Device, method and apparatus for communication, and computer readable medium

Abstract

The embodiment of the invention discloses equipment, a method and a device for communication. In an embodiment, a network device receives, from a terminal device, at least one of a Reference Signal Received Power (RSRP) and a Reference Signal Received Quality (RSRQ) associated with a serving cell of the terminal device. The network device then determines a signal quality level associated with a neighbor cell of the serving cell using a machine learning (ML) or artificial intelligence (AI) model based on at least one of the RSRP and the RSRQ associated with the serving cell. In this way, the throughput of the communication system may be improved.

Description

Technical Field

[0001] Embodiments of the present disclosure generally relate to the field of telecommunications, and more particularly, to devices, methods, apparatuses, and computer-readable storage media for communication. Background Art

[0002] With the development of communication technologies, terminal devices are enabled to operate on one or more frequency carriers or bands. Further, in order to enhance the coverage of a cell, one or more network devices may provide multiple cells, and each of the multiple cells corresponds to a respective frequency carrier and / or a respective radio access technology (RAT). In an example, a wider bandwidth may be configured for a cell with a higher frequency carrier in order to increase the traffic throughput of terminal devices resident in that cell. On the other hand, another cell with a lower frequency carrier may have a larger coverage area than the cell with the higher frequency carrier in order to provide seamless coverage for terminal devices.

[0003] In some cases, a terminal device needs to measure neighboring cells during communication with a serving cell, and the terminal device also sends a measurement report of the neighboring cells to the network device. Then, the network device may schedule the terminal device accordingly, such as handover or other cell-level operations. Therefore, improving the estimation of the cell quality associated with the terminal device is a key aspect related to communication performance. Summary of the Invention

[0004] Generally speaking, example embodiments of the present disclosure provide devices, methods, apparatuses, and computer-readable storage media for estimating neighboring cells.

[0005] In a first aspect, a network device is provided. The network device may include at least one processor; and at least one memory storing instructions that, when executed by the at least one processor, cause the network device to: receive from a terminal device at least one of a reference signal received power (RSRP) and a reference signal received quality (RSRQ) associated with a serving cell of the terminal device. The network device also determines a signal quality level associated with a neighboring cell of the serving cell using a machine learning (ML) or artificial intelligence (AI) model based on at least one of the RSRP and RSRQ associated with the serving cell.

[0006] In a second aspect, a terminal device is provided. The terminal device may include at least one processor; and at least one memory storing instructions that, when executed by the at least one processor, cause the terminal device to: send at least one of RSRP and RSRQ associated with a serving cell of the terminal device to a network device, and at least one of RSRP and RSRQ associated with the serving cell will be used to determine a signal quality level associated with an adjacent cell of the serving cell based on an ML or AI model.

[0007] In a third aspect, a method implemented at a network device is provided. The method includes: receiving at least one of RSRP and RSRQ associated with a serving cell of the terminal device from the terminal device; and using an ML or AI model to determine a signal quality level associated with an adjacent cell of the serving cell based on at least one of RSRP and RSRQ associated with the serving cell.

[0008] In a fourth aspect, a method implemented at a terminal device is provided. The method includes: sending at least one of RSRP and RSRQ of a serving cell of the terminal device to a network device, and at least one of RSRP and RSRQ of the serving cell will be used to determine a signal quality level of an adjacent cell of the serving cell based on an ML or AI model.

[0009] In a fifth aspect, an apparatus for a network device is provided. The apparatus includes: means for receiving at least one of RSRP and RSRQ associated with a serving cell of the terminal device from the terminal device; and means for using an ML or AI model to determine a signal quality level associated with an adjacent cell of the serving cell based on at least one of RSRP and RSRQ associated with the serving cell.

[0010] In a sixth aspect, an apparatus for a terminal device is provided. The apparatus includes: means for sending at least one of RSRP and RSRQ of a serving cell of the terminal device to a network device, and at least one of RSRP and RSRQ of the serving cell will be used to determine a signal quality level of an adjacent cell of the serving cell based on an ML or AI model.

[0011] In a seventh aspect, a non-transitory computer-readable medium is provided, which has component program instructions for causing an apparatus to at least execute the method according to any one of the third to fourth aspects.

[0012] In an eighth aspect, there is provided a computer program comprising instructions which, when executed by a device, cause the device to at least: receive from a terminal device at least one of RSRP and RSRQ associated with a serving cell of the terminal device; and determine a signal quality level associated with an adjacent cell of the serving cell using an ML or AI model based on at least one of RSRP and RSRQ associated with the serving cell.

[0013] In a ninth aspect, there is provided a computer program comprising instructions which, when executed by a device, cause the device to at least: send to a network device at least one of RSRP and RSRQ associated with a serving cell of a terminal device, and at least one of RSRP and RSRQ associated with the serving cell will be used to determine a signal quality level associated with an adjacent cell of the serving cell based on an ML or AI model.

[0014] In a tenth aspect, there is provided a network device. The network device includes a receiving circuit configured to: receive from a terminal device at least one of RSRP and RSRQ associated with a serving cell of the terminal device. The network device further includes a determining circuit configured to: determine a signal quality level associated with an adjacent cell of the serving cell using an ML or AI model based on at least one of RSRP and RSRQ associated with the serving cell.

[0015] In an eleventh aspect, there is provided a terminal device. The terminal device includes a sending circuit configured to: send to a network device at least one of RSRP and RSRQ associated with a serving cell of the terminal device, and at least one of RSRP and RSRQ associated with the serving cell will be used to determine a signal quality level associated with an adjacent cell of the serving cell based on an ML or AI model.

[0016] It should be understood that the summary section is not intended to identify key or essential features of the embodiments of the present disclosure, nor is it intended to be used to limit the scope of the present disclosure. Other features of the present disclosure will become readily apparent through the following description. Brief Description of the Drawings

[0017] Some example embodiments will now be described with reference to the drawings, wherein:

[0018] Figure 1A An example network environment in which example embodiments of the present disclosure can be implemented is illustrated;

[0019] Figure 1B An example measurement gap for cross-frequency and cross-radio access technology (RAT) measurements is illustrated;

[0020] Figure 2Illustrates an example signaling procedure for estimating the quality of neighboring cells according to some embodiments of the present disclosure;

[0021] Figure 3A Illustrates an example path loss of a reference signal transmitted from a network device according to some embodiments of the present disclosure;

[0022] Figure 3B Illustrates an example of the direction of arrival (DOA) of a signal transmitted from a terminal device to a network device according to some embodiments of the present disclosure;

[0023] Figure 4A Illustrates example estimation results corresponding to different input parameters of a machine learning (ML) or artificial intelligence (AI) model according to some embodiments of the present disclosure;

[0024] Figure 4B Illustrates example estimation results corresponding to different output parameters of a machine learning (ML) or artificial intelligence (AI) model according to some embodiments of the present disclosure;

[0025] Figure 4C Illustrates an example comparison between an estimation result and an actual result according to some embodiments of the present disclosure;

[0026] Figure 5A Illustrates example estimation results corresponding to different machine learning (ML) or artificial intelligence (AI) models according to some embodiments of the present disclosure;

[0027] Figure 5B Illustrates example operation costs corresponding to different machine learning (ML) or artificial intelligence (AI) models according to some embodiments of the present disclosure;

