Method and base station for UE selection for mobile load balancing in a RAN network

Abstract

The object of the present invention is to provide a method, a base station and a RAN intelligent controller for UE selection for mobile load balancing in a RAN. According to a set of continuous signal-related measurement information of each UE in all cells of the RAN in the current time window, determine the predicted moving direction of each UE relative to its serving cell; when one of the cells is overloaded, determine the candidate UEs in this cell to be switched to other cells; according to the predicted moving direction of the candidate UEs relative to the current cell, determine the target UE therefrom. Compared with the prior art, the present invention introduces the prediction of the moving direction of the UE to assist in the selection of the cut-off UE during the mobile load balancing process. This moving direction will be used as a key criterion for UE selection. For example, UEs moving towards the edge of the serving cell are more likely to be selected to unload from the current cell. The combination of static measurement and future direction prediction can greatly improve the accuracy of UE selection, thereby improving the performance of mobile load balancing.

Description

Technical Field

[0001] The present invention relates to the technical field of mobile load balancing, and in particular to a technology for UE selection for mobile load balancing in a RAN network. Background Art

[0002] Mobile load balancing (MLB) is a key component of a radio access network (RAN). It provides a service for distributing cell load among multiple network entities by moving some user equipment (UE) of an overloaded cell to an appropriate adjacent idle cell. Thus, the workload of each network entity can be balanced. Consequently, the end-user experience (QoE) and network quality (QoS) in such scenarios can also be improved.

[0003] UE selection is a main step in the mobile load balancing process. The purpose of UE selection is to find candidate UEs to be off-loaded from an overloaded cell. Usually, a set of UE selection criteria is predefined and will be used to determine which UE can be added to the candidate list of the MLB. Therefore, how to design the UE selection criteria is crucial. Summary of the Invention

[0004] The object of the present invention is to provide a method, a base station, and a RAN intelligent controller for UE selection for mobile load balancing in a RAN.

[0005] According to one aspect of the present invention, there is provided a method for UE selection for mobile load balancing in a RAN. When an overload occurs in a current cell, the base station determines candidate UEs in the cell to be switched to other cells; the base station obtains a predicted moving direction of the candidate UEs relative to the current cell, and the predicted moving direction is determined according to a set of continuous signal-related measurement information of the candidate UEs in a current time window; the base station determines a target UE from the candidate UEs according to the predicted moving direction.

[0006] According to one aspect of the present invention, there is also provided a method for assisting UE selection for mobile load balancing in a RAN. The RAN intelligent controller continuously collects signal-related measurement information of each UE in each cell at the current moment from each base station; the RAN intelligent controller determines a predicted moving direction of each UE relative to its serving cell according to a set of continuous signal-related measurement information of each UE in a current time window; the RAN intelligent controller feeds back the determined predicted moving direction of each UE relative to its serving cell to its corresponding base station.

[0007] According to one aspect of the present invention, there is also provided a base station for UE selection for mobile load balancing of the RAN. The base station includes a selection device, an acquisition device, and a screening device. Among them, the selection device is used to determine candidate UEs to be handed over to other cells in the cell when the cell is overloaded. The acquisition device is used to acquire the predicted moving direction of the candidate UE relative to the current cell, and the predicted moving direction is determined according to a set of continuous signal-related measurement information of the candidate UE in the current time window. The screening device is used to determine the target UE from the candidate UEs according to the predicted moving direction.

[0008] According to one aspect of the present invention, there is also provided a RAN intelligent controller for UE selection for mobile load balancing of the RAN. The RAN intelligent controller includes a collection device and a prediction device. Among them, the collection device is used to continuously collect signal-related measurement information of each UE in each cell at the current moment from each base station. The prediction device is used to determine the predicted moving direction of each UE relative to its serving cell according to a set of continuous signal-related measurement information of each UE in the current time window, and feed back the determined predicted moving direction of each UE relative to its serving cell to its corresponding base station.

