Apparatus, method, and computer program for load prediction

By using machine learning models in a wireless communication system to share the measured values ​​and occasional load indications of user equipment between source nodes and target nodes, predicting the load after user equipment switches, solving the problem of inaccurate prediction of occasional load in the prior art. , improve load balancing and mobility optimization capabilities.

CN119072951APending Publication Date: 2024-12-03NOKIA TECHNOLOGIES OY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202380034535.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2022-04-26
Filing Date
2023-04-14
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

Existing wireless communication systems have inaccuracies in predicting the occasional load when switching user equipment, especially in a short period of time, which affects load balancing and mobility optimization.

Method used

Using machine learning models to collect and share measurements and occasional load indications of user equipment between source nodes and target nodes, train the model to predict the occasional load after user equipment switches, and use these prediction results in mobility decisions to optimize users The switching process of the device.

Benefits of technology

Improve the prediction accuracy of occasional loads caused by user equipment switching, enhance load balancing and mobility optimization capabilities, and reduce signaling complexity and operating costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119072951A_ABST
    Figure CN119072951A_ABST
Patent Text Reader

Abstract

There is provided an apparatus comprising means for: receiving, at a source node, measurements from a plurality of user equipments; providing, from the source node to a respective one of the plurality of target nodes, a handover request for each of the plurality of user equipments, the handover request comprising a request to measure an accidental load; determining to execute switching from the subsequent user equipment to the subsequent target node; determining a predicted accidental load based on the received measurement value and the measured accidental load; using the predicted accidental load in a mobility decision; and providing the predicted accidental load to a subsequent target node in a handover request for a subsequent user equipment.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to a method, apparatus, system, and computer program, particularly but not limited to machine learning driven episodic load prediction prior to a mobile event. Background Art

[0002] A communication system may be considered as a facility that enables communication sessions between two or more entities (such as user terminals, base stations, and / or other nodes) by providing a bearer between the various entities involved in the communication path. A communication system may be provided, for example, by means of a communication network and one or more compatible communication devices. A communication session may include, for example, data communications for bearer communications, such as voice, video, electronic mail (email), text messaging, multimedia, and / or content data. Non-limiting examples of the services provided include two-way or multi-way calls, data communications or multimedia services, and access to a data network system (such as the Internet).

[0003] In a wireless communication system, at least a portion of a communication session between at least two stations occurs over a wireless link. Examples of wireless systems include public land mobile networks (PLMNs), satellite-based communication systems, and different wireless local networks, such as wireless local area networks (WLANs). Some wireless systems can be divided into cells, and are therefore often referred to as cellular systems.

[0004] A user may access a communication system by means of an appropriate communication device or terminal. The user's communication device may be referred to as a user equipment (UE) or user device. The communication device is provided with appropriate signal receiving and sending means for enabling communication, such as enabling access to a communication network or direct communication with other users. The communication device may access a carrier provided by a station (e.g., a base station of a cell) and send and / or receive communications on the carrier.

[0005] A communication system and associated equipment typically operate according to a given standard or specification that specifies what the various entities associated with the system are permitted to do and how this should be achieved. The communication protocols and / or parameters used for the connection are also typically defined. An example of a communication system is UTRAN (3G radio). Other examples of communication systems include the Long Term Evolution (LTE) of the Universal Mobile Telecommunications System (UMTS) radio access technology and so-called 5G or New Radio (NR) networks. NR is being standardized by the Third Generation Partnership Project (3GPP). Summary of the invention

[0006] In a first aspect, a device is provided, comprising components for: receiving measurement values ​​from a plurality of user devices at a source node; providing a switching request for each of the plurality of user devices from the source node to a corresponding one of a plurality of target nodes, the switching request comprising a request to measure an occasional load; determining to perform a switching of a subsequent user device to a subsequent target node; determining a predicted occasional load based on the received measurement values ​​and the measured occasional load; using the predicted occasional load in a mobility decision; and providing the predicted occasional load to the subsequent target node in the switching request for the subsequent user device.

[0007] The component for determining the predicted sporadic load may include at least one machine learning model, which, when executed, is configured to: determine the predicted sporadic load based on measurements of subsequent user equipment.

[0008] The machine learning model, when executed, is configured to determine a predicted episodic load based on an identity of a target node.

[0009] The apparatus may include means for providing the received measurements to a host node for use in training a machine learning model.

[0010] The apparatus may include components for receiving an indication of measured sporadic load from each of a plurality of target nodes and providing the indication of measured sporadic load to a host node for use in training a machine learning model.

[0011] The apparatus may include means for receiving, from each of the plurality of destination nodes, an indication of a measured capacity at an occasional load.

[0012] The sporadic load may include an absolute amount of resources allocated to the user equipment in at least one of the downlink and uplink, or a ratio of resources allocated to the user equipment in at least one of the uplink and downlink to all resources scheduled in a given time period.

[0013] The apparatus may include means for receiving an indication of a measured episodic load from each of a plurality of target nodes for use in training a machine learning model at a source node.

[0014] The apparatus may include means for determining a predicted contingent load further based on an identifier of at least one of the subsequent target node and the source node.

[0015] In a second aspect, an apparatus is provided, comprising components for: receiving, at a target node, a handover request for a user device from a source node, wherein either the handover request includes a request to measure an occasional load, and if so, providing an indication of the measured occasional load to at least one of the source node and the host node for use in training a machine learning model, or the handover request includes an indication of a predicted occasional load associated with the user device.