[0028] Figure 5C Illustrates an example extreme random tree regression model according to some embodiments of the present disclosure;

[0029] Figure 5D Illustrates an example of reconstructing an extreme random tree regression model according to some embodiments of the present disclosure;

[0030] Figure 6A Illustrates an example comparison between the estimation result of an extreme random tree regression model and the estimation result of a reconstructed extreme random tree regression model according to some embodiments of the present disclosure;

[0031] Figures 6B - 6E Illustrates the gain of cell average throughput and resource block utilization according to some embodiments of the present disclosure;

[0032] Figure 7 FIG. shows a flowchart of a method implemented at a terminal device according to an exemplary embodiment of the present disclosure;

[0033] Figure 8 FIG. shows an example flowchart of a method implemented at a network device according to an exemplary embodiment of the present disclosure;

[0034] Figure 9 FIG. shows an example simplified block diagram of an apparatus suitable for implementing an embodiment of the present disclosure; and

[0035] Figure 10 FIG. shows an example block diagram of an example computer-readable medium according to some embodiments of the present disclosure.

[0036] In all the figures, the same or similar reference numerals denote the same or similar elements. Detailed Description of the Embodiments

[0037] Now, the principles of the present disclosure will be described with reference to some exemplary embodiments. It should be understood that these embodiments are described for illustrative purposes only and help those skilled in the art to understand and implement the present disclosure, without implying any limitation to the scope of the present disclosure. The present disclosure described herein can be implemented in various ways other than those described below.

[0038] In the following description and claims, unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs.

[0039] References in the present disclosure to "an embodiment", "embodiment", "exemplary embodiment", etc. indicate that the described embodiment may include a particular feature, structure, or characteristic, but not necessarily every embodiment includes the particular feature, structure, or characteristic. Moreover, these phrases do not necessarily refer to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with an embodiment, it should be considered that such feature, structure, or characteristic is within the knowledge of those skilled in the art in connection with other embodiments, whether or not explicitly described.

[0040] It should be understood that although the terms "first" and "second" etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. As used herein, the term "and / or" includes any and all combinations of one or more of the listed terms.

[0041] The terms used herein are for the purpose of describing particular embodiments only and are not intended to limit the example embodiments. As used herein, unless the context clearly dictates otherwise, the singular forms "a", "an" and "the" also include the plural forms. It should also be understood that when the terms "comprises", "has" and / or "includes" are used herein, they specify the presence of the stated features, elements and / or components, etc., but do not preclude the presence or addition of one or more other features, elements, components and / or combinations thereof.

[0042] As used in this application, the term "circuitry" may refer to one or more or all of the following:

[0043] (a) A pure hardware circuit implementation (such as an implementation only in analog and / or digital circuitry) and

[0044] (b) A combination of hardware circuits and software, such as, where applicable:

[0045] (i) A combination of (one or more) analog and / or digital hardware circuits and software / firmware, and

[0046] (ii) Any part of (one or more) hardware processors (including (one or more) digital signal processors) with software, software and (one or more) memories, which work together to enable a device (such as a mobile phone or a server) to perform various functions) and

[0047] (c) (One or more) hardware circuits and / or (one or more) processors that require software (e.g., firmware) to operate, such as (one or more) microprocessors or a part of (one or more) microprocessors, but the software may not be present when the operation does not require software

[0048] This definition of circuitry applies to all uses of the term herein, including all uses in any claims. As a further example, as used in this application, the term "circuitry" also encompasses an implementation of only hardware circuits or processors (or one or more processors) or a part of a hardware circuit or processor and the software and / or firmware attached thereto. By way of example and where applicable to a particular claim element, the term "circuitry" also encompasses a baseband integrated circuit or a processor integrated circuit for a mobile device, or a similar integrated circuit in a server, a cellular network device or other computing or network device.

[0049] As used herein, the term "communication network" refers to a network that follows any suitable communication standard, such as Long-Term Evolution (LTE), LTE-Advanced (LTE-A), Wideband Code Division Multiple Access (WCDMA), High-Speed Packet Access (HSPA), NarrowBand Internet of Things (NB-IoT), etc. Additionally, communication between a terminal device and a network device in a communication network can be performed according to any suitable generation of communication protocol, including but not limited to the third generation (3G), fourth generation (4G), 4.5G, fifth generation (5G) communication protocol, 5G-A, and / or later communication protocols. Embodiments of the present disclosure can be applied to various communication systems. Given the rapid development of communication, there will of course also be future types of communication technologies and systems in which the present disclosure can be embodied. It should not be regarded as limiting the scope of the present disclosure to the above systems and / or later communication protocols.

[0050] As used herein, the term "network device" refers to a node in a communication network through which a terminal device accesses the network and receives services therefrom. The network device can refer to a base station (BS) or an access point (AP), such as Node B (NodeB or NB), evolved Node B (eNodeB or eNB), NR NB (also known as gNB), Remote Radio Unit (RRU), Radio Head (RH), Remote Radio Head (RRH), relay, low-power node (such as femto, pico, etc.), depending on the terminology and technology applied.

[0051] The term "terminal device" refers to any terminal device capable of wireless communication. By way of example and not limitation, the terminal device may also be referred to as a communication device, user equipment (UE), subscriber station (SS), portable subscriber station, mobile station (MS), or access terminal (AT). The terminal device may include but is not limited to mobile phones, cellular phones, smart phones, Voice over Internet Protocol (VoIP) phones, wireless local loop phones, tablets, wearable terminal devices, personal digital assistants (PDA), portable computers, desktop computers, image capture terminal devices such as digital cameras, game terminal devices, music storage and playback devices, in-vehicle wireless terminal devices, wireless endpoints, mobile stations, laptop embedded devices (LEE), laptop mounted devices (LME), USB dongles, smart devices, wireless customer premise equipment (CPE), Internet of Things (IoT) devices, watches or other wearable devices, head-mounted displays (HMD), vehicles, drones, medical devices and applications (e.g., remote surgery), industrial devices and applications (e.g., robots and / or other wireless devices operating in an industrial and / or automation processing chain environment), consumer electronic devices, devices operating on commercial and / or industrial wireless networks, etc. In the following description, the terms "terminal device", "communication device", "terminal", "user equipment", and "UE" may be used interchangeably.

[0052] As described above, the improvement in the assessment of the cell quality associated with the terminal device is a key aspect related to communication performance. Communication systems operate on an increasing number of higher frequency bands. In 5G systems, higher frequency bands may require a larger number of base stations to provide coverage, but not limited thereto. Network cells are becoming increasingly dense, and the frequency bands are thus also becoming more numerous. For a multi-frequency network, when communicating with the serving cell, the terminal device can easily measure the signals associated with other cells transmitted on the same frequency as the serving cell. However, when communicating with the serving cell, the terminal device cannot perform measurements on other cells configured with a carrier frequency or RAT different from that of the serving cell. Even when performing intra-frequency measurements, the terminal device in a 5G system cannot perform measurements outside the current active bandwidth part (BWP) of the terminal device.