[0009] Compared with the prior art, the present invention introduces the prediction of the UE moving direction to assist in the selection of the cut-off UE during the mobile load balancing process. This moving direction will be used as a key criterion for UE selection. For example, among the UEs that meet the basic conditions of load balancing, the UEs moving towards the edge of the serving cell are more likely to be selected to unload from the current cell. The combination of static measurement and future direction prediction can greatly improve the accuracy of UE selection, thereby improving the performance of mobile load balancing. Brief Description of the Drawings

[0010] By reading the detailed description of the non-limiting embodiments with reference to the following drawings, other features, objectives, and advantages of the present invention will become more obvious:

[0011] Figure 1 Schematic diagram showing the effect that the existing UE selection algorithm fails to consider the UE moving direction;

[0012] Figure 2 Flowchart showing the method for a base station to select a UE for mobile load balancing according to an embodiment of the present invention;

[0013] Figure 3 Schematic diagram showing the effect that the UE selection algorithm according to the present invention considers the UE moving direction;

[0014] Figure 4 Schematic diagram showing the process for a base station to obtain the predicted moving direction of a UE by interacting with the RIC according to an embodiment of the present invention;

[0015] Figure 5 Schematic diagram of a base station apparatus for UE selection for mobile load balancing according to an embodiment of the present invention;

[0016] Figures 6(a) and (b) respectively show schematic diagrams of an apparatus for UE selection for mobile load balancing of a RAN according to an embodiment of the present invention.

[0017] Identical or similar reference numerals in the drawings represent identical or similar components. Detailed implementation manners

[0018] Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments of the present invention are described as apparatuses represented by block diagrams and processes or methods represented by flowcharts. Although the flowcharts describe the operation processes of the present invention as sequential processes, many of the operations can be implemented in parallel, concurrently, or simultaneously. In addition, the order of the operations can be rearranged. The processes of the present invention can be terminated when their operations are completed, but can also include additional steps not shown in the flowcharts. The processes of the present invention can correspond to methods, functions, procedures, subroutines, subprograms, etc.

[0019] The methods shown by the flowcharts and the apparatuses shown by the block diagrams discussed below can be implemented by hardware, software, firmware, middleware, microcode, hardware description language, or any combination thereof. When implemented by software, firmware, middleware, or microcode, the program code or code segments for performing the necessary tasks can be stored in a machine or a computer-readable medium such as a storage medium. One or more processors can execute the necessary tasks.

[0020] Similarly, it will also be understood that any flowchart, flow diagram, state transition diagram, and the like, representing various processes, can be fully described as program code stored in a computer-readable medium and thus executed by a computing device or a processor, whether or not these computing devices or processors are explicitly shown.

[0021] Herein, the term "storage medium" can represent one or more devices for storing data, including read-only memory (ROM), random access memory (RAM), magnetic RAM, core memory, disk storage media, optical storage media, flash devices, and / or other machine-readable media for storing information. The term "computer-readable medium" can include, but is not limited to, portable or fixed storage devices, optical storage devices, and various other media capable of storing and / or containing instructions and / or data.

[0022] A code segment can represent a procedure, function, subroutine, program, routine, subroutine, module, package, class, or any combination of instructions, data structures, or program descriptions. A code segment can be coupled to another code segment or a hardware circuit by passing and / or receiving information, data, arguments, parameters, or stored content. Information, arguments, parameters, data, etc., can be passed, forwarded, or emitted via any suitable means including storage sharing, information transfer, token passing, network transmission, etc.

[0023] The specific structural and functional details disclosed herein are merely representative and are for the purpose of describing exemplary embodiments of the present invention. However, the present invention can be embodied in many alternative forms and should not be construed as limited only to the embodiments set forth herein.

[0024] It should be understood that although terms such as "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit can be called the second unit, and similarly the second unit can be called the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0025] The terms used herein are only for the purpose of describing specific embodiments and are not intended to limit the exemplary embodiments. Unless the context clearly dictates otherwise, the singular forms "a", "an" used herein are also intended to include the plural. It should also be understood that the terms "comprises" and / or "comprising" specify the presence of the stated features, integers, steps, operations, units, and / or components, and do not preclude the presence or addition of one or more other features, integers, steps, operations, units, components, and / or combinations thereof.