[0016] In a third aspect, a device is provided, comprising components for receiving, from a source node and one of a plurality of target nodes, at least one indication of an occasional load measured for a user device at a corresponding one of a plurality of target nodes, and a measurement value for the user device, and using at least one indication and the measurement value when training a machine learning model, which, when executed, is configured to determine a predicted occasional load based on subsequent measurement values ​​of the user device.

[0017] In a fourth aspect, a method is provided, comprising: receiving measurement values ​​from a plurality of user equipments at a source node, providing a switching request for each of the plurality of user equipments from the source node to a corresponding one of a plurality of target nodes, the switching request comprising a request to measure an occasional load, determining to perform a switching of a subsequent user equipment to a subsequent target node, determining a predicted occasional load based on the received measurement values ​​and the measured occasional load, using the predicted occasional load in a mobility decision, and providing the predicted occasional load to the subsequent target node in the switching request for the subsequent user equipment.

[0018] Determining the predicted sporadic load may include at least one machine learning model, which, when executed, is configured to: determine the predicted sporadic load based on measurements of subsequent user devices.

[0019] The machine learning model, when executed, is configured to determine a predicted episodic load based on an identity of a target node.

[0020] The method may include providing the received measurements to a host node for use in training a machine learning model.

[0021] The method may include receiving an indication of a measured sporadic load from each of a plurality of target nodes, and providing the indication of the measured sporadic load to a host node for use in training a machine learning model.

[0022] The method may include receiving, from each of the plurality of target nodes, an indication of a measured capacity at an occasional load.

[0023] The sporadic load may include an absolute amount of resources allocated to the user equipment in at least one of the downlink and uplink, or a ratio of resources allocated to the user equipment in at least one of the uplink and downlink to all resources scheduled in a given time period.

[0024] The method may include receiving an indication of a measured episodic load from each of a plurality of target nodes for use in training a machine learning model at a source node.

[0025] The method may include determining a predicted episodic load further based on an identifier of at least one of the subsequent target node and the source node.

[0026] In a fifth aspect, a method is provided, comprising: at a target node, receiving a handover request for a user device from a source node, wherein either the handover request includes a request to measure an occasional load, and if so, providing an indication of the measured occasional load to at least one of the source node and the host node for use in training a machine learning model, or the handover request includes an indication of a predicted occasional load associated with the user device.

[0027] In a sixth aspect, a method is provided, comprising: receiving from a source node and one of a plurality of target nodes at least one indication of an occasional load measured for a user device at a corresponding one of a plurality of target nodes, and a measurement value for the user device, and using at least one indication and the measurement value when training a machine learning model, which machine learning model is configured when executed to: determine a predicted occasional load based on subsequent measurement values ​​of the user device.

[0028] In a seventh aspect, a device is provided, comprising: at least one processor; and at least one memory comprising computer program code, wherein the at least one memory and the computer program code are configured to, together with the at least one processor, enable the device to at least: receive measurement values ​​from a plurality of user devices at a source node, provide a switching request for each of the plurality of user devices from the source node to a corresponding one of the plurality of target nodes, the switching request comprising a request to measure an occasional load, determine to perform a switching of a subsequent user device to a subsequent target node, determine a predicted occasional load based on the received measurement values ​​and the measured occasional load, use the predicted occasional load in a mobility decision, and provide the predicted occasional load to the subsequent target node in the switching request for the subsequent user device.

[0029] The component for determining the predicted sporadic load may include at least one machine learning model, which, when executed, is configured to: determine the predicted sporadic load based on measurements of subsequent user equipment.

[0030] The machine learning model, when executed, is configured to determine a predicted episodic load based on an identity of a target node.

[0031] The apparatus may be configured to provide the received measurements to a host node for use in training a machine learning model.

[0032] The apparatus may be configured to receive an indication of measured sporadic load from each of a plurality of target nodes and provide the indication of measured sporadic load to a host node for use in training a machine learning model.

[0033] The apparatus may be configured to receive, from each of the plurality of target nodes, an indication of a measured capacity at an occasional load.

[0034] The sporadic load may include an absolute amount of resources allocated to the user equipment in at least one of the downlink and uplink, or a ratio of resources allocated to the user equipment in at least one of the uplink and downlink to all resources scheduled in a given time period.

[0035] The apparatus may be configured to receive an indication of measured episodic load from each of a plurality of target nodes for use in training a machine learning model at a source node.

[0036] The apparatus may be configured to determine the predicted contingent load further based on an identifier of at least one of the subsequent target node and the source node.

[0037] In an eighth aspect, there is provided an apparatus comprising: at least one processor; at least one memory comprising computer program code, the at least one memory and the computer program code being configured to, together with the at least one processor, cause the apparatus to at least: receive, at a target node, a switching request for a user device from a source node, wherein either the switching request comprises a request to measure an occasional load, and if so, provide an indication of the measured occasional load to at least one of the source node and the host node for use in training a machine learning model, or the switching request comprises an indication of a predicted occasional load associated with the user device.

[0038] In a ninth aspect, a device is provided, comprising: at least one processor; and at least one memory comprising computer program code, wherein the at least one memory and the computer program code are configured to, together with the at least one processor, enable the device to at least: receive from a source node and one of a plurality of target nodes at least one indication of an occasional load measured for a user device at a corresponding one of a plurality of target nodes, and a measurement value for the user device, and use at least one indication and the measurement value when training a machine learning model, wherein the machine learning model is configured, when executed, to: determine a predicted occasional load based on the measurement values ​​of subsequent user devices.