[0053] To measure neighboring cells operating on a carrier frequency different from that of the serving cell (which can also be referred to as inter-frequency measurement) and / or neighboring cells operating on a RAT different from that of the serving cell (which can also be referred to as inter-RAT measurement), the communication service between the terminal device and the serving cell must be suspended. The time duration during which the terminal device suspends the communication service with the serving cell and performs inter-frequency measurement or inter-RAT measurement is called a Measurement Gap, which is defined in the 3GPP specification. In addition, the terminal device sends the measurement report of the inter-frequency measurement or inter-RAT measurement to the corresponding network device. In other words, to perform inter-frequency measurement and / or inter-RAT measurement, the terminal device needs to suspend the communication service with the serving cell. In this way, the terminal device may consume additional power and resources due to inter-frequency measurement and / or inter-RAT measurement.

[0054] In view of the above and to improve the performance of the communication system, a scheme for estimating neighboring cells is proposed. In this scheme, the network device receives at least one of the reference signal received power (RSRP) and the reference signal received quality (RSRQ) associated with the serving cell of the terminal device from the terminal device. Then, based on at least one of the RSRP and RSRQ associated with the serving cell, the network device uses a machine learning (ML) or artificial intelligence (AI) model to determine the signal quality level associated with the neighboring cells of the serving cell. In this case, the network device can directly determine the quality of the neighboring cells without having to receive the measurement report of the neighboring cells from the terminal device. Furthermore, the terminal device can omit the steps of measuring the neighboring cells and transmitting the measurement report.

[0055] In this way, based only on the RSRP and RSRQ associated with the serving cell, the network device can accurately determine the signal quality level associated with the neighboring cell. Thus, the terminal device does not need to perform cross-frequency and / or cross-RAT measurements and can therefore save measurement power and reporting resources.

[0056] The principles and embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. Figure 1A An example network environment 100 in which example embodiments of the present disclosure can be implemented is illustrated. The environment 100 can be part of a communication network, which includes a terminal device and a network device.

[0057] As Figure 1A illustrated, the network environment 100 can include a network device 110, a terminal device 120, and another terminal device 130. The network device 110 can provide multiple cells, and each cell can have a corresponding carrier frequency or RAT, but is not limited thereto. For illustrative purposes only and not by way of limitation, in the example shown in FIG. 1, the cell 113 provided by the network device 110 can have a lower carrier frequency and a larger coverage area. Further, the cells 115 and 117 provided by the network device 110 can have higher carrier frequencies to enhance the traffic throughput with the terminal devices 120 and 130. In addition, the cell 115 can be configured with a different RAT from the cell 117. As mentioned above, the terminal device 120 can suspend communication with the serving cell 115 and measure the signals sent from neighboring cells (e.g., the cell 113 or the cell 117).

[0058] It should be understood that the number of network devices and terminal devices given is for illustrative purposes only and does not imply any limitation. The system 100 can include any suitable number of network devices and / or terminal devices suitable for implementing the embodiments of the present disclosure. Although not shown, it should be understood that one or more terminal devices can be located in the environment 100.

[0059] Communication in the network environment 100 can be implemented according to any suitable communication protocol, including but not limited to the third generation (3G), fourth generation (4G), fifth generation (5G), advanced 5G or higher (6G), wireless local area network communication protocols (such as those of the Institute of Electrical and Electronics Engineers (IEEE) 802.11, etc.), and / or any other currently known or future-developed protocols. In addition, the communication can utilize any suitable wireless communication technology, including but not limited to: multiple-input multiple-output (MIMO), orthogonal frequency division multiplexing (OFDM), time division multiplexing (TDM), frequency division multiplexing (FDM), code division multiplexing (CDM), Bluetooth, ZigBee, machine type communication (MTC), enhanced mobile broadband (eMBB), massive machine type communication (mMTC), ultra-reliable low latency communication (URLLC), carrier aggregation (CA), dual connectivity (DC), and new unlicensed radio (NR-U) technology.

[0060] Figure 1B An example measurement gap for inter-frequency measurement and radio access technology (RAT) inter-measurement is illustrated.

[0061] As mentioned above, the measurement gap gives the terminal device the opportunity to perform measurements on downlink signals in order to obtain inter-frequency or RAT inter-measurements. In addition, the measurement gap repetition period defines the period at which the measurement gap repeats. In some embodiments, the measurement gap repetition period can be configured. For example, in the 3GPP specification, the gap repetition period can be configured to 20, 40, 80, and 160 milliseconds. Figure 1B The example in shows measurement gaps within subframes #4, #5, #6, and #7, with a measurement gap repetition period of 40 milliseconds, and subframes #4, #5, #6, and #7 having subframe numbers (SFNs) of 20, 21, 22, and 23, respectively. Figure 1B Another subsequent measurement gap with an SFN of 24 in subframe #8 is also shown. The time interval between this measurement gap and the subsequent measurement gap is equal to the configured measurement gap repetition period.

[0062] Thus, a shorter measurement gap repetition period results in more measurements and will therefore require more dedicated time to perform the measurements. During the measurement gap, the terminal device cannot communicate with the serving cell, and the average throughput will also decrease accordingly. To reduce the cost of cross-frequency measurements, the present disclosure provides a process for predicting the signal strength and quality of inter-frequency carriers and RAT inter-carriers, which is based only on measurements on the serving carrier. Simulations in actual network deployments show that the prediction accuracy is up to 98.5%.

[0063] Figure 2Illustrated is an example signaling procedure 200 for estimating the quality of an adjacent cell according to some embodiments of the present disclosure. For the purpose of discussion, the procedure 200 will be described with reference to Figure 1A to describe the procedure 200. It should be understood that although the procedure 200 has been described in the Figure 1A communication environment 100, the procedure 200 can also be applied to other communication scenarios.

[0064] In the signaling procedure 200, the terminal device 120 (or the terminal device 130, but not limited thereto, and the following embodiments will be discussed with reference to the terminal device 120) transmits (210) at least one of the RSRP and RSRQ associated with the serving cell 115 of the terminal device 120.

[0065] In some embodiments, the RSRP may be the Synchronization Signal Reference Signal Receiving Power (SS-RSRP). The SS-RSRP is the average power received from a single resource unit allocated to the synchronization signal reference signal, as shown in FIG. 3 (which may be further discussed below). Assume that terminal devices in the network (e.g., terminal devices 120 and 130) send SS-RSRP measurement reports to the network device. The RSRP may be the receiving power of any other reference signal, such as the Demodulation Reference Signal (DMRS), the Channel State Information Reference Signal (CSI-RS), etc., but not limited thereto. In one solution, the SS-RSRP report includes measurements of the serving cell and other measurements of up to eight adjacent cells of the primary carrier of the serving cell. However, in this case, the terminal device 120 needs to perform measurements on the adjacent cells and report the measurement reports on the corresponding resources.

[0066] According to some embodiments of the present disclosure, the terminal device 120 may only perform measurements on the serving cell (which should be performed as a matter of course and is an in-band measurement or an in-BWP measurement), and send measurement information about the serving cell (such as the RSRP of the serving cell) to the network device 110. In addition, the terminal device 120 may not perform measurements on other adjacent cells and send measurement information about other adjacent cells to the network device 110. Further, the network device 110 may only receive the RSRP (and / or RSRQ) associated with the serving cell without receiving the signal quality levels associated with the adjacent cells of the serving cell. In this case, measurement power and the corresponding uplink (UL) resources can be saved. For example, the gain in cell average throughput can be increased by up to 30.7%, and there is little impact on other network key performance indicators (KPIs) such as the Drop Rate (DR) and the Handover Scheduling Request (HOSR).