[0026] It should also be noted that in some alternative implementations, the functions / actions mentioned may occur in a different order than that indicated in the figures. For example, depending on the functions / actions involved, two successive figures shown may actually be executed substantially simultaneously or sometimes in the reverse order.

[0027] Specific embodiments of the present invention will be further described below in conjunction with the accompanying drawings.

[0028] The criteria for UE selection are very important for the performance of MLB. The problem with the current UE selection criteria is that it only considers static measurement factors of the UE state and does not consider factors such as the next UE location and the moving direction of the UE.

[0029] For example, refer to Figure 1, where UE11 is currently moving towards the cell edge and UE12 is moving towards the cell center. In the current UE selection criteria, since it does not consider the moving direction of the UE, UE12 which is closer to the cell edge is considered to have a higher priority to be cut off from this cell. However, in fact, in this scenario, UE11 should be cut off from the current cell more. And if UE12 is selected to be cut off from the current cell, operations similar to ping-pong handover will occur next.

[0030] When the load of the serving cell is high, it will trigger mobile load balancing. The base station will select target users for offloading. The criteria of the UE selection algorithm are usually based on operator configuration and UE measurements of the serving cell or target cell. For example, RSRP (Reference Signal Receiving Power), RSRQ (Reference Signal Receiving Quality).

[0031] Once the base station detects that the current cell is overloaded, some UEs will be offloaded to other cells. The criteria for candidate UE selection are static factors, such as A4 measurement reports. The selected UEs will be guided to target cells with higher available capacity through handovers.

[0032] For UEs in the connected mode, the current mobile load balancing process is as follows:

[0033] 1. The base station performs load monitoring and cell available capacity calculation, which includes load measurement, cell load evaluation, and active load balancing status;

[0034] 2. The base station selects candidate UEs. Among them, all UEs that meet specific conditions become candidates. Start A4 measurement for candidate UEs, and use the UEs that report A4 events as target UEs;

[0035] 3. The base station selects the target cell. Cells with higher available capacity will have higher priority;

[0036] 4. The base station performs load balancing, and the handover of the target UE towards the target cell is triggered.

[0037] Obviously, the above existing UE selection mechanism only considers the serving cell / neighboring cell load, RSRP / RSRQ, operator configuration, etc. The moving direction of the UE relative to its serving cell is not a consideration factor in the existing UE selection algorithm.

[0038] To this end, the present invention introduces the prediction of the UE's moving direction to assist in the selection of offloading UEs in mobile load balancing. This moving direction will be used as a key criterion for UE selection. For example, a UE moving towards the edge of the serving cell has more chances to be selected for offloading. The combination of static measurement and future prediction can significantly improve the accuracy of UE selection, thereby improving the performance of mobility load balancing.

[0039] Specifically, the present invention introduces a neural network based on deep learning to predict the UE's moving direction. This neural network model will hereinafter be referred to as the prediction model. The RAN collects the measurement reports of each UE, and these measurement reports will be used as the input of the prediction model to predict the UE's moving direction. That is, the collected historical measurements of the UE are used as the input features of the prediction model, while the moving direction of the UE relative to its serving cell is used as the output of the prediction model. Here, the signal-related measurement information of the UE may include, for example, CQI (Channel Quality Indication) / RSSI (Received Signal Strength Indication) / RSRP / RSRQ in the LTE RAN network. Therefore, the present invention does not require any modification to the UE side, and all these measurements already exist in the current network.

[0040] Refer to Figure 2 , which shows a flowchart of a method for a base station to select a UE for mobile load balancing according to an embodiment of the present invention.