[0039] In a tenth aspect, a computer-readable medium comprising program instructions is provided, the program instructions being used to cause an apparatus to perform at least the following operations: receiving measurement values ​​from a plurality of user devices at a source node; providing a switching request for each of the plurality of user devices from the source node to a corresponding one of the plurality of target nodes, the switching request comprising a request to measure an occasional load; determining to perform a switching of a subsequent user device to a subsequent target node; determining a predicted occasional load based on the received measurement values ​​and the measured occasional load; using the predicted occasional load in a mobility decision; and providing the predicted occasional load to the subsequent target node in a switching request for the subsequent user device.

[0040] Determining the predicted sporadic load may include at least one machine learning model, which, when executed, is configured to: determine the predicted sporadic load based on measurements of subsequent user devices.

[0041] The machine learning model, when executed, is configured to determine a predicted episodic load based on an identity of a target node.

[0042] The apparatus may be caused to perform: providing the received measurements to a host node for use in training a machine learning model.

[0043] The apparatus may be caused to perform: receiving an indication of measured sporadic load from each of a plurality of target nodes, and providing the indication of measured sporadic load to a host node for use in training a machine learning model.

[0044] The apparatus may be caused to perform: receiving, from each of a plurality of target nodes, an indication of a capacity at a measured occasional load.

[0045] The sporadic load may include an absolute amount of resources allocated to the user equipment in at least one of the downlink and uplink, or a ratio of resources allocated to the user equipment in at least one of the uplink and downlink to all resources scheduled in a given time period.

[0046] The apparatus may be caused to perform: receiving an indication of measured episodic load from each of a plurality of target nodes for use in training a machine learning model at a source node.

[0047] The apparatus may be caused to perform: determining a predicted contingent load further based on an identifier of at least one of the subsequent target node and the source node.

[0048] In an eleventh aspect, a computer-readable medium comprising program instructions is provided, the program instructions being used to cause an apparatus to perform at least the following operations: at a target node, receiving a switching request for a user device from a source node, wherein either the switching request includes a request to measure an occasional load, and if so, providing an indication of the measured occasional load to at least one of the source node and the host node for use in training a machine learning model, or the switching request includes an indication of a predicted occasional load associated with the user device.

[0049] In a twelfth aspect, a computer-readable medium comprising program instructions is provided, the program instructions being used to cause an apparatus to perform at least the following operations: receiving from a source node and one of a plurality of target nodes at least one indication of an occasional load measured for a user device at a corresponding one of a plurality of target nodes, and a measurement value for the user device, and using at least one indication and the measurement value when training a machine learning model, the machine learning model being configured, when executed, to determine a predicted occasional load based on subsequent measurement values ​​of the user device.

[0050] In a thirteenth aspect, a non-transitory computer-readable medium is provided, comprising program instructions for causing an apparatus to execute a method according to at least the third or fourth aspect.

[0051] In the above, many different embodiments have been described. It should be understood that further embodiments can be provided by combining any two or more of the embodiments described above. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Embodiments will now be described, by way of example only, with reference to the accompanying drawings, in which:

[0053] Figure 1 A schematic diagram of an example 5GS communication system is shown;

[0054] Figure 2 A schematic diagram of an example mobile communication device is shown;

[0055] Figure 3 A schematic diagram of an example control device is shown;

[0056] Figure 4 shows a high-level signaling flow for a load balancing use case;

[0057] Figure 5 A schematic diagram of a predictive mobility concept based on machine learning is shown;

[0058] Figure 6 shows the signaling flow of resource status prediction assisted mobility;

[0059] Figure 7 A flow chart showing a method according to an example embodiment is shown;

[0060] Figure 8 shows a high level signaling flow according to an example embodiment;

[0061] Fig. 9 shows a high level signaling flow according to an example embodiment;

[0062] Fig.10 A schematic diagram of a machine learning model according to an example embodiment is shown;

[0063] Fig.11 shows a signaling flow for training a machine learning model according to an example embodiment;

[0064] Fig.12 shows a signaling flow for training a machine learning model according to an example embodiment;

[0065] Fig.13 shows a signaling flow for occasional load prediction for baseline switching according to an example embodiment;

[0066] Fig.14 A signaling flow for occasional load prediction for conditional switching according to an example embodiment is shown. DETAILED DESCRIPTION

[0067] Before explaining the examples in detail, refer to Figures 1 to 3 Certain general principles of wireless communication systems and mobile communication devices are briefly explained to aid the understanding of the technology on which the described examples are based.

[0068] An example of a suitable communication system is a 5G system (5GS). The network architecture in 5GS may be similar to that of Advanced LTE. The base station of the NR system may be referred to as a next-generation node B (gNB). Changes to the network architecture may depend on the demand for various radio technologies and more refined QoS support, as well as some on-demand requirements, such as QoS levels that support QoE from the user's perspective. In addition, network-aware services and applications, as well as service- and application-aware networks, can bring changes to the architecture. These are related to information-centric networks (ICNs) and user-centric content delivery networks (UC-CDN) approaches. NR can use multiple-input multiple-output (MIMO) antennas, more base stations or nodes than LTE (the so-called small cell concept), including macro sites that operate in cooperation with small sites, and can also adopt various radio technologies to achieve better coverage and higher data rates.

[0069] 5G networks may utilize network function virtualization (NFV), a network architecture concept that proposes virtualizing network node functions into "building blocks" or entities that can be operably connected or linked together to provide services. A virtualized network function (VNF) may include one or more virtual machines that run computer program code using standard or common types of servers rather than customized hardware. Cloud computing or data storage may also be used. In radio communications, this may mean that node operations are performed at least in part in a server, host, or node that is operably coupled to a remote radio head. Node operations may also be distributed among multiple servers, nodes, or hosts. It should also be understood that the distribution of labor between core network operations and base station operations may be different from LTE or even non-existent.