[0067] Additionally or alternatively, in some embodiments, the RSRQ may be the Secondary Synchronization Signal Reference Signal Received Quality (SS-RSRQ). In an example, the SS-RSRQ is determined by:

[0068]

[0069] where N is the number of resource blocks in the carrier measurement bandwidth, and RSSI is the Reference Signal Strength Indicator. Assume that a terminal device in the network sends an SS-RSRQ measurement report. The RSRQ may be the received quality of any other reference signal, such as the Demodulation Reference Signal (DMRS), the Channel State Information Reference Signal (CSI-RS), etc., but is not limited thereto. In one solution, the SS-RSRQ measurement report includes measurements of the serving cell and up to eight neighboring cells on the primary carrier. In this way, the signal quality from each cell in the network can be compared. This is crucial for load balancing, handover, and secondary cell selection.

[0070] According to some embodiments of the present disclosure, similarly, the terminal device 120 may perform only RSRQ measurements on the serving cell (which should be performed as a matter of course, and this is an in-frequency measurement or an in-BWP measurement), and send RSRQ measurement information about the serving cell (such as the RSRQ of the serving cell), without performing measurements on other neighboring cells and sending measurement information about other neighboring cells.

[0071] Returning to Figure 2 , the network device 110 receives (210) at least one of the RSRP and RSRQ associated with the serving cell 115 from the terminal device 120. Then, based on at least one of the RSRP and RSRQ, the network device 110 uses an ML or AI model to determine (220) the signal quality level associated with the neighboring cells of the serving cell 115.

[0072] In this way, the network device 110 can use only at least one of the RSRP and RSRQ associated with the serving cell 115 as an input parameter for the ML or AI model used.

[0073] In addition, the network device 110 may also determine the signal quality level associated with the neighboring cells based on the angle information of the terminal device 120. In one solution, the angle information may be determined according to the Precoding Matrix Indicator (PMI) information received from the terminal device 120. In this case, the angle information is determined by the terminal device 120, and the measurement accuracy at the terminal device may be poor. In addition, the terminal device 120 may calculate the angle information by using additional resources. Furthermore, additional power may be consumed accordingly.

[0074] In some embodiments of the present disclosure, the network device 110 may also determine a signal quality level associated with an adjacent cell based on the direction of arrival (DOA) of a signal transmitted from the terminal device 120 to the network device 110. In addition, the DOA is calculated at the network device 110. In this way, the measurement accuracy of the DOA can reach the level of 0.1 degree. In addition, the calculation step of the angle information may be omitted at the terminal device 120, and the corresponding cost can be saved.

[0075] Thus, the network device 110 may use only the DOA of the signal transmitted from the terminal device 120 and at least one of the RSRP and RSRQ associated with the serving cell 115 as input parameters for the used ML or AI model. Then, the network device 110 may determine the signal quality level associated with the adjacent cell through the output of the used ML or AI model. In some embodiments, the output of the ML or AI model may be at least one of the RSRP or RSRQ associated with the adjacent cell. For example, the output may be the RSRP of the inter-frequency carrier and the RSRP of the inter-RAT carrier, and / or the RSRQ of the inter-frequency carrier and the inter-RAT carrier.

[0076] The network device 110 can achieve good performance (which can be shown hereinafter) by using only such a small number of input parameters because these input parameters can implicitly indicate the position of the terminal device within the serving cell. For illustrative purposes, reference may be made to Figure 3A and Figure 3B to discuss these input parameters further.

[0077] Figure 3A Illustrates an example path loss of a reference signal transmitted from a network device according to some embodiments of the present disclosure. As Figure 3A shown, the SS-RSRP at the terminal device 120 is related to the path loss between the network device 110 and the terminal device 120. In addition, the path loss is related to the distance between the network device 110 and the terminal device 120.

[0078] Figure 3B Illustrates an example of the DOA of a signal transmitted from a terminal device to a network device according to some embodiments of the present disclosure. As Figure 3BAs shown, the DOA is the direction in which the propagation wave arrives at the network device, usually the location where a set of antenna arrays is located. The DOA reflects the orientation of the terminal device. For example, using beamforming technology, the network device can estimate a given direction of a signal in the horizontal and vertical directions. In this way, the above input parameters (RSRP, RSRQ, and / or DOA) can reflect the distance and orientation information of the terminal device within the serving cell 115. Therefore, the ML or AI model can accurately learn how to estimate the signal quality associated with neighboring cells based on carefully selected input parameters. In this way, the location of the terminal device 120 can be implicitly indicated or determined at the network device 110 without the need for the terminal device 120 to send dedicated location information. Additionally, since the RSRP and RSRQ associated with the serving cell already sent by the terminal device 120 can be used for other purposes, in the present disclosure, the RSRP and RSRQ are reused for estimating neighboring cells. Thus, the terminal device does not need to perform additional operations to measure neighboring cells.

[0079] Regarding different input parameters and different output parameters for the ML or AI model, Figures 4A - 4C the corresponding estimation performance is shown.

[0080] Figure 4A The figure illustrates example estimation results corresponding to different input parameters of a machine learning (ML) or artificial intelligence (AI) model according to some embodiments of the present disclosure. Different measurements associated with the serving cell can be used as input parameters to predict the RSRP and RSRQ of neighboring cells. Additionally, the estimation is performed on an ML or AI model (i.e., an extreme tree regression algorithm). As Figure 4A shown, adding DOA information as an input parameter can achieve better performance (lower RMSE) than using only RSRP or RSRQ as input parameters. RMSE (root mean square error) is used as a metric for the accuracy of estimation or inference. RMSE is calculated by the following formula:

[0081]

[0082] where y i is the i-th actual signal quality associated with the neighboring cell, and is the corresponding predicted value.

[0083] In Figure 4AIn the simulation (which used logs from gNodeB field tests and obtained over 26,000 data samples, of which 23,000 data samples were used to train the ML or AI model and 3,000 data samples were used for validation), the best performance can be obtained using RSRP, RSRQ, and DOA as input parameters. Specifically, the RMSE for predicting the RSRP of neighboring cells is less than 1.5%, and the RMSE for predicting the RSRQ of neighboring cells is approximately 1.3%.

[0084] Figure 4B FIG. illustrates example estimation results corresponding to different output parameters of a machine learning (ML) or artificial intelligence (AI) model according to some embodiments of the present disclosure. As Figure 4B shown, an output having an RSRQ associated with a neighboring cell can achieve better performance.

[0085] Additionally, Figure 4C FIG. illustrates an example comparison between the estimation results and the actual results according to some embodiments of the present disclosure. As Figure 4C shown, the predicted RSRP associated with neighboring cell 115 can perfectly fit the actual RSRP measured by the terminal device. Specifically, Table 1 below further lists the number of handover times and the corresponding handover ratios triggered respectively based on the predicted RSRP and the actual RSRP.