[0041] As Figure 2 shown, in step 210, when the current cell is overloaded, the base station determines candidate UEs in this cell to be handed over to other cells; in step 220, the base station obtains the predicted moving direction of the candidate UEs relative to the current cell, and the predicted moving direction is determined according to a set of continuous signal-related measurement information of the candidate UEs in the current time window; in step 230, the base station determines the target UE from the candidate UEs according to the predicted moving direction.

[0042] Specifically, in step 210, when the current cell is overloaded, the base station determines candidate UEs in this cell to be handed over to other cells. When the serving cell is in an active load balancing state, the base station monitors the load of the serving cell and calculates the available capacity of the cell. When it is found that the serving cell is overloaded, the UEs that meet the basic conditions of load balancing are used as candidate UEs to be handed over to other cells. For example, the base station selects candidate UEs to be handed over to other cells according to the A4 measurement reports of each UE in the cell.

[0043] In step 220, the base station obtains a predicted moving direction of the candidate UE relative to the current cell, where the predicted moving direction is determined according to a set of continuous signal-related measurement information of the candidate UE in the current time window.

[0044] A prediction model may be configured in the base station, and the prediction model may predict the moving direction of the UE relative to the current cell according to a set of continuous signal-related measurement information of the UE in the current time window.

[0045] The signal-related measurement information of the UE may include at least any one of the following: CQI / RSSI / RSRP / RSRQ.

[0046] According to an example of the present invention, the predicted moving direction may include two directions: a moving direction toward the center of the serving cell and a moving direction toward the edge of the serving cell.

[0047] The prediction model can be based on a set of continuous signal-related measurement information {M t ,M t-1 ,M t-2 ,…} to predict the moving direction D of the UE at time t t By taking a set of historical signal-related measurement information of the UE as the input of the prediction model and taking the moving direction of the UE relative to its serving cell as the output of the prediction model, the prediction model can be trained to predict the moving direction of the UE relative to its serving cell.

[0048] Specifically, the prediction model can be constructed based on a neural network. The present invention can use a deep neural network or a recurrent neural network. t represents the signal correlation measurement value of a UE at time t. Similarly, M t-1 Indicates the signal correlation measurement value of the UE at time t-1, etc. And, D t Indicates the moving direction of the UE relative to its serving cell at time t.

[0049] The prediction model uses a set of UE signal-related measurement information {M} collected in a specified time window to predict the movement direction D of the UE relative to its serving cell. For example, the prediction model can use {M t , M t-1 , M t-2 , …} to predict the moving direction D of UE at time t t , and can use {M t+1 , M t , M t-1, …} to predict the moving direction D of the UE at time t+1 t+1 . The time window is a hyperparameter of the prediction model and can be adjusted during the training process.

[0050] According to an example of the present invention, for example, D can be defined by only two directions {ToCenter, ToEdge}. ToCenter indicates that the UE is moving towards the center of the serving cell, while ToEdge indicates that the UE is moving towards the edge of the serving cell. Therefore, in machine learning, it can be classified as a binary classification problem. However, in actual implementation, more moving directions may be defined, which will become a multi-classification problem.

[0051] The prediction model needs to be pre-trained. The prediction model can be obtained, for example, by training a classification neural network with labeled samples, where each sample is labeled with a set of consecutive signal-related measurement information of a UE in a predetermined time window and the moving direction of the UE relative to its serving cell.

[0052] There are multiple ways to construct the training dataset. First, the dataset in the RAN network simulator can be used, which has different path loss models. Second, training data can be collected from MDT (Minimization Drive Test) in the RAN network, where the signal-related measurement values and labels can be inferred from the UE location. Third, for GPS-enabled UEs, training data can be collected from the real-time network. Finally, if GPS is not available, training data can be collected from the real-time network through the handover state.

[0053] Here, the prediction model can continuously predict the moving direction of each UE in the current cell relative to the current cell, so that the predicted moving direction of the UE is always up-to-date, or it can only predict the moving direction of the candidate UE for the UE to select.

[0054] According to an example of the present invention, the base station can determine the predicted moving direction D of the candidate UE relative to the current cell at the current time t through the prediction model only when overload occurs t .