[0070] Figure 1 A schematic diagram of a 5G system (5GS) 100 is shown. The 5GS may include a user equipment (UE) 102 (also referred to as a communication device or terminal), a 5G radio access network (5GRAN) 104, a 5G core network (5GCN) 106, one or more application functions (AFs) 108, and one or more data networks (DNs) 110.

[0071] An example 5G core network (CN) includes functional entities. The 5GCN 106 may include: one or more access and mobility management functions (AMFs) 112, one or more session management functions (SMFs) 114, an authentication server function (AUSF) 116, a unified data management (UDM) 118, one or more user plane functions (UPFs) 120, a unified data repository (UDR) 122, and / or a network exposure function (NEF) 124. The UPF is controlled by the SMF (session management function) which receives policies from the PCF (policy control function).

[0072] The CN is connected to the terminal device via a radio access network (RAN). The 5G RAN may include one or more gNodeB (GNB) distributed unit functions connected to one or more gNodeB (GNB) centralized unit functions. The RAN may include one or more access nodes.

[0073] The role of UPF (User Plane Function) is called PSA (Protocol Data Unit (PDU) Session Anchor), which is responsible for forwarding frames back and forth between the DN (Data Network) and the tunnels that are established over 5G towards (multiple) UEs that exchange traffic with the DN.

[0074] Now refer to Figure 2 describing in more detail possible mobile communication devices, Figure 2 A schematic partial cross-sectional view of a communication device 200 is shown. Such a communication device is generally referred to as a user equipment (UE) or terminal. Any device capable of sending and receiving radio signals can provide a suitable mobile communication device. Non-limiting examples include a mobile station (MS) or mobile device, such as a mobile phone or so-called "smart phone", a computer provided with a wireless interface card or other wireless interface facilities (e.g., a USB dongle), a personal data assistant (PDA) or a tablet computer provided with wireless communication capabilities, an IP voice (VoIP) phone, a portable computer, a desktop computer, an image capture terminal device, such as a digital camera, a game terminal device, a music storage and playback device, a vehicle-mounted wireless terminal device, a wireless endpoint, a mobile station, a laptop embedded device (LEE), a laptop mounted device (LME), an intelligent device, a wireless customer premises equipment (CPE), or any combination of these or similar devices. The mobile communication device can provide, for example, data communications for carrying communications such as voice, email (email), text messages, multimedia, etc. Therefore, a variety of services can be provided to users via their communication devices. Non-limiting examples of these services include two-way or multi-way calls, data communications or multimedia services, or simply access to a data communications network system such as the Internet. Broadcast or multicast data may also be provided to the user. Non-limiting examples of content include downloads, television and radio programs, videos, advertisements, various alerts, and other information.

[0075] The mobile device is typically provided with at least one data processing entity 201, at least one memory 202, and other possible components 203 for use in the software and hardware-assisted execution of the tasks it is designed to perform, including controlling access to and communication with access systems and other communication devices. Data processing, storage and other related control devices may be provided on appropriate circuit boards and / or in chipsets. This feature is indicated by reference numeral 204. The user can control the operation of the mobile device by means of a suitable user interface (such as a keyboard 205, voice commands, a touch screen or touchpad, a combination thereof, etc.). A display 208, a speaker, and a microphone may also be provided. In addition, the mobile communication device may include appropriate connectors (wired or wireless) to connect to other devices and / or for connecting external accessories (such as hands-free devices) thereto.

[0076] The mobile device 200 may receive signals over the air or radio interface 207 via appropriate receiving means, and may transmit signals via appropriate means for transmitting radio signals. Figure 2 In the embodiment of the present invention, the transceiver arrangement is schematically designated by block 206. The transceiver arrangement 206 may be provided, for example, by means of a radio part and an associated antenna arrangement. The antenna arrangement may be arranged internally or externally to the mobile device.

[0077] Figure 3 An example of a control device 300 for a communication system is shown, for example, the control device 300 is coupled to a station of the access system, and / or a station for controlling the access system, such as a RAN node, such as a base station, an eNB or a gNB, a relay node or a core network node (such as an MME or an S-GW or a P-GW), or a core network function (such as an AMF / SMF), or a server or a host. The method can be implemented in a single control device or across multiple control devices. The control device can be integrated with a node or module of a core network or RAN / or external to a node or module of a core network or RAN. In some embodiments, the base station includes a separate control device unit or module. In other embodiments, the control device can be another network element (such as a radio network controller or a spectrum controller). In some embodiments, each base station can have such a control device and a control device provided in a radio network controller. The control device 300 can be arranged to provide control of communications in a service area of ​​the system. The control device 300 includes at least one memory 301, at least one data processing unit 302, 303, and an input / output interface 304. Via the interface, the control device may be coupled to a receiver and a transmitter of the base station.The receiver and / or the transmitter may be implemented as a radio front end or a remote radio head.

[0078] The following relates to the enabling and functional framework of AI in the RAN, including the inputs and outputs required for AI functionality and ML algorithms, and the impact of standardization in the network interfaces or at nodes in the existing architecture to communicate this input / output data through them.

[0079] gNB load prediction is an enabler for many other use cases that can benefit from accurate load prediction, including mobility optimization, load balancing, and energy conservation.

[0080] Load balancing is an example of a use case where predictive load information can improve performance by providing higher QoS and enhanced system performance. Figure 4 An example high-level signaling flow for AI / ML use cases related to load balancing is shown in FIG.