[0086] Table 1

[0087] Prior Art The Present Invention Number of Switches 615 602 Total 25858 25856 Switching Ratio 2.378％ 2.328％

[0088] The totals in Table 1 are the total number of predicted RSRP values and the total number of actual RSRP values respectively. As listed in Table 1, the predicted RSRP has little impact on the HO SR of the terminal device 120 as compared to the handover scheduling request (HO SR) triggered by the actual RSRP measured by the terminal device 120.

[0089] Still referring to Figure 2 , the network device 110 can adopt any existing ML or AI model. Alternatively, the network device 110 can also adopt an ML or AI model that will be designed, studied, and constructed in the future. In the example, the ML or AI model can include one of the following: an extra tree regression model, a random forest regression model, a linear regression model, a K-nearest neighbor regression model, or a deep neural network (DNN) model. For illustrative purposes, reference can be made to Figures 5A to 5D for further discussion of various ML or AI models.

[0090] Figure 5AIllustrates example estimation results corresponding to different machine learning (ML) or artificial intelligence (AI) models according to some embodiments of the present disclosure.

[0091] As Figure 5A shown, tree-based regression models or algorithms (e.g., random forest regression models and extremely randomized tree regression models) have better performance than linear regression, K-nearest neighbor regression, and deep neural network (DNN) models or algorithms. Without any limitation, Figure 5A the ML or AI models in

[0092] are trained based on the following parameters: the RSRP and RSRQ associated with the serving cell and the direction of arrival (DOA) of the terminal device as input data samples; the RSRP associated with the neighboring cell as the output of the model. In some embodiments, the neighboring cell model can be trained based on historical data of at least one of the RSRP and RSRQ measured by multiple terminal devices in their respective serving cells. Additionally, the ML or AI model can be further trained based on the DOA of the terminal device. In the simulation, the DNN model has the worst performance, and although different methods have been used to optimize the neural network and hyperparameters within the DNN model, the performance improvement is limited. Figure 5B Illustrates example operating costs corresponding to different machine learning (ML) or artificial intelligence (AI) models according to some embodiments of the present disclosure. As Figure 5B shown in the example of Figure 5B the operating cost or resource consumption of the DNN model, as well as that of the linear regression model, is higher than that of other ML or AI models. Specifically,

[0093] actually shows the complexity of each ML or AI model (which can correspond to running time, operating cost, or computational cost). The extremely randomized tree regression may require a higher computational cost than linear regression and the nearest neighbor algorithm. However, the computational cost of the extremely randomized tree regression is still lower than that of the random forest regression and the DNN model. In the simulation, the training running time of the extremely randomized tree regression is approximately 2.5 seconds. Figure 5A and Figure 5B shown, tree-based regression models or algorithms can achieve better performance with a shorter running time or lower computational cost.

[0094] Since the serving cell may have multiple different neighboring cells, and the RSRP and RSRQ associated with the neighboring cells depend at least in part on cell deployment and radio environment, predicting the RSRP and RSRQ of neighboring cells is a complex non-linear problem. Therefore, using a tree-based regression model can obtain better performance. As discussed above, the normalized RMSE of the tree-based regression model is approximately 1.47%.

[0095] As Figure 5A and Figure 5B shown in the simulations of Figure 5B , a tree-based regression model or algorithm can be a suitable ML or AI model for determining the signal quality level associated with neighboring cells. An extra tree is a powerful alternative random forest ensemble method, and it is an ensemble learning technique that can aggregate the results of different decorrelated decision trees similar to a random forest. In some cases, an extra tree model can achieve better performance than a typical random forest model. In some embodiments, the network device 110 can build an extra tree regression model by training multiple trees based on multiple training data sets, which will be further discussed with reference to Figure 5C . Figure 5C Further discussion.

[0096] Figure 5C FIG. illustrates an example extra tree regression model according to some embodiments of the present disclosure.

[0097] As Figure 5C shown in Figure 5C , the extra tree regression model can split the training data into N groups (which is also the number of trees) of training data. In addition, for each group of training data, an extra tree regression model is built by training the corresponding decision tree model. Then, the extra tree regression model combines the decision trees into a random forest. The average result of all decision tree results is the output of the built extra tree regression model. As shown in Figure 5B , the running time of the extra tree regression model is about 2.5 seconds, which is too heavy for online training of an embedded system like a gNB. In some embodiments, the network device 110 can reconstruct the extra tree regression model to reduce the running time or complexity of the model. Figure 5B As shown in Figure 5B , the running time of the extra tree regression model is about 2.5 seconds, which is too heavy for online training of an embedded system like a gNB. In some embodiments, the network device 110 can reconstruct the extra tree regression model to reduce the running time or complexity of the model.

[0098] The input parameters of the above ML or AI model have high typical values and correlations. In this case, the structure of the extra tree regression model can be reconstructed to reduce the complexity. Furthermore, the reconstructed extra tree regression model can be used in an embedded system while only using limited computing resources.

[0099] The complexity of the extra tree regression model is at the O(n 2 ) level, where n is the number of trees in the extra tree regression model. This can be achieved by modifying the modeling parameters of the trees (decision trees) in the extra tree regression model, especially the number of trees. For illustrative purposes, the reconstruction of the extra tree regression model is discussed with reference to Figure 5D . Figure 5D The reconstruction of the extra tree regression model is discussed.

[0100] Figure 5D FIG. illustrates an example of the reconstructed extra tree regression model according to some embodiments of the present disclosure. As shown in Figure 5D Figure 5DAs shown, the network device 110 can reconstruct the extremely randomized tree regression model by reducing the number of trees within the extremely randomized tree regression model. In Figure 5D In the example shown, the first number of trees (510) within the extremely randomized tree regression model can be reduced to the second number of trees (520) in the constructed extremely randomized tree regression model. Additionally, for illustrative purposes, reference will be made to Figures 6A - 6E discuss the performance and operating cost of the reconstructed extremely randomized tree regression model.

[0101] Figure 6A illustrates an example comparison between the estimation results of an extremely randomized tree regression model and the estimation results of a reconstructed extremely randomized tree regression model according to some embodiments of the present disclosure.

[0102] As Figure 6A shown, block 610 represents the running time of a typical extremely randomized tree regression model, and block 620 represents the running time of the reconstructed extremely randomized tree regression model. It can be seen that the running time or complexity of the reconstructed extremely randomized tree regression model is significantly reduced. Specifically, approximately 90% of the complexity, running time, or computational cost is saved. Additionally, block 630 represents the RMSE performance of the typical extremely randomized tree regression model, and block 640 represents the RMSE performance of the reconstructed extremely randomized tree regression model. It can be seen that the performance difference between the typical extremely randomized tree regression model and the reconstructed extremely randomized tree regression model is negligible.

[0103] Returning to Figure 2 , as described above, the estimation of neighboring cells can be processed at the network device 110 without receiving information about neighboring cell measurements from the terminal device 120. In this way, there is no need to perform neighboring cell measurements and transmit the corresponding measurement reports at the terminal device 120. Thus, the resources in the measurement gap and the resources for transmitting the corresponding measurement reports can be reused.