[0055] According to another example of the present invention, the base station can continuously determine the latest predicted moving direction of each UE relative to the current cell through the prediction model according to a set of consecutive signal-related measurement information of each UE in the current cell in the current time window, and update the latest determined predicted moving direction of each UE relative to the current cell to the UE movement status database. Accordingly, the base station can always obtain the latest predicted moving direction of the candidate UE from the UE movement status database.

[0056] Furthermore, the process may further include the following steps: continuously collect signal-related measurement information of each UE in the current cell at the current moment. By continuously collecting the signal-related measurement information of the UE at each moment, the base station can always obtain a set of UE signal-related measurement information of each UE in a current time window, thereby predicting the moving direction of the candidate UE in real time or continuously predicting the latest moving direction of each UE.

[0057] In step 230, the base station determines a target UE from the candidate UEs according to the predicted moving direction of the candidate UE relative to the current cell.

[0058] According to an example of the present invention, the base station traverses all candidate UEs, removes the candidate UEs whose predicted moving direction is towards the center of the current cell, and marks the candidate UEs whose predicted moving direction is towards the edge of the current cell as target UEs, so as to obtain the target UEs to be handed over to other cells.

[0059] With reference to Figure 3 , when selecting UEs during the "mobility load balancing" process, there are two candidate UEs that meet the basic conditions for offloading. At the same time, UE 32 is predicted to be moving from the edge of the serving cell towards the cell center, and UE 31 is predicted to be moving towards the cell edge direction. In this case, UE 32 will not be added to the target UE list, while UE 31 will be added to the target UE list.

[0060] Next, according to an embodiment of the present invention, Figure 2 The process shown may further include the following steps: hand over the target UE to other target cells. Here, adjacent cells with higher available capacity will have higher priority. After determining the target cell for handover, the base station communicates with the target base station about the handover process of the target UE to trigger the handover of the target UE towards the target cell.

[0061] According to another embodiment of the present invention, wherein the prediction model is arranged in a RAN intelligent controller (RIC, RAN Intelligence Controller), and the RIC is introduced by an ORAN (Open RAN organization) system. With reference to Figure 2 and Figure 4 , step 220 is further divided into the following steps 4201-4203, so that the base station obtains the predicted moving direction of the candidate UE relative to the current cell from the prediction model by interacting with the RAN intelligent controller.

[0062] In step 4201, the base station continuously transmits the signal-related measurement information of each UE in the current cell at the current moment to the RAN intelligent controller; accordingly, the RAN intelligent controller can continuously collect the signal-related measurement information of each UE in each cell at the current moment from each base station.

[0063] In step 4202, based on a set of continuous signal-related measurement information of each UE in each cell within the current time window, the RAN intelligent controller determines the predicted moving direction of each UE relative to its serving cell through a prediction model.

[0064] The prediction model in the RAN intelligent controller can predict the moving direction of each UE for all cells. For each cell, the prediction model can continuously predict the moving direction of each UE therein relative to the current cell.

[0065] In step 4203, the RAN intelligent controller feeds back the predicted moving direction of each UE in each cell relative to its serving cell to its corresponding base station; accordingly, the current base station can obtain the predicted moving direction of each UE in its cell relative to the current cell.

[0066] According to an example of the present invention, the base station may be arranged with a UE movement status database, and the predicted moving direction of each UE in the cell continuously fed back by the RAN intelligent controller can be updated to the UE movement status database. Accordingly, the base station can always obtain the latest predicted moving direction of each UE in its cell relative to the current cell, and obtain the latest predicted moving direction of the candidate UE from the UE movement status database for use in subsequent step 203 to screen out the target UE to be switched to the target cell from the candidate UEs.

[0067] According to the UE selection scheme of the present invention, the selection of the target UE can be more accurate. Compared with blind UE selection, the present invention can effectively eliminate potential ping-pong handovers.