[0081] exist Figure 4 In step 0, the AI / ML model training is located at NG-RAN node 1. It is assumed that NG-RAN node 2 is capable of providing useful input information to NG-RAN node 1, including predicted resource status and / or mobility prediction.

[0082] In steps 1-3, the NG-RAN node 1 configures and obtains UE measurement and location information (e.g., RRM measurements, MDT measurements, speed, location).

[0083] In step 4, NG-RAN node 1 receives input information from neighboring NG-RAN node 2 for load balancing model training.

[0084] In step 5, load balancing model training occurs at NG-RAN Node 1.

[0085] In step 6, the NG-RAN node 1 receives inputs from the UE that will be used in the reasoning phase, such as UE measurements and location information.

[0086] In step 7, NG-RAN node 1 receives input data from neighboring NG-RAN node 2 for load balancing reasoning.

[0087] In step 8, the NG-RAN node 1 performs mobility load balancing prediction (e.g., for the cells of the NG-RAN node 1).

[0088] In step 9, NG-RAN node 1 makes a mobility load balancing decision / action based on the mobility load balancing prediction and moves the UE from NG-RAN node 1 to NG-RAN node 2.

[0089] In step 10, the NG-RAN node 2 sends feedback information (eg, resource status update after load balancing) to the NG-RAN node 1. The feedback information may be signaled after receiving the feedback request.

[0090] To improve mobility decisions, the gNB can request mobility feedback from neighboring nodes. The source NG-RAN node can be provided with predicted resource status and performance information for candidate target NG-RAN nodes.

[0091] The Resource Status Report Initiation procedure may be used by an NG-RAN node to request the reporting of load measurements to another NG-RAN node. For example, NG-RAN node 1 initiates the procedure by sending a Resource Status Request message to NG-RAN node 2 to start measurements, stop measurements, or add cells for which measurements are to be reported. When starting measurements, the Reporting Characteristics IE in the Resource Status Request indicates the type of objects for which the NG-RAN node 2 should perform measurements. If requested, the Radio Resource Status IE may be included as part of the Resource Status Response (RESOURCE STATUS RESPONSE), which indicates the usage of PRBs for each cell and each SSB area for all services in the downlink and uplink, and the usage of PDCCH CCEs for downlink and uplink scheduling.

[0092] In the concept of predictive mobility based on machine learning (ML), a machine learning model is trained to make mobility decisions based on a series of UE measurements over time.

[0093] Figure 5 A concept is illustrated that uses ML-based mobility prediction for determining the best target and time to trigger a handover based on UE radio measurements (RSRP, RSRQ, SINR) using a recurrent neural network (RNN). Training frames with time series measurements and labels are used as input to train the ML classification model. New frames including time series measurements are input to the model and the predicted probability of the target cell is given as output from the model. Another concept involves predicting when and where to trigger a Layer 3 handover based on the serving beam index reported by the UE. The results of the studies on these concepts indicate that machine learning-based sequence prediction has potential in optimizing mobility use cases.

[0094] The current concepts for gNB load prediction are performed locally in the gNB. The load variations caused by UE mobility (especially inbound mobility, such as UE switching from other cells to the gNB) can be statistically predicted, but not accurately for a short period of time per UE. This so-called sporadic load caused by inbound mobility may have a significant impact on the load of the gNB. The sporadic load caused by the UE on the radio interface depends on its traffic requirements and type (GBR, non-GBR) and QCI, as well as the radio conditions once the UE is switched to the target cell, which in turn depends on the UE trajectory and the radio environment conditions at the time. The sporadic load also causes load on the transport and backhaul interfaces (e.g., RAN-RAN and RAN-core network interfaces, and software and hardware processing components of the network nodes).

[0095] One challenge with predicting sporadic load before a handover occurs is that the data and knowledge to enable meaningful predictions is split between the source and target nodes. The UE is reporting its measurements to the source, and the source also knows the current traffic throughput and may have learned some of the traffic patterns of this UE. However, the target node knows which load values ​​(i.e., the load values ​​of which managed elements) are relevant to its operation and load prediction.

[0096] During the handover preparation process (assuming Rel-17 network implementation), the source node forwards the UE access stratum context (containing the UE's source RRC configuration, which includes the PDU sessions and DRBs assigned to the UE and the UE's radio access capabilities) to the target node. The target node performs admission control on this UE based on PDU sessions, QoS requirements, load in the target cell, UE measurements, etc., and ultimately derives a radio reconfiguration message (handover command) that is returned to the source node to be provided to the UE.

[0097] The entire resource allocation process only takes into account the instant in time that the UE is admitted to the target node and does not include information on various aspects that may help the target node make better or more reliable admission control decisions, such as considering the immediate and short-term / medium-term aspects (e.g., in terms of 10-100 milliseconds of traffic) of the (multiple) PDU sessions of the admitted UE.

[0098] Figure 6 The method in conditional handover (CHO) is illustrated, where the source node can provide the target node with the information required for the predicted occasional load. Such information can include the predicted trajectory of the UE and the predicted traffic demand. Using these inputs, the target node can estimate the additional occasional load caused by the UE after the handover.

[0099] As part of the CHO request message, the source node provides the target node with a resource utilization request information element, which contains the trajectory and traffic demand forecast required to estimate the occasional load. The target node can use the predicted occasional load in its admission control and can also provide it back to the source node in the resource utilization response in the CHO request acknowledgment.

[0100] The source node may also later use the resource utilization request information and the resource utilization request element in the mobility update request and response messages to trigger an update procedure to obtain an updated occasional load estimate from the target node.

