5), where 0 < M < N is the largest value such that Xn=o °CPU -NCPU ~ £ holds. If P DCPU are occupied for logging of CSI in a given OFDM symbol, the UE has NLog- P unoccupied DCPUs. If N CSI calculations for data logging start occupying their respective DCPUs on the same OFDM symbol on which NLog- P DCPUs are unoccupied, where each CSI calculation for data logging n = 0, ...,N - 1 corresponds to Opu, the UE is not required to log the N - M requested CSI report configurations with lowest priority (defined separately), where 0 < M < N is the largest value such that X

[0263]

[0264] n=o O£pu ^Log-P holds.

[0265] Logging for a CSI report configuration occupies a number of DCPUs for a number of symbols as follows:

[0266] ODCPU = ^>Ocpu = 0 for a CSI report with CSI-ReportConfig with higher layer parameter reportQuantity set to 'none' or ‘datalogging’ and CSI-RS-ResourceSet with a new parameter for data collection configuration set to 'L1-RSRP.

[0267] °

[0268]

[0269] DCPU = °cpu = 0 for a CSI report with CSI-ReportConfig with higher layer parameter reportQuantity set to 'none' or ‘datalogging’ and CSI-RS-ResourceSet with a new parameter for data collection configuration set to 'cri-RI-PMI-CQI'

[0270] ODCPU = Y> Ocpu= 0 for a CSI report with CSI-ReportConfig with higher layer parameter reportQuantity set to 'none' or ‘datalogging’ and CSI-RS-ResourceSet with a new parameter for data collection configuration set to 'cri-RI-PMI-CQI' with codebookType set to 'typell-Doppler-r18'.

[0271] For a CSI report configuration for logging, the DCPU(s) are occupied for a number of OFDM symbols as follows:

[0272] occupies DCPU(s) from the first symbol of the earliestone of each transmission occasion of periodic or semi-persistent CSI-RS / SSB resource for data collectionmeasurement, wherein the earliest transmission occasion is the first transmission occasion after activation of data collection, until the last symbol of the latest one of the CSI-RS / SSB resource for data collection measurement, wherein the latest transmission occasions if the last transmission occasion before deactivation of data collection.

[0273] The term "non-transitory," as used herein, is a limitation of the medium itself (i.e., tangible, not a signal) as opposed to a limitation on data storage persistency (e.g., RAM vs. ROM).

[0274] As used herein, "at least one of the following: " and "at least one of " and similar wording, where the list of two or more elements are joined by "and" or "or", mean at least any one of the elements, or at least any two or more of the elements, or at least all the elements.

[0275] The apparatus may include a processor configured to provide signals to and receive signals and to control the functioning of the apparatus. The processor may be configured to control other elements of apparatus by effecting control signaling via electrical leads connecting processor to the other elements, such as a display or a memory. The processor may, for example, be embodied in a variety of ways including circuitry, at least one processing core, one or more microprocessors with accompanying digital signal processor(s), one or more processor(s) without an accompanying digital signal processor, one or more coprocessors, one or more multi-core processors, one or more controllers, processing circuitry, one or more computers, various other processing elements including integrated circuits (for example, an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), and / or the like), or some combination thereof.

[0276] The term "circuitry" may refer to one or more or all of the following:

[0277] (a) hardware-only circuit implementations (such as implementations in only analog and / or digital circuitry) and

[0278] (b) combinations of hardware circuits and software, such as (as applicable): (i) a combination of analog and / or digital hardware circuit(s) with software / firmware and(ii) any portions of hardware processor(s) with software (including digital signal processor(s)), software, and memory(ies) that work together to cause an apparatus, such as a mobile phone or server, to perform various functions) and

[0279] hardware circuit(s) and / or processor(s), such as a microprocessor(s) or a portion of a microprocessor(s), that requires software (e.g., firmware) for operation, but the software may not be present when it is not needed for operation.

[0280] This definition of circuitry applies to all uses of this term in this application, including in any claims. As a further example, as used in this application, the term circuitry also covers an implementation of merely a hardware circuit or processor (or multiple processors) or portion of a hardware circuit or processor and its (or their) accompanying software and / or firmware. The term circuitry also covers, for example and if applicable to the particular claim element, a baseband integrated circuit or processor integrated circuit for a mobile device ora similar integrated circuit in server, a cellular network device, or other computing or network device.