[0068] Figure 5 Schematic diagram of a device of a base station for UE selection for mobile load balancing of a RAN showing an embodiment of the present invention.

[0069] As Figure 5 shown, the base station includes a selection device 521, an acquisition device 522, and a screening device 523. Among them, when the current cell is overloaded, the selection device 521 determines the candidate UEs in the cell to be switched to other cells; the acquisition device 522 acquires the predicted moving direction of the candidate UEs relative to the current cell, and the predicted moving direction is determined according to a set of continuous signal-related measurement information of the candidate UEs within the current time window; the screening device 523 determines the target UE from the candidate UEs according to the predicted moving direction.

[0070] Specifically, when the current cell is overloaded, the selection device 521 determines candidate UEs in this cell that are to be handed over to other cells. For example, the selection device 521 regards all UEs in the current cell that meet the basic conditions of mobile load balancing as candidate UEs.

[0071] Subsequently, the obtaining device 522 obtains the predicted moving direction of the candidate UE relative to the current cell, and the predicted moving direction is determined based on a set of consecutive signal-related measurement information of the candidate UE in the current time window.

[0072] The signal-related measurement information of the UE may include at least any one of the following: CQI / RSSI / RSRP / RSRQ.

[0073] According to an example of the present invention, the predicted moving direction may include two directions: the moving direction towards the center of the serving cell and the moving direction towards the edge of the serving cell.

[0074] The predicted moving direction can be determined by a prediction model. The obtaining device 522 may be integrated with this prediction model or obtain its prediction result from the prediction model.

[0075] The prediction model can predict the moving direction D of a UE at time t based on a set of consecutive signal-related measurement information {M t , M t-1 , M t-2 ,...} of the UE in the current time window {t, t - 1, t - 2...}. t .

[0076] According to an example of the present invention, the obtaining device 522 may only call the prediction model to determine the predicted moving direction D of the candidate UE relative to the current cell at the current time t when the current cell is overloaded. t .

[0077] According to another example of the present invention, the base station is further configured with a UE movement status database, and the latest predicted moving direction of each UE in the current cell is stored and updated in the UE movement status database. The prediction model continuously predicts the moving direction of each UE in the cell relative to the current cell, and the latest predicted moving direction is updated to the UE movement status database. The obtaining device 522 may read the latest predicted moving direction of the candidate UE from the UE movement status database.

[0078] Furthermore, the base station may further include a collection device ( Figure 5(not shown). The collection device continuously collects the signal-related measurement information of each UE in the current cell at the current moment. By continuously collecting the signal-related measurement information of the UE at each moment through the collection device, the acquisition device 522 can always obtain a set of UE signal-related measurement information of each UE in a current time window, and thus can call the prediction model to predict the moving direction of the candidate UE in real time or continuously predict the latest moving direction of each UE.

[0079] The screening device 523 determines the target UE from the candidate UEs according to the predicted moving direction of the candidate UE relative to the current cell.

[0080] According to an example of the present invention, the screening device 523 traverses all candidate UEs, removes the candidate UEs with the predicted moving direction towards the center of the current cell, and marks the candidate UEs with the predicted moving direction towards the edge of the current cell as the target UEs, so as to obtain the target UEs to be handed over to other cells.

[0081] Subsequently, the base station may further include a handover device ( Figure 5 (not shown) to hand over the target UE to other cells to be handed over to. That is, the UE handover process between the current cell and the target cell is triggered.

[0082] Referring to FIG. 6(a), which specifically shows a schematic diagram of the device of the base station for mobile load balancing in the RAN according to an embodiment of the present invention.

[0083] As shown in FIG. 6(a), the collection device 611 continuously collects the signal-related measurement information of each UE in the current cell at the current moment; the prediction device 612 continuously determines the predicted moving direction of each UE relative to the current cell according to a set of continuous signal-related measurement information of each UE in the current time window, and updates the latest predicted moving method of each determined UE to the UE movement status database 613.