[0104] In some embodiments, the network device 110 can send (225) first configuration information to the terminal device 120, and the first configuration information disables the measurement gap for performing inter-frequency carrier measurements and inter-RAT carrier measurements. Subsequently, after receiving (225) the first configuration information, the terminal device 120 can disable the measurement of neighboring cells. In this way, cross-frequency measurements that cause battery consumption can be avoided. Thus, battery consumption can be reduced and a green environment can be promoted.

[0105] Additionally or alternatively, the network device 110 may send (225) second configuration information to the terminal device 120, where the second configuration information indicates that resources in the measurement gap are reused for receiving transmissions from or sending transmissions to the serving cell. Further, after receiving (225) the second configuration information, the terminal device 120 may reuse the resources in the measurement gap to perform data services. In this way, the service throughput for the terminal device 120 can be improved. In some cases, the gain in cell average throughput can be as high as 30.7%, and there is little impact on other network KPIs (such as DR and HO SR). For illustrative purposes, reference is made to Figures 6B - 6D discusses the throughput gain.

[0106] Figure 6B illustrates the cell average throughput without measurement gaps according to some embodiments of the present disclosure. Figure 6C illustrates the cell average throughput with measurement gaps according to some embodiments of the present disclosure. As Figure 6B and Figure 6C shown, the cell average throughput without measurement gaps is improved compared to that with measurement gaps.

[0107] Figure 6D and Figure 6E illustrate the detailed gain values under different measurement gap configurations according to some embodiments of the present disclosure.

[0108] As Figure 6D shown, blocks 610-1, 610-2, and 610-3 represent the minimum cell throughput, average cell throughput, and maximum throughput without measurement gaps, respectively. Blocks 620-1, 620-2, and 620-3 represent the minimum cell throughput, average cell throughput, and maximum throughput with measurement gaps, respectively. In this example, when the measurement gap is disabled or reused for other purposes, the gain in average cell throughput is 30.7%.

[0109] Figure 6E illustrates the gain in resource block utilization according to some embodiments of the present disclosure. As Figure 6E shown, block 630 represents the utilization of physical resource blocks (PRBs) of the physical uplink shared channel (PUSCH) with measurement gaps. Block 640 represents the utilization of physical resource blocks (PRBs) of the physical uplink shared channel (PUSCH) without measurement gaps. In this example, when the measurement gap is disabled or reused for other purposes, significant savings in PRBs of the PUSCH can be achieved.

[0110] In view of this, for embodiments of the present disclosure, only the RSPR and RSRQ associated with the serving cell and the DOA of the terminal device are used as inputs to the ML or AI model. In addition, the prediction accuracy of the ML or AI model can be as high as 98.5%. In addition, the embodiments can be compatible with different ML or AI models, such as linear regression, K-nearest neighbor regression, random forest regression, extremely randomized tree regression, and deep neural network (DNN). Specifically, the extremely randomized tree regression can be reconstructed to have better performance and lower complexity. The embodiments in the present disclosure can also be easily used in other RAN-level user scenarios, such as load balancing, handover, Scell selection, etc. Without any limitation, although the embodiments in the present disclosure are provided in New Radio (NR), the embodiments have backward compatibility for LTE, 3G, 2G, and cross-RAT networks. In addition, since measurement reports are no longer required, UL PRB resource consumption can be reduced, and thus battery consumption for cross-frequency and cross-RAT measurements can be reduced accordingly.

[0111] Figure 7 FIG. 4 shows a flowchart of an exemplary method 700 implemented at a network device (e.g., network device 110) according to some embodiments of the present disclosure. For ease of discussion, method 700 will be described from the perspective of network device 110 with reference to FIG. 1.

[0112] At 710, network device 110 receives at least one of RSRP and RSRQ associated with the serving cell of terminal device 120 from terminal device 120. At 720, network device 110 uses an ML or AI model to determine a signal quality level associated with an adjacent cell of the serving cell based on at least one of RSRP and RSRQ associated with the serving cell.

[0113] In some embodiments, the signal quality level associated with the adjacent cell is further determined based on the DOA of the signal sent from terminal device 120 to network device 110, and the DOA is calculated by network device.

[0114] In some embodiments, the ML or AI model may include an extremely randomized tree regression model, a random forest regression model, a linear regression model, a K-nearest neighbor regression model, or a DNN model.

[0115] In some embodiments, the ML or AI model includes an extremely randomized tree regression model, and network device 110 may also construct the extremely randomized tree regression model by training multiple trees based on multiple training data sets; and reconstruct the extremely randomized tree regression model by reducing the number of multiple trees within the extremely randomized tree regression model.

[0116] In some embodiments, the ML or AI model is trained based on historical data of at least one of RSRP and RSRQ measured by multiple terminal devices in their respective serving cells.

[0117] In some embodiments, the network device 110 may also send first configuration information for disabling a measurement gap to the terminal device, where the measurement gap is used for performing inter-frequency carrier measurement and inter-radio RAT carrier measurement. Alternatively or additionally, the network device 110 may send second configuration information to the terminal device, where the second configuration information indicates that resources in the measurement gap are reused for receiving transmissions from the serving cell or for sending transmissions to the serving cell.

[0118] In some embodiments, the serving cell is configured with a first carrier frequency and a first RAT, and the neighboring cell is configured with at least one of the following: a second carrier frequency different from the first carrier frequency; or a second RAT different from the first RAT.

[0119] In some embodiments, the signal quality level associated with the neighboring cell includes at least one of RSRP or RSRQ.

[0120] Figure 8 A flowchart of an example method 800 implemented at a terminal device (e.g., terminal device 120) according to some embodiments of the present disclosure is shown. For the purpose of discussion, method 800 will be described from the perspective of terminal device 120 with reference to FIG. 1.

[0121] At 810, the terminal device 120 sends at least one of RSRP and RSRQ associated with the serving cell of the terminal device to the network device 110. At least one of RSRP and RSRQ associated with the serving cell will be used to determine the signal quality level associated with the neighboring cell of the serving cell based on the ML or AI model.

[0122] In some embodiments, the signal quality level associated with the neighboring cell is also determined based on the DOA of the signal sent from the terminal device 120 to the network device 110. The DOA is calculated by the network device 110.

[0123] In some embodiments, the ML or AI model is trained based on historical data of at least one of RSRP and RSRQ measured by multiple terminal devices in their respective serving cells.

[0124] In some embodiments, the ML or AI model includes an extreme random tree regression model, a random forest regression model, a linear regression model, a K-nearest neighbor regression model, or a DNN model.

[0125] In some embodiments, the terminal device 120 may further receive, from the network device 110, first configuration information for disabling a measurement gap that is used to perform inter-frequency carrier and inter-RAT carrier measurements; or receive second configuration information from the network device 110, where the second configuration information indicates that resources in the measurement gap are reused for receiving transmissions from or sending transmissions to the serving cell.

[0126] In some embodiments, the serving cell is configured with a first frequency carrier and a first RAT, and the neighboring cell is configured with a second frequency carrier different from the first frequency carrier; or a second RAT different from the first RAT.

[0127] In some embodiments, the signal quality level associated with the neighboring cell includes at least one of RSRP or RSRQ.

[0128] In some embodiments, an apparatus (e.g., the network device 110) capable of performing any operation of method 700 may include: components for receiving, from the terminal device 120, at least one of RSRP and RSRQ associated with the serving cell of the terminal device; and components for determining, based on at least one of RSRP and RSRQ associated with the serving cell, the signal quality level associated with the neighboring cell of the serving cell using an ML or AI model.