[0281] The apparatus may comprise memory which may store information elements. The apparatus may include volatile memory and / or non-volatile memory. For example, volatile memory may include Random Access Memory (RAM) including dynamic and / or static RAM, on-chip or off-chip cache memory, and / or the like. Non-volatile memory, which may be embedded and / or removable, may include, for example, read-only memory, flash memory, magnetic storage devices, for example, hard disks, floppy disk drives, magnetic tape, optical disc drives and / or media, non-volatile random access memory (NVRAM), and / or the like. Like volatile memory, non-volatile memory may include a cache area for temporary storage of data. At least part of the volatile and / or non-volatile memory may be embedded in processor. The memories may store one or more software programs, instructions, pieces of information, data, and / or the like which may be used by the apparatus for performing operations disclosed herein.

[0282] The apparatus may be a User Equipment (UE), for example a mobile or handled device, for example a mobile phone or a smart phone, or may be at least comprised in a UE.

[0283] Fig. 1 shows example signaling according to subject-matter described herein. In particular, Fig. 1 shows an example of dedicated CPU resource which are available in addition to anumber of CPU resources for measuring CSI- Reference Signals (RS), in particular wherein the number of dedicated CPU resources for data collection are independent of the number of CPU resources for measuring CSI-RS. Fig. 1 shows an example of periodic data collection. It should be noted that any reference related to a specific CSI resource, reporting, measurement and / or message or the like may be considered optional and, in particular, not limiting. Fig. 1 may be described by the following non-limiting steps:

[0284] Step 1 : (optional) A data collection trigger may be received at a network device or base station, e.g. the gNB.

[0285] Step 2: The apparatus, e.g. the UE, may be notified about the trigger. In an example, a UE attach procedure may be completed, e.g. along with a capability exchange. The apparatus, e.g. the UE, may transmit at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are available in addition to a number of CPU resources for measuring CSI- Reference Signals (RS). In an example, the number of dedicated CPU resources for data collection may be independent of the number of CPU resources for measuring CSI-RS and max dedicated CPU resources for data collection.

[0286] Step 3: The network device, e.g. the gNB, may configure the apparatus, e.g. the UE, with necessary data collection configuration(s). In other words, the apparatus, e.g. the UE, may receive at least one data collection configuration for the number of dedicated CPU resources for data collection. In this case, the apparatus, e.g. the UE may be configured with, e.g. periodic, logging of the data measured, in particular based at least in part on data collection specific CSI-RS configurations for, e.g. Set A / B. It may be noted that, herein, a ResourceSet may belong to input to AI / ML model (which may be referred to as "Set B") and / or ResourceSet may belong to output to AI / ML model (which may be referred to as "Set A").

[0287] Step 4: The apparatus, e.g. the UE, may acknowledge (ACK) the DC configuration(s), in particular using an RRC configuration complete message. Otherwise, theapparatus, e.g. the UE, may be unable to comply with the configuration. In such a case, it may send either RRC reconfiguration failure or a new failure indication message, e.g. dedicated for data collection.

[0288] Step 5: The apparatus, e.g. the UE, may initialize a count for the number of dedicated CPU resources for data collection that is occupied to be zero. In particular it may initialize the Occupied CPU count for data collection purposes to zero (0), e.g. (O_DCPU=0). The initializing may be based at least in part on or be conducted based at least in part on acknowledging the at least one data collection configuration

[0289] Step 6: The apparatus, e.g. the UE may start counting of Occupied CPU count for data collection purposes. In other words, it may start to count, or it may count a number of, a quantity of and / or how many of the number of dedicated CPU resources for data collection are occupied, e.g. in association with conducting the data collection.

[0290] Step 7: The apparatus, e.g. the UE may log or may start logging the measured, e.g.

[0291] L1 , quantity in its AS buffer.

[0292] Step 8: The apparatus, e.g. the UE may send a data availability indication to the network device.

[0293] Step 9: The network device, e.g. the gNB, may request data using UElnformationRequest message.

[0294] Step 10: The apparatus, e.g. the UE may send logged data using UElnformationResponse.

[0295] Step 11 : The network device, e.g. the gNB, may map the collected data to relevant trace collection sessions.