[0101] In this approach, in order to obtain the sporadic load predictions, the source node needs to request them from the target node. The source node will have the predicted sporadic load only after having sent the CHO request. The source node cannot utilize the prediction to select the target candidate to prepare. Similarly, whenever the source node wants to obtain updates to the predicted sporadic load, it needs to request them from the target node and send the required request information in the resource utilization request information, which increases signaling.

[0102] Figure 6 The approach involves linking trajectories and business demand forecasts to the model to determine the occasional load induced. The challenge of linking ML models is that the cumulative inaccuracy of each model in the model and the inaccuracy of the linked ML model inputs can have a significant impact on model performance. It is usually better to build a predictive model from observed, measured input values ​​to predicted KPIs that is also trained as one entity rather than linking several independently trained models.

[0103] A method is needed to predict the occasional load directly at the source node.

[0104] Figure 7 A flow chart of a method according to an example embodiment is shown.

[0105] In S1 , the method comprises: receiving, at a source node, measurement values ​​from a plurality of user equipments.

[0106] In S2, the method comprises providing, from a source node to a corresponding one of a plurality of target nodes, a handover request for each of a plurality of user equipments, the handover request comprising a request to measure an occasional load.

[0107] In S3, the method includes: determining to perform a handover of a subsequent user equipment to a subsequent target node.

[0108] In S4, the method comprises determining a predicted occasional load based on the received measurement value and the measured occasional load.

[0109] In S5, the method comprises using the predicted episodic load in mobility decisions.

[0110] In S6, the method includes: providing the predicted incidental load to a subsequent target node in a handover request for a subsequent user equipment.

[0111] The component for determining the predicted sporadic load may include at least one machine learning model, which, when executed, is configured to: determine the predicted sporadic load based on measurements of subsequent user devices. The machine learning model, when executed, may be configured to: determine the predicted sporadic load based on the identity of the target node.

[0112] The method may provide a learning method (eg, based on supervised learning) to predict the occasional load on the target node caused by UE handover.

[0113] The sporadic load may include an absolute amount of resources allocated to the user equipment in at least one of the downlink and uplink, or a ratio of resources allocated to the user equipment in at least one of the uplink and downlink to all resources scheduled in a given time period.

[0114] The given time interval may be a configured time interval T after the UE has connected to the target cell and sent and / or received data over the air interface. 事件 .

[0115] Occasional load can be defined as the 事件 The absolute number of physical resource blocks (PRBs) allocated to the switching UE in the downlink (DL) and / or uplink (UL).

[0116] Alternatively or additionally, the occasional load may be defined as T 事件 The ratio of the PRBs allocated to the switched UE to all scheduled PRBs.

[0117] The definition of occasional load may be extended to cover other additional loads caused by the UE when it switches to a cell. The definition of occasional load may apply to different contexts, i.e., context attributes may include a combination of managed elements, where the load is aggregated: gNB (DU), cell, SSB area, beam, uplink / downlink, RAN slice, multiple connectivity and CA settings, transport connectivity (e.g., IP throughput to external nodes), and software / hardware resources (i.e., memory and CPU).

[0118] Occasional load may be affected by network congestion, in which case the UE may not be able to obtain its requested throughput and will therefore also incur an occasional load lower than the requested service. However, for load balancing and other similar use cases, predicting the load that will satisfy the UE's requirements may be more important. Therefore, the following content focuses on predicting the occasional load of the requested service. This should also be taken into account when collecting training data.

[0119] In order to learn to predict load rather than estimate it, the target node measures the occasional load.

[0120] Training for the machine learning model may be performed on a training host located at the RAN (e.g., at a source node), or a host node external to the RAN (e.g., an OAM).

[0121] The method may include providing the received measurements to a host node for use in training a machine learning model.

[0122] The method may include receiving an indication of a measured sporadic load from each of a plurality of target nodes, and providing the indication of the measured sporadic load to a host node for use in training a machine learning model.

[0123] Alternatively or additionally, the method may include receiving an indication of a measured episodic load from each of a plurality of target nodes for use in training a machine learning model at a source node.

[0124] The method may include receiving, from each of the plurality of target nodes, an indication of a capacity measured at an occasional load. For example, the target gNB may also report its (average) composite available capacity (CAC) when measuring the occasional load, which may be used to filter measurements affected by congestion or to train the effect of target node congestion into a predictive model.

[0125] The measurements may be L1 beam measurements, or L3 cell quality measurements. The measurements may include: a time series of beam indices of past serving beams, a time series of indices of past serving cells, a time series of UE radio measurements (e.g., RSRP, RSRQ, SINR), a time series of UE locations (expressed in coordinates), a list of carriers owned by the UE and their priorities (e.g., their QoS Class Identifiers (QCIs)), a time series of past traffic demands for each carrier, and (slice-specific) traffic demands for each PDU session (and corresponding DRB).

[0126] The occasional load measurements may be used to train an ML model to predict the occasional load based on the KPIs measured and reported by the UE to the source node. The input features for the prediction model depend on the implementation and on which managed object the load is predicted, and may include one or more of the following: for example source and target node IDs - if the same ML model instance covers several source and / or target nodes and / or UE measurements. The model may cover several potential mobility source nodes and target nodes, in which case identifiers of these nodes may be provided as input to the model. The identifier of the node may be a physical cell ID (PCI) or an NR-CGI (cell global ID). That is, the method may include determining the predicted occasional load also based on an identifier of at least one of the subsequent target nodes and the source node.

[0127] Using this approach, the sporadic load can be predicted directly from the reported measurements without the need for intermediate predictions of UE trajectories and traffic demand.

[0128] Figure 8 A signaling flow according to an example embodiment is shown.