[0084] Among them, the data proxy 614 and the control proxy 615 are embedded in the base station to assist in data collection and forwarding the feedback of the prediction service.

[0085] The collection device 611 receives the measurement information of all UEs in each cell, such as signal-related measurement information such as CQI, RSSI, RSRP, and RSPQ. Each cell regularly reports the measurement results of the UEs to the collection device 611 through the data proxy 614. The collection device 611 preprocesses these data and stores the formatted data in a certain type of storage, such as the UE movement status database 613.

[0086] The prediction device 612 can be a prediction model, which predicts the latest moving direction D of a UE relative to its serving cell based on a set of consecutive signal-related measurement information {M t ,M t-1 ,M t-2 ,…} of each UE in each cell within the current time window t 。

[0087] The prediction device 612 updates the latest predicted moving direction of all UEs in each cell to the UE mobility status database 613 through the control agent 615

[0088] The base station simultaneously monitors the load balancing of the cell

[0089] When the current cell is overloaded, the selection device 621 determines the candidate UEs in this cell to be handed over to other cells; the acquisition device 622 obtains the latest predicted moving direction of the candidate UEs relative to the current cell from the UE mobility status database 613; the screening device 623 determines the target UEs from the candidate UEs according to the predicted moving directions of the candidate UEs

[0090] Subsequently, the handover device 624 hands over the target UEs to other cells to be handed over to. That is, the UE handover process between the current cell and the target cell is triggered

[0091] Refer to FIG. 6(b), which specifically shows a schematic diagram of the device of a system for mobile load balancing in the RAN according to an embodiment of the present invention. Among them, the system includes a base station and a Radio Access Network Intelligent Controller (RIC), and the RIC includes a collection device 611 and a prediction device 612

[0092] Since the RIC is a new network element proposed in ORAN, for traditional RAN networks without this network element, if there are sufficient computing resources, the collection device 512 and the prediction device 511 can be deployed to the base station as shown in FIG. 6(a)

[0093] As shown in FIG. 6(b), the collection device 611 continuously collects the signal-related measurement information of each UE in each cell at the current moment from each base station; the prediction device 612 continuously determines the predicted moving direction of each UE relative to the current cell according to a set of consecutive signal-related measurement information of each UE in each cell within the current time window, and feeds back the latest predicted moving method of each UE in each cell determined to the corresponding base station

[0094] Among them, the data agent 614 and the control agent 615 are embedded in the base station to assist in data collection and forwarding the feedback of the prediction service

[0095] The collection device 611 continuously collects UE measurement metrics from each base station, such as signal-related measurement information like CQI, RSSI, RSRP, RSPQ, etc. Each cell periodically reports the measurement results of the UE to the collection device 611 through the data proxy 514.

[0096] The collection device 611 receives all UE measurement information, preprocesses this data and stores the formatted data in a certain type of storage, such as the UE mobility status database 513.

[0097] The prediction device 612 can be a prediction model, which predicts the latest moving direction D of the UE relative to its serving cell according to a set of consecutive signal-related measurement information {M t , M t-1 , M t-2 , …} of each UE in each cell within the current time window. t .

[0098] The prediction device 612 feeds back the latest predicted moving direction of all UEs in each cell to the corresponding base station through the control agent 615. Further, the latest predicted moving direction of each UE in the cell returned by the prediction device 612 is updated to the UE mobility status database 613 of the base station.

[0099] The base station simultaneously monitors the load balancing of the cell:

[0100] When the current cell is overloaded, the selection device 621 determines the candidate UEs in this cell to be switched to other cells; the acquisition device 622 obtains the latest predicted moving direction of the candidate UEs relative to the current cell from the UE mobility status database 613; the screening device 623 determines the target UEs from the candidate UEs according to the predicted moving directions of the candidate UEs.

[0101] Subsequently, the handover device 624 hands over the target UEs to the other cells to be switched to. That is, the UE handover process between the current cell and the target cell is triggered.