[0296] Step 12: The collected data may be send to a trace collection entity.Figs. 2A-2B show example signaling according to subject-matter described herein. In particular, Figs. 2A-2B show an example of dedicated CPU resources for data collection which are available in addition to a number of CPU resources for measuring CSI-Reference Signals (RS), in particular wherein the number of dedicated CPU resources for data collection are independent of the number of CPU resources for measuring CSI-RS. Figs. 2A-2B show an example of event-based data collection, wherein Fig. 2B may be regarded as a continuation of Fig. 2A, which is why these figures may be commonly described. It should be noted that any reference related to a specific CSI resource, reporting, measurement and / or message or the like may be considered optional and, in particular, not limiting. Figs. 2A-2B may be described by the following non-limiting steps:

[0297] Step 1 : (optional) A data collection trigger may be received at a network device or base station, e.g. the gNB.

[0298] Step 2: The apparatus, e.g. the UE, may be notified about the trigger. In an example, a UE attach procedure may be completed, e.g. along with a capability exchange. The apparatus, e.g. the UE, may transmit at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are available in addition to a number of CPU resources for measuring CSI- Reference Signals (RS). In an example, the number of dedicated CPU resources for data collection may be independent of the number of CPU resources for measuring CSI-RS and max dedicated CPU resources for data collection.

[0299] Step 3: The network device, e.g. the gNB, may configure the apparatus, e.g. the UE, with necessary data collection configuration(s). In other words, the apparatus, e.g. the UE, may receive at least one data collection configuration for the number of dedicated CPU resources for data collection. In this case, the apparatus, e.g. the UE may be configured with, e.g. more than one e.g.

[0300] L1RSRP thresholds received from data collection specific CSI-RS configurations, in particular for Set A / B. It may be noted that, herein, a ResourceSet may belong to input to AI / ML model (which may be referred to as "Set B") and / or ResourceSet may belong to output to AI / ML model (whichmay be referred to as "Set A").

[0301] Step 4: The apparatus, e.g. the UE, may acknowledge (ACK) the threshold-based configuration(s), in particular using an RRC configuration complete message. Otherwise, the apparatus, e.g. the UE, may be unable to comply with the configuration. In such a case, it may send either RRC reconfiguration failure or a new failure indication message, e.g. dedicated for data collection.

[0302] Step 5: The apparatus, e.g. the UE, may initialize a count for the number of dedicated CPU resources for data collection that is occupied to be zero. In particular it may initialize the Occupied CPU count for data collection purposes to zero (0), e.g. (O_DCPU=0). The initializing may be based at least in part on or be conducted based at least in part on acknowledging the at least one data collection configuration

[0303] Step 6: The apparatus, e.g. the UE, may start measuring CSI-RS and / or may generate the specified quantity, e.g. based at least in part on the report quantity indicated in data collection configuration such as cri-RSRP values etc. When the value of measured quantity is above the configured threshold (e.g. event A1), it may start preparing to send a report to the network device.

[0304] Step 7: The apparatus, e.g. the UE, may send a Layer 3 (L3) measurements report to the network device.

[0305] Step 8: The network device may send at least one indication to the apparatus, e.g.

[0306] the UE, to log or start logging the L1 measurement.

[0307] Step 9: The apparatus, e.g. the UE may count or start counting the Occupied CPU count, e.g. for data collection purposes.

[0308] Step 10: The apparatus, e.g. the UE, may log or start logging the measured L1

[0309] quantity in its AS buffer.

[0310] Step 11 : When the value of the measured quantity for the event is below theconfigured threshold (e.g. an exit condition for event A1), the apparatus, e.g. the UE, may start preparing to send a report to the network device.

[0311] Step 12: The apparatus, e.g. the UE, may send the L3 measurement report to the network device, e.g. indicating exit condition for A1.

[0312] Step 13: The network device may send at least one indication to stop the logging of the data by the apparatus, e.g. the UE.

[0313] Step 14: The apparatus, e.g. the UE, may stop logging of the measured L1 quantity in its AS buffer.

[0314] Step 15: The apparatus, e.g. the UE, may stop counting of the Occupied CPU count for data collection purposes.

[0315] Step 16: The apparatus, e.g. the UE, may send data availability indication to the network device.

[0316] Step 17: The network device, e.g. the gNB, may request data using UElnformationRequest message.

[0317] Step 18: The apparatus, e.g. the UE, may send logged data using UElnformationResponse.