[0129] In step 1, the source gNB configures the UE to perform measurements for training data collection.

[0130] In step 2, the UE reports measurements, such as L1 beam measurements, or L3 cell quality measurements. In the case of a split architecture, L1 beam measurements including neighboring cells will be reported to the gNB-DU, and L3 cell quality measurements will be reported to the gNB-CU.

[0131] In step 3, measurements associated with the UE context are stored by the source gNB.

[0132] In step 4, handover to a neighboring cell is prepared. In the handover request, the source gNB may request the target gNB to measure the occasional load caused by the UE being handed over.

[0133] In step 5, the handover is performed.

[0134] In step 6, the target gNB measures T 事件 occasional load on the

[0135] In step 7, the target gNB sends feedback information to the source gNB by reporting the measured occasional load back to the source gNB. The target gNB may also report its (average) composite available capacity (CAC) at the time of measurement, which can be used to filter measurements affected by congestion, or to train the impact of target node congestion into a prediction model.

[0136] In step 8, the source gNB stores and appends the measured load to the collected input data.

[0137] Once the source gNB has collected the configured number of training samples, it reports the training dataset to the training host in step 9.

[0138] The training host trains the ML using the provided training dataset in step 10 .

[0139] The trained model is deployed back to the source gNB in ​​step 11. In this example, the training host is located outside the source gNB. In an alternative embodiment, the source gNB may include a training host.

[0140] In step 12, the source gNB configures another UE for inferential load prediction using events that triggered data collection. These could be mobility events, such as A3 events.

[0141] In step 13, the UE reports the configured measurements.

[0142] In step 14, if the UE is handed over to the target gNB, the source gNB uses the reported measurements and the trained ML model to predict the occasional load.

[0143] In step 15, the source gNB may exploit the predicted contingency, e.g. in order to optimize load balancing in its mobility decisions.

[0144] In step 16, a handover is requested to the target gNB. In the handover request, the source gNB includes the predicted occasional load.

[0145] In step 17, the target gNB may use the predicted occasional load, for example in its admission control.

[0146] Fig. 9 A signaling flow according to an alternative example embodiment is shown. In this embodiment, the target gNB reports the measured occasional load directly to the training master.

[0147] After executing HO, if Figure 8 As shown in , the target gNB reports the measured occasional load and (optionally) its average CAC directly to the training master in step 7.

[0148] The source gNB only reports the collected input feature data to the training host in step 8.

[0149] The training master correlates the input data with the corresponding occasional load measurement(s) in step 9. The correlation may require the generation of a "universal token" to identify a specific UE and handover of that UE.

[0150] Fig.10An example machine learning model is shown. In this example, the sporadic load is predicted based on a time series of past service beam indices, service quality level indices, and service demands (for all carriers). As an additional input, the ID of the target node for which the sporadic load is predicted is given. The model uses a long short-term memory (LSTM) recurrent neural network (RNN), which excels at making predictions from data sequences. In the presented model, there are two LSTM layers, one that is unaware of the target node ID and the other that is aware of the target node ID.

[0151] Fig.11 A signaling diagram for training a sporadic load prediction model according to an example embodiment is shown.

[0152] The AMF provides mobility configuration information, which may include configuration for collecting training data for sporadic load prediction. The source gNB configures the UE with an event to start measurements for sporadic load prediction. This can be an existing event or a new event.

[0153] The UE reports configured measurements, which may come from the input feature list above.

[0154] The source gNB records the collected data and tracks it using the UE context to identify the UE.

[0155] The source gNB decides on the handover.

[0156] The source gNB sends a handover request to the target gNB. This request may contain a configuration for measuring the occasional load.

[0157] Once the UE starts receiving and transmitting with the target gNB, the occasional load measurement starts. The target gNB measures the load of the target gNB by the UE for a configured period of time T. 事件 Occasional load caused by

[0158] The target gNB reports the measured occasional load back to the source gNB. This can be done in existing messages such as the UE Context Release message (3GPP TS 38.300 Fig. 9 .2.3.2.1-1)) or in a newly defined message. Reporting measurements may delay the UE context release message. The target gNB may also report its (average) composite available capacity (CAC) at the time of measurement, which can be used to filter measurements affected by congestion or to train the impact of target node congestion into a prediction model.

[0159] The source gNB stores the measured occasional load together with the input data it collected from the UE before the handover. These steps are repeated a pre-configured N times to collect a sufficient amount of training data.

[0160] The source gNB reports the collected training data to the training host.

[0161] The training host uses the provided data to train an episodic load prediction model.

[0162] The training host deploys the trained incident load prediction model to the source gNB.

[0163] Fig.12 The signaling flow for training the occasional load prediction model of an example embodiment is shown, where the required data (input data and measured occasional load) is collected directly at the training host. It is similar to the training master except that the target and source gNBs report the measured occasional load and predicted input directly to the training master (steps 15 and 17), respectively. Fig.11 The duration of the occasional load measurement of the target node can be configured by the source node, and the timing of sending the measurement from the source node to the training host can be aligned with the duration of the occasional load measurement at the target node. The training host correlates the two data sources in step 18 and uses the data sets to train the model. The training host maps the occasional load to the input data based on the "universal token" created in step 2, which identifies that these are from the same UE. For a given UE, the generation of the "universal token" is unique, but does not reveal its identity. Once the data is consumed to train the model, the token does not need to be retained on the training host.

[0164] exist Fig.11 In the embodiment shown in , all training data collection is integrated in the source gNB, which avoids the complexity of more distributed data collection and does not require the training host to be able to relate which reported measurement loads to which input sequences. Furthermore, in this embodiment, the target gNB does not need to be integrated into the training host, which may be simpler in some scenarios.