[0129] In some embodiments, the signal quality level associated with the neighboring cell is further determined based on the DOA of the signal sent from the terminal device 120 to the network device 110, and the DOA is calculated by the network device.

[0130] In some embodiments, the ML or AI model includes an extremely randomized trees regression model, a random forest regression model, a linear regression model, a K-nearest neighbors regression model, or a DNN model.

[0131] In some embodiments, the ML or AI model includes an extremely randomized trees regression model, and further causes the network device 110 to: build an extremely randomized trees regression model by training multiple trees based on multiple training data sets; and rebuild the extremely randomized trees regression model by reducing the number of multiple trees within the extremely randomized trees regression model.

[0132] In some embodiments, the ML or AI model is trained based on historical data of at least one of RSRP and RSRQ measured by multiple terminal devices in their respective serving cells.

[0133] In some embodiments, the apparatus further includes: a component for sending, to a terminal device, first configuration information for disabling a measurement gap for performing inter-frequency carrier measurement and inter-radio RAT carrier measurement; or a component for sending, to the terminal device, second configuration information indicating that resources in the measurement gap are reused for receiving transmissions from or sending transmissions to a serving cell.

[0134] In some embodiments, the serving cell is configured with a first carrier frequency and a first RAT, and the neighboring cell is configured with at least one of the following: a second carrier frequency different from the first carrier frequency; or a second RAT different from the first RAT.

[0135] In some embodiments, the signal quality level associated with the neighboring cell includes at least one of RSRP or RSRQ.

[0136] In some embodiments, the apparatus further includes a component for performing other steps in some embodiments of method 700. In some embodiments, the component includes at least one processor and at least one memory including computer program code, and the at least one memory and the computer program code are configured to, together with the at least one processor, implement the execution of the apparatus.

[0137] In some embodiments, an apparatus (e.g., terminal device 120) capable of performing any method 800 may include a component for sending, to a network device, at least one of the RSRP and RSRQ of the serving cell of the terminal device, and at least one of the RSRP and RSRQ of the serving cell will be used to determine the signal quality level of the neighboring cell of the serving cell based on an ML or AI model.

[0138] In some embodiments, the signal quality level associated with the neighboring cell is further determined based on the DOA of the signal sent from the terminal device 120 to the network device 110, and the DOA is calculated by the network device 110.

[0139] In some embodiments, the ML or AI model is trained based on historical data of at least one of the RSRP and RSRQ measured by multiple terminal devices in their respective serving cells.

[0140] In some embodiments, the ML or AI model includes one of the following: an extremely randomized tree regression model; a random forest regression model; a linear regression model; a K-nearest neighbor regression model; or a DNN model.

[0141] In some embodiments, the apparatus further comprises at least one of the following: a component for receiving, from a network device 110, first configuration information for disabling a measurement gap for performing inter-frequency carrier measurement and inter-radio access technology (RAT) carrier measurement; or a component for receiving, from the network device 110, second configuration information indicating that resources in the measurement gap are reused for receiving transmissions from or transmitting transmissions to a serving cell.

[0142] In some embodiments, the serving cell is configured with a first frequency carrier and a first RAT, and the neighboring cell is configured with at least one of the following: a second frequency carrier different from the first frequency carrier; or a second RAT different from the first RAT.

[0143] In some embodiments, the signal quality level associated with the neighboring cell includes at least one of RSRP or RSRQ.

[0144] In some embodiments, the apparatus further comprises a component for performing other steps in some embodiments of method 800. In some embodiments, the component includes at least one processor and at least one memory including computer program code, the at least one memory and the computer program code being configured to, together with the at least one processor, implement the execution of the apparatus.

[0145] Figure 9 is a simplified block diagram of a device 900 suitable for implementing embodiments of the present disclosure. The device 900 may be provided to implement a communication device, such as the network device 110 or the terminal device 120 shown in FIG. 1. As shown, the device 900 includes one or more processors 910, one or more memories 940 coupled to the processors 910, and one or more transmitters and / or receivers (TX / RX) 940 coupled to the processors 910.

[0146] The TX / RX 940 is for two-way communication. The TX / RX 940 has at least one antenna to facilitate communication. The communication interface may represent any interface required for communicating with other network elements.

[0147] The processor 910 may be of any type suitable for a local technical network and, by way of non-limiting example, may include one or more of the following: a general-purpose computer, a special-purpose computer, a microprocessor, a digital signal processor (DSP), and a processor based on a multi-core processor architecture. The device 900 may have multiple processors, such as an application-specific integrated circuit chip that is subordinate in time to a clock that synchronizes with a main processor.

[0148] The memory 920 may include one or more non-volatile memories and one or more volatile memories. Examples of non-volatile memories include, but are not limited to, read-only memory (ROM) 924, electrically programmable read-only memory (EPROM), flash memory, hard disk, compact disk (CD), digital video disk (DVD), and other magnetic storage devices and / or optical storage devices. Examples of volatile memories include, but are not limited to, random access memory (RAM) 922 and other volatile memories that do not persist during a power outage duration.

[0149] The program 930 includes executable instructions to be executed by the associated processor 910. The program 930 may be stored in the ROM 924. The processor 910 may execute any appropriate actions and processes by loading the program 930 into the RAM 922.

[0150] Embodiments of the present disclosure may be implemented by means of a program such that the device 900 may execute any process of the present disclosure as discussed with reference to Figures 2 to 8 The embodiments of the present disclosure may also be implemented by hardware or a combination of software and hardware.

[0151] In some embodiments, the program 930 may be tangibly embodied in a readable storage medium, which may be included in the device 900 (such as in the memory 920), or in other storage devices accessible by the device 900. The device 900 may load the program 930 from the storage medium into the RAM 922 for execution. The storage medium may include any type of tangible non-volatile memory, such as ROM, EPROM, flash memory, hard disk, CD, DVD, etc. Figure 10 An example of a storage medium 1000 in the form of a CD or DVD is shown. The storage medium has processor instructions 930 stored therein.

[0152] Generally, various embodiments of the present disclosure may be implemented in hardware or special-purpose circuits, software, logic, or any combination thereof. Some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software, which may be executed by a controller, a microprocessor, or other computing devices. Although the various aspects of the embodiments of the present disclosure are illustrated and described as block diagrams, flowcharts, or using some other graphical representation, it should be understood that, by way of non-limiting example, the blocks, devices, systems, techniques, or methods described herein may be implemented in hardware, software, firmware, special-purpose circuits or logic, general hardware or a controller or other computing devices, or some combination thereof.

[0153] The present disclosure also provides at least one program product tangibly stored on a non-transitory readable storage medium. The program product includes executable instructions, such as those included in program modules, which are executed in a device on a target real or virtual processor to perform the processes 200, methods 700, or 800 as described above with reference to Figure 2 FIGS. 4 and 5. Generally, program modules include routines, programs, libraries, objects, classes, components, data structures, etc. that perform particular tasks or implement particular abstract data types. In various embodiments, the functionality of program modules may be combined or split among program modules as desired. The machine-executable instructions of program modules may be executed within local or distributed devices. In a distributed device, program modules may be located in local and remote storage media.