[0318] Step 19: The network device, e.g. the gNB, may map the collected data to relevant trace collection sessions.

[0319] Step 20: The collected data may be send to a trace collection entity.

[0320] Figs. 3A-3B show example signaling according to subject-matter described herein. In particular, Figs. 3A-3B show an example of dedicated CPU resources for data collection which selected from a number of CPU resources which are available for measuring CSI-Reference Signals (RS), in particular wherein the number of CPU resources for measuring CSI-RS may be included in or part of a same CPU pool which is shared for data collectionand for measuring CSI-RS. Figs. 3A-3B show an example of event-based data collection, wherein Fig. 3B may be regarded as a continuation of Fig. 3A, which is why these figures may be commonly described. It should be noted that any reference related to a specific CSI resource, reporting, measurement and / or message or the like may be considered optional and, in particular, not limiting. Figs. 3A-3B may be described by the following nonlimiting steps:

[0321] Step 1 : (optional) A data collection trigger may be received at a network device or base station, e.g. the gNB.

[0322] Step 2: The apparatus, e.g. the UE, may be notified about the trigger. In an example, a UE attach procedure may be completed, e.g. along with a capability exchange. The apparatus, e.g. the UE, may transmit at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are selected from a number of CPU resources which are available for measuring CSI-Reference Signals (RS). In an example, the number of CPU resources for measuring CSI-RS are included in or part of a same CPU pool which is shared for data collection and for measuring CSI-RS. UE can optionally indicate the total number of separate hardware limit for data collection from the shared CPU pool. The apparatus, e.g. the UE, can optionally indicate the total number of separate hardware limit for data collection from the shared CPU pool.

[0323] Step 3: The network device, e.g. the gNB, may configure the apparatus, e.g. the UE, with necessary data collection configuration(s). In other words, the apparatus, e.g. the UE, may receive at least one data collection configuration for the number of dedicated CPU resources for data collection. In this case, the apparatus, e.g. the UE may be configured with, e.g. Layer 3 (L3) event-based logging (e.g., A1 / A2) of the data measured based at least in part on data collection specific CSI-RS configurations, in particular for Set A / B. It may be noted that, herein, a ResourceSet may belong to input to AI / ML model (which may be referred to as "Set B") and / or ResourceSet may belong to output to AI / ML model (which may be referred to as "Set A").Step 4: The apparatus, e.g. the UE, may acknowledge (ACK) the data collection and logging configuration(s), in particular using an RRC configuration complete message. Otherwise, the apparatus, e.g. the UE, may be unable to comply with the configuration. In such a case, it may send either RRC reconfiguration failure ora new failure indication message, e.g. dedicated for data collection.

[0324] Step 5: The apparatus, e.g. the UE, may initialize a count for the number of dedicated CPU resources for data collection that is occupied to be zero. In particular it may initialize the Occupied CPU count for data collection purposes to zero (0), e.g. (O_DCPU=0). The initializing may be based at least in part on or be conducted based at least in part on acknowledging the at least one data collection configuration

[0325] Step 6: The apparatus, e.g. the UE, may detect an A1 event.

[0326] Step 7: The apparatus, e.g. the UE, may send a Layer 3 (L3) measurements report associated with the event to the network device.

[0327] Step 8: The network device may send at least one indication to the apparatus, e.g.

[0328] the UE, to log or start data logging.

[0329] Step 9: The apparatus, e.g. the UE may count or start counting the Occupied CPU count, e.g. for data collection purposes. If the apparatus, e.g. the UE, provides a separate hardware limit, then maximum number of simultaneous CSI calculation for logging can be up to NJog, which may be less than or equal to the maximum number of simultaneous CSI calculation for legacy CSI reporting that the apparatus, e.g. the UE, may indicate in the capability exchange. Otherwise, the occupancy of CPU for CSI calculation for logging may be less than or equal to the total number of simultaneous CSI calculation used for legacy CSI reporting.

[0330] Step 10: The apparatus, e.g. the UE, may log or start logging the measured L1

[0331] quantity in its AS buffer.Step 11 : When the value of the measured quantity for the event is below the configured threshold (e.g. an exit condition for event A1), the apparatus, e.g. the UE, may start preparing to send a report to the network device.