[0102] It should be noted that the present invention can be implemented in software and / or a combination of software and hardware. For example, it can be implemented using an application specific integrated circuit (ASIC), a general purpose computer or any other similar hardware device. In one embodiment, the software program of the present invention can be executed by a processor to implement the above-mentioned steps or functions. Similarly, the software program (including related data structures) of the present invention can be stored in a computer-readable recording medium, such as a RAM memory, a magnetic or optical drive or a floppy disk and similar devices. In addition, some steps or functions of the present invention can be implemented using hardware, for example, as a circuit cooperating with the processor to execute each step or function.

[0103] In addition, at least a part of the present invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computing device, can invoke or provide the methods and / or technical solutions according to the present invention through the operations of the computing device. The program instructions for invoking / providing the methods of the present invention may be stored in a fixed or removable recording medium, and / or transmitted through a data stream in a broadcast or other signal-bearing medium, and / or stored in the working memory of a computing device operating according to the program instructions.

[0104] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above-described exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be encompassed within the present invention. Any reference signs in the claims should not be construed as limiting the claims involved. In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. The multiple units or devices stated in the system claims can also be implemented by one unit or device through software or hardware. First, second, etc. are used to denote names and do not denote any particular order.

Claims

1. A method for UE selection for mobile load balancing in RAN, wherein, At the base station, the method includes the following steps: When the current cell is overloaded, determine candidate UEs in the cell to be handed over to other cells, where the candidate UEs are determined according to the A4 measurement reports of each UE in the cell; Obtain the latest predicted moving direction of the candidate UEs relative to the current cell from the UE mobility status database, where the predicted moving direction is determined by a prediction model based on a set of consecutive signal-related measurement information of the candidate UEs in the current time window; Determine the target UE from the candidate UEs according to the predicted moving direction; Wherein, the method further includes the following steps: Continuously determine the latest predicted moving direction of each UE relative to the current cell by the prediction model according to a set of consecutive signal-related measurement information of each UE in the current time window; Update the latest determined predicted moving direction of each UE relative to the current cell to the UE mobility status database.

2. The method according to claim 1, wherein The prediction model is obtained by training a classification neural network with labeled samples, where each sample is labeled with a set of consecutive signal-related measurement information of a UE in a predetermined time window and the moving direction of the UE relative to its serving cell.

3. The method according to claim 1, wherein The method further includes the following steps: Continuously collect the signal-related measurement information of each UE in the current cell at the current moment.

4. The method according to claim 1, wherein, The prediction model is arranged in the RAN intelligent controller; Wherein, the method further includes: Continuously transmit the signal-related measurement information of each UE in the current cell at the current moment to the RAN intelligent controller; Wherein, the UE mobility status database is updated according to the predicted moving direction of each UE relative to the current cell, which is newly determined by the prediction model and continuously fed back by the RAN intelligent controller.

5. The method according to claim 1, wherein, The predicted moving direction includes two directions: the moving direction towards the center of the current cell and the moving direction towards the edge of the current cell.

6. The method according to claim 1, wherein The method further includes the following steps: Hand over the target UE to the other cell.

7. A base station for UE selection for mobile load balancing in the RAN, wherein, The base station includes: A selection device for determining candidate UEs in the cell to be handed over to other cells when its cell is overloaded, where the candidate UEs are determined according to the A4 measurement reports of each UE in the cell; An acquisition device for obtaining the latest predicted moving direction of the candidate UEs relative to the current cell from the UE mobility status database, where the predicted moving direction is determined by a prediction model based on a set of consecutive signal-related measurement information of the candidate UEs in the current time window; A screening device for determining the target UE from the candidate UEs according to the predicted moving direction; A prediction device for continuously determining the latest predicted moving direction of each UE relative to the current cell by the prediction model according to a set of consecutive signal-related measurement information of each UE in the current time window; and updating the latest determined predicted moving direction of each UE relative to the current cell to the UE mobility status database.

8. The base station according to claim 7, wherein, The base station further includes: A handover module for handing over the target UE to the other cell.

Images