[0154] The program code for performing the methods of the present disclosure may be written in any combination of one or more programming languages. The program code may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus such that, when the processor or controller executes the program code, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The program code may be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0155] In the context of the present disclosure, the program code or related data may be carried by any suitable carrier such that a device, apparatus, or processor can perform the various processes and operations as described above. Examples of carriers include signals, readable storage media, etc.

[0156] The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may include, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the readable storage medium include: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. As used herein, the term "non-transitory" is a limitation of the medium itself (i.e., tangible, rather than a signal), rather than a limitation of data storage persistence (e.g., RAM versus ROM).

[0157] In addition, although the operations are depicted in a particular order, this should not be construed as requiring that the operations be performed in the particular order shown or in sequential order, or that all of the illustrated operations be performed to achieve the desired result. In some scenarios, multitasking and parallel processing may be advantageous. Similarly, although several specific implementation details are included in the above discussion, these should not be construed as limitations on the scope of the present disclosure, but rather as descriptions of features that may be specific to particular embodiments. Certain features described in the context of separate embodiments may also be implemented in combination in a single embodiment. Conversely, the various features described in the context of a single embodiment may also be implemented separately in multiple embodiments or in any suitable sub-combination.

[0158] Although the present disclosure has been described in language specific to structural features and / or methodological acts, it is to be understood that the disclosure defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the above specific features and acts are disclosed as example forms of implementing the claims.

Claims

1. A network device, comprising: At least one processor; And At least one memory storing instructions which, when executed by the at least one processor, cause the network device to at least: Receive at least one of a reference signal received power (RSRP) and a reference signal received quality (RSRQ) associated with a serving cell of the terminal device from the terminal device; and Based on the at least one of the RSRP and the RSRQ associated with the serving cell, use a machine learning (ML) or artificial intelligence (AI) model to determine a signal quality level associated with an adjacent cell of the serving cell.

2. The network device according to claim 1, wherein the signal quality level associated with the adjacent cell is further determined based on a direction of arrival (DOA) of a signal sent from the terminal device to the network device, and wherein the DOA is calculated by the network device.

3. The network device according to claim 1 or 2, wherein the ML or AI model comprises one of the following: An extremely randomized tree regression model; A random forest regression model; A linear regression model; A K-nearest neighbor regression model; or A deep neural network (DNN) model.

4. The network device according to claim 3, wherein the ML or AI model comprises the extremely randomized tree regression model, and the network device is further caused to: Build the extremely randomized tree regression model by training a plurality of trees based on a plurality of training data sets; and Rebuild the extremely randomized tree regression model by reducing the number of the plurality of trees within the extremely randomized tree regression model.

5. The network device according to any one of claims 1 to 4, wherein the ML or AI model is trained based on historical data of at least one of RSRP and RSRQ measured by a plurality of terminal devices in respective serving cells.

6. The network device according to any one of claims 1 to 5, wherein the network device is further caused to perform at least one of the following: Send first configuration information for disabling a measurement gap for performing inter-frequency carrier measurement and radio access technology (RAT) inter-carrier measurement to the terminal device; or Send second configuration information to the terminal device, the second configuration information indicating that resources in the measurement gap are reused for receiving transmissions from the serving cell or for sending transmissions to the serving cell.

7. The network device according to any one of claims 1 to 6, wherein the serving cell is configured with a first carrier frequency and a first RAT, and wherein the adjacent cell is configured with at least one of the following: A second carrier frequency different from the first carrier frequency; or A second RAT different from the first RAT.

8. The network device according to any one of claims 1 to 7, wherein the signal quality level associated with the adjacent cell includes: At least one of RSRP or RSRQ.

9. A terminal device, comprising: At least one processor; And At least one memory storing instructions which, when executed by the at least one processor, cause the terminal device to at least: Transmit at least one of a reference signal received power (RSRP) and a reference signal received quality (RSRQ) associated with the serving cell of the terminal device to the network device. The at least one of the RSRP and the RSRQ associated with the serving cell will be used to determine a signal quality level associated with an adjacent cell of the serving cell based on a machine learning (ML) or artificial intelligence (AI) model.

10. The terminal device according to claim 9, wherein the signal quality level associated with the adjacent cell is further determined based on a direction of arrival (DOA) of a signal transmitted from the terminal device to the network device, and wherein the DOA is calculated by the network device.

11. The terminal device according to claim 9 or 10, wherein the ML or AI model is trained based on historical data of at least one of RSRP and RSRQ measured by a plurality of terminal devices in respective serving cells.

12. The terminal device according to any one of claims 9 to 11, wherein the ML or AI model is trained based on measurement information about the serving cell, and the ML or AI model includes one of the following: An extremely randomized tree regression model; A random forest regression model; A linear regression model; A K-nearest neighbor regression model; or A deep neural network (DNN) model.

13. The terminal device according to any one of claims 9 to 12, wherein the terminal device is further caused to perform at least one of the following: Receive first configuration information from the network device for disabling a measurement gap for performing inter-frequency carrier measurement and radio access technology (RAT) inter-carrier measurement; or Receive second configuration information from the network device, the second configuration information indicating that resources in the measurement gap are reused for receiving a transmission from the serving cell or for transmitting a transmission to the serving cell.

14. The terminal device according to any one of claims 9 to 13, wherein the serving cell is configured with a first frequency carrier and a first RAT, and wherein the adjacent cell is configured with at least one of the following: A second frequency carrier different from the first frequency carrier; or A second RAT different from the first RAT.

15. The terminal device according to any one of claims 9 to 14, wherein the signal quality level associated with the adjacent cell includes: At least one of RSRP or RSRQ.

16. A method, comprising: At a network device, receive at least one of a reference signal received power (RSRP) and a reference signal received quality (RSRQ) associated with the serving cell of the terminal device; and Based on the at least one of the RSRP and the RSRQ associated with the serving cell, use a machine learning (ML) or artificial intelligence (AI) model to determine a signal quality level associated with an adjacent cell of the serving cell.

17. A method, comprising: At a terminal device, transmit at least one of a reference signal received power (RSRP) and a reference signal received quality (RSRQ) of the serving cell of the terminal device to the network device, and At least one of the RSRP and the RSRQ of the serving cell will be used to determine a signal quality level of a neighboring cell of the serving cell based on a machine learning (ML) or artificial intelligence (AI) model.

18. An apparatus, comprising: means for receiving from a terminal device at least one of a reference signal received power (RSRP) and a reference signal received quality (RSRQ) associated with a serving cell of the terminal device; and means for determining a signal quality level associated with a neighboring cell of the serving cell using a machine learning (ML) or artificial intelligence (AI) model based on at least one of the RSRP and the RSRQ associated with the serving cell.

19. An apparatus, comprising: means for sending to a network device at least one of a reference signal received power (RSRP) and a reference signal received quality (RSRQ) of a serving cell of a terminal device, and at least one of the RSRP and the RSRQ of the serving cell will be used to determine a signal quality level of a neighboring cell of the serving cell based on a machine learning (ML) or artificial intelligence (AI) model.

20. A non-transitory computer-readable medium, comprising program instructions stored thereon for at least performing the method according to claim 16 or 17.

Images