[0332] Step 12: The apparatus, e.g. the UE, may send the L3 measurement report to the network device, e.g. indicating exit condition for A1.

[0333] Step 13: The network device may send at least one indication to stop the logging of the data by the apparatus, e.g. the UE.

[0334] Step 14: The apparatus, e.g. the UE, may stop logging of the measured L1 quantity in its AS buffer. The apparatus, e.g. the UE, may stop counting of the Occupied CPU count for data collection purposes.

[0335] Step 15: The apparatus, e.g. the UE, may send data availability indication to the network device.

[0336] Step 16: The network device, e.g. the gNB, may request data using UElnformationRequest message.

[0337] Step 17: The apparatus, e.g. the UE, may send logged data using UElnformationResponse.

[0338] Step 18: The network device, e.g. the gNB, may map the collected data to relevant trace collection sessions.

[0339] Step 19: The collected data may be sent to a trace collection entity.

[0340] Fig. 4 shows a flow chart 400 of an example aspect of subject-matter described herein. The flow chart may comprise steps of an apparatus, a method and / or a non-transitory computer readable medium as mentioned herein.The flow chart 400 may include the step of transmitting 410 at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are available in addition to a number of CPU resources for measuring CSI- Reference Signals (RS), wherein the number of dedicated CPU resources for data collection are independent of the number of CPU resources for measuring CSI-RS. The flow chart 400 may include the step of receiving 420 at least one data collection configuration for the number of dedicated CPU resources for data collection. The flow chart 400 may include the step of conducting data collection 430 based at least in part on the data collection configuration by using at least one of the number of dedicated CPU resources.

[0341] The flow chart 400 may comprise or be combined with one or more additional steps associated with some of the example aspects and / or associate example embodiments described herein.

[0342] Fig. 5 shows a flow chart 500 of an example aspect of subject-matter described herein. The flow chart may comprise steps of an apparatus, a method and / or a non-transitory computer readable medium as mentioned herein.

[0343] The flow chart 500 may include the step of transmitting 510 at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are selected from a number of CPU resources which are available for measuring CSI-Reference Signals (RS), wherein the number of CPU resources for measuring CSI-RS are included in or part of a same CPU pool which is shared for data collection and for measuring CSI-RS. The flow chart 500 may include the step of receiving 520 at least one data collection configuration for the number of dedicated CPU resources for data collection. The flow chart 500 may include the step of conducting data collection 530 based at least in part on the data collection configuration by using at least one of the number of dedicated CPU resources.

[0344] The flow chart 500 may comprise or be combined with one or more additional steps associated with some of the example aspects and / or associate example embodiments described herein.While the foregoing is directed to embodiments of the disclosure, other and further embodiments of the disclosure may be devised without departing from the basic scope thereof, and the scope thereof is determined by the claims that follow.

[0345] LIST OF ABBREVIATIONS

[0346] Al Artificial Intelligence

[0347] BM Beam management

[0348] CSI Channel State Information

[0349] CPU CSI Processing Unit

[0350] L1 Layer 1

[0351] L2 Layer 2

[0352] L3 Layer 3

[0353] NG Next Generation

[0354] MAC Medium Access Control

[0355] ML Machine Learning

[0356] NW Network I Network device

[0357] RAN Radio Access Network

[0358] UE User equipment

[0359] TCI Transmission Configuration Index

[0360] RRC Radio Resource Control

[0361] RSRP Reference Signal Received Power

[0362] RS Reference Signal

[0363] SSB Synchronization Signal Block

[0364] SS Synchronization Signal

[0365] SI NR Signal to Interference Noise Ratio

[0366] PCI Physical Cell Identity

[0367] PBCH Physical Broadcast Channel

[0368] PDCCH Physical Downlink Control Channel

[0369] PDSCH Physical Data Shared Channel

[0370] LIST OF REFERENCE NUMERALS

[0371] 410, 510 Transmitting Step420, 520 Receiving Step

[0372] 430, 530 Conducting data collection

Claims

WE CLAIM:

1. Apparatus, comprising:at least one processor; andat least one memory storing instructions that, when executed by the at least one processor, cause the apparatus at least to perform:transmitting at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are selected from a number of CPU resources which are available for measuring CSI-Reference Signals (RS), wherein the number of CPU resources for measuring CSI-RS are included in or part of a same CPU pool which is shared for data collection and for measuring CSI-RS,receiving at least one data collection configuration for the number of dedicated CPU resources for data collection, andconducting data collection based at least in part on the data collection configuration by using at least one of the number of dedicated CPU resources.

2. Apparatus according to claim 1 , wherein the CPU pool is further for measuring CSI-RS for at least one of: an Artificial Intelligence (Al) model, or a Machine Learning (ML) model.

3. Apparatus according to any one of the preceding claims, wherein the number of dedicated CPU resources for data collection corresponds to at least a subgroup of the number of CPU resources for measuring CSI-RS.

4. Apparatus according to any one of the preceding claims, wherein the at least one indication indicates a different limit for the number of dedicated CPU resources for data collection and for the number of CPU resources which are available for measuring CSI-RS.

575. Apparatus according to any one of the preceding claims, wherein the at least one indication indicates a maximum available number of dedicated CPU resources for data collection from the number of CPU resources for measuring CSI-RS.

6. Apparatus according to any one of the preceding claims, wherein the at least one indication indicates the number of dedicated CPU resources for data collection to be at least initially zero.

7. Apparatus according to any one of the preceding claims, wherein the at least one indication indicates that the apparatus supports simultaneous measurements for multiple data collection configurations.

8. Apparatus according to any one of the preceding claims, wherein the at least one indication indicates a limit on the number of dedicated CPU resources available for simultaneous or concurrent logging of collected data.

9. Apparatus according to any one of the preceding claims, wherein the at least one data collection configuration corresponds to or includes a CSI measurement configuration.

10. Apparatus according to any one of the preceding claims, wherein the at least one data collection configuration is provided in dependence of the indicated number of dedicated CPU resources for data collection.

11. Apparatus according to any one of the preceding claims, wherein the at least one data collection configuration indicates at least one type of logging that is to be conducted for collected data.

12. Apparatus according to any one of the preceding claims, wherein the at least one data collection configuration indicates at least one data collection quantity13. Apparatus according to claim 12, wherein the data collection quantity includes at least one of:- at least one beam prediction in a spatial domain,- at least one beam prediction in a time domain,- at least one CSI prediction, or- at least one CSI compression.

14. Apparatus according to any one of the preceding claims, wherein the at least one data collection configuration indicates a periodic logging of collected data.

15. Apparatus according to any one of the preceding claims 1-13, wherein the at least one data collection configuration indicates an event-based logging of collected data.

16. Apparatus according to any one of the preceding claims, wherein the apparatus is further caused to perform:initializing a count for the number of dedicated CPU resources for data collection that is occupied to be zero.

17. Apparatus according to claim 16, wherein the initializing is based at least in part on acknowledging the at least one data collection configuration.

18. Apparatus according to any one of the preceding claims, wherein the apparatus is further caused to perform:starting to count how many of the number of dedicated CPU resources for data collection are occupied in association with conducting the data collection.

19. Apparatus according to any one of the preceding claims, wherein the at least one indication is provided when an attach procedure between the apparatus and the network device is completed.

20. Apparatus according to any one of the preceding claims, wherein the at least one indication is provided as part of a capability exchange procedure between the apparatus and the network device.

21. Apparatus according to any one of the preceding claims, wherein the data collection is conducted to collect at least one of: Artificial Intelligence (Al) related data, or Machine Learning (ML) related data.

22. Apparatus according to any one of the preceding claims, wherein the data collection is conducted to collect beam management related data.

23. Apparatus according to any one of the preceding claims, wherein the apparatus is a User equipment (UE) or wherein the apparatus is comprised in a UE.

24. Apparatus comprising means for:transmitting (510) at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are selected from a number of CPU resources which are available for measuring CSI-Reference Signals (RS), wherein the number of CPU resources for measuring CSI-RS are included in or part of a same CPU pool which is shared for data collection and for measuring CSI-RS,receiving (520) at least one data collection configuration for the number of dedicated CPU resources for data collection, andconducting data collection (530) based at least in part on the data collection configuration by using at least one of the number of dedicated CPU resources.

25. Apparatus according to claim 24, wherein the CPU pool is further for measuring CSI-RS for at least one of: an Artificial Intelligence (Al) model, or a Machine Learning (ML) model.

26. Apparatus according to any one of the preceding claims 24-25, wherein the number of dedicated CPU resources for data collection corresponds to at least a subgroup of the number of CPU resources for measuring CSI-RS.

27. Apparatus according to any one of the preceding claims 24-26, wherein the at least one indication indicates a different limit for the number of dedicated CPU resources for data collection and for the number of CPU resources which are available for measuring CSI-RS.

28. Apparatus according to any one of the preceding claims 24-27, wherein the at least one indication indicates a maximum available number of dedicated CPU resources for data collection from the number of CPU resources for measuring CSI-RS.

29. Apparatus according to any one of the preceding claims 24-28, wherein the at least one indication indicates the number of dedicated CPU resources for data collection to be at least initially zero.

30. Apparatus according to any one of the preceding claims 24-29, wherein the at least one indication indicates that the apparatus supports simultaneous measurements for multiple data collection configurations.

31. Apparatus according to any one of the preceding claims 24-30, wherein the at least one indication indicates a limit on the number of dedicated CPU resources available for simultaneous or concurrent logging of collected data.

32. Apparatus according to any one of the preceding claims 24-31 , wherein the at least one data collection configuration corresponds to or includes a CSI measurement configuration.

33. Apparatus according to any one of the preceding claims 24-32, wherein the at least one data collection configuration is provided in dependence of the indicated number of dedicated CPU resources for data collection.

34. Apparatus according to any one of the preceding claims 24-33, wherein the at least one data collection configuration indicates at least one type of logging that is to be conducted for collected data.

35. Apparatus according to any one of the preceding claims 24-34, wherein the at least one data collection configuration indicates at least one data collection quantity36. Apparatus according to claim 35, wherein the data collection quantity includes at least one of:- at least one beam prediction in a spatial domain,- at least one beam prediction in a time domain,- at least one CSI prediction, or- at least one CSI compression.

37. Apparatus according to any one of the preceding claims 24-36, wherein the at least one data collection configuration indicates a periodic logging of collected data.

38. Apparatus according to any one of the preceding claims 24-36, wherein the at least one data collection configuration indicates an event-based logging of collected data.

39. Apparatus according to any one of the preceding claims 24-38, wherein the means are configured for:initializing a count for the number of dedicated CPU resources for data collection that is occupied to be zero.

40. Apparatus according to claim 39, wherein the initializing is based at least in part on acknowledging the at least one data collection configuration.

41. Apparatus according to any one of the preceding claims 24-40, wherein the means are configured for:starting to count how many of the number of dedicated CPU resources for data collection are occupied in association with conducting the data collection.

42. Apparatus according to any one of the preceding claims 24-41 , wherein the at least one indication is provided when an attach procedure between the apparatus and the network device is completed.

43. Apparatus according to any one of the preceding claims 24-42, wherein the at least one indication is provided as part of a capability exchange procedure between the apparatus and the network device.

44. Apparatus according to any one of the preceding claims 24-43, wherein the data collection is conducted to collect at least one of: Artificial Intelligence (Al) related data, or Machine Learning (ML) related data.

45. Apparatus according to any one of the preceding claims 24-44, wherein the data collection is conducted to collect beam management related data.

46. Apparatus according to any one of the preceding claims 24-45, wherein the apparatus is a User equipment (UE) or wherein the apparatus is comprised in a UE.

47. Method, comprising:transmitting at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are selected from a number of CPU resources which are available for measuring CSI-Reference Signals (RS), wherein the number of CPU resources for measuring CSI-RS are included in or part of a same CPU pool which is shared for data collection and for measuring CSI-RS,receiving at least one data collection configuration for the number of dedicated CPU resources for data collection, andconducting data collection based at least in part on the data collection configuration by using at least one of the number of dedicated CPU resources.

48. A non-transitory computer readable storage medium comprising program instructions that, when executed by an apparatus, causes operations comprising:transmitting at least one indication to a network device that indicates a number of dedicated Channel State Information (CSI) Processing Unit (CPU) resources for data collection which are selected from a number of CPU resources which are available for measuring CSI-Reference Signals (RS), wherein the number of CPU resources for measuring CSI-RS are included in or part of a same CPU pool which is shared for data collection and for measuring CSI-RS,receiving at least one data collection configuration for the number of dedicated CPU resources for data collection, andconducting data collection based at least in part on the data collection configuration by using at least one of the number of dedicated CPU resources.

Images