[0165] Fig.12 The embodiment shown in allows a longer load measurement period after switching. Fig.11 In an embodiment of the present invention, the measured load needs to be reported in the UE context release message so that it can be correlated with the input data based on the UE context before it is released. Fig.12 In the embodiment shown in , no such restriction is needed, which allows measuring occasional loads of UEs with bursty traffic that may not be sent or received immediately after a handover.

[0166] For example, the training host can be the source gNB or in OAM. The communication between the source and target gNB occurs over the Xn interface.

[0167] Fig.13 The signaling flow for inferring the occasional load prediction model for baseline HO is shown.

[0168] Furthermore, the AMF can be configured to activate occasional load prediction and events required to trigger data collection at the UE.

[0169] UE reports its measurements and KPIs. Events may trigger additional data collection for sporadic load forecasting. This may be an existing event, such as A3, or new events may be defined.

[0170] The source gNB collects the input data required for the prediction and can use the occasional load prediction model to predict the load if the UE is handed over to a cell in the target gNB. The inference may be triggered by a mobility event.

[0171] The source gNB can use the predicted episodic load in its handover decision, e.g. for load balancing or optimizing network quality.

[0172] The source gNB sends a handover request to the target gNB. The handover preparation message may also include the predicted occasional load.

[0173] The target gNB may use the predicted episodic load in its admission control procedure or when predicting its future load.

[0174] Fig.14 The inference approach in CHO preparation is shown.

[0175] The occasional load prediction is performed as above and the predicted load is given to both target gNB candidates in CHO preparation. The target gNB may consider the predicted load in, for example, its admission control. In addition, the source gNB may consider the predicted load in its decision which target candidate to prepare.

[0176] For other contexts, such as, for example, predicting other episodic load KPIs instead of or in addition to radio interface PRB usage, separate ML model instances may be trained.

[0177] The method may provide the capability for predicting the UE occasional load on a target (candidate) node at a source node, where it is needed, for example, for load balancing, selecting CHO preparation, etc. Learning to predict the end-to-end occasional load from UE measurements adapted to a specific context (e.g., a specific cell border) may achieve higher accuracy because the prediction is not linked. In addition, less signaling may be required between the source node and the target node.

[0178] An apparatus may include components for receiving, at a source node, measurement values ​​from a plurality of user equipments, providing, from the source node to a corresponding one of a plurality of target nodes, a handover request for each of the plurality of user equipments, the handover request including a request to measure an occasional load, determining to perform a handover of a subsequent user equipment to a subsequent target node, determining a predicted occasional load based on the received measurement values ​​and the measured occasional load, using the predicted occasional load in a mobility decision, and providing the predicted occasional load to the subsequent target node in the handover request for the subsequent user equipment.

[0179] Alternatively, an apparatus may include components for receiving, at a target node, a handover request for a user device from a source node, wherein either the handover request includes a request to measure an occasional load, and if so, providing an indication of the measured occasional load to at least one of the source node and the host node for use in training a machine learning model, or the handover request includes an indication of a predicted occasional load associated with the user device.

[0180] Alternatively, an apparatus may include components for receiving, from a source node and one of a plurality of target nodes, at least one indication of an occasional load measured for a user device at a corresponding one of a plurality of target nodes and a measurement value for the user device, and using at least one indication and the measurement value when training a machine learning model, the machine learning model being configured, when executed, to determine a predicted occasional load based on subsequent measurement values ​​of the user device.

[0181] It should be understood that the apparatus may include or be coupled to other units or modules for transmission and / or reception, such as a radio part or a radio head, etc. Although the apparatus has been described as one entity, the different modules and memories may be implemented in one or more physical or logical entities.

[0182] It is worth noting that although some embodiments have been described with respect to 5G networks, similar principles may be applied to other networks and communication systems. Therefore, although certain embodiments are described above by way of example with reference to certain example architectures of wireless networks, technologies, and standards, the embodiments may be applied to any other suitable form of communication system other than the communication system illustrated and described herein.

[0183] It is also noted herein that while the above describes exemplifying embodiments, there are several variations and modifications which may be made to the disclosed solution without departing from the scope of the present invention.

[0184] In general, various embodiments may be implemented in hardware or dedicated circuits, software, logic, or any combination thereof. Some aspects of the present disclosure may be implemented in hardware, while other aspects may be implemented in firmware or software that may be executed by a controller, microprocessor, or other computing device, but the present disclosure is not limited thereto. Although various aspects of the present disclosure may be illustrated and described in block diagrams, flow charts, or other graphical representations, it is fully understood that, as non-limiting examples, these blocks, devices, systems, techniques, or methods described herein may be implemented in hardware, software, firmware, dedicated circuits or logic, general hardware, or controllers, or other computing devices, or some combination thereof.

[0185] In this application, the term "circuitry" may refer to one or more or all of the following:

[0186] (a) pure hardware circuit implementation (such as implementation with only analog and / or digital circuit systems) and

[0187] (b) a combination of hardware circuitry and software such as (where applicable):

[0188] (i) a combination of analog and / or digital hardware circuits and software / firmware, and

[0189] (ii) any portion of hardware processor(s) with software (including digital signal processor(s), software and memory(s) that work together to enable a device such as a mobile phone or server to perform various functions), and

[0190] (c) Hardware circuit(s) and / or processor(s) that require software (e.g., firmware) to operate, such as microprocessor(s) or portion(s) of microprocessor(s), but the software may not be present when it is not required for operation.

[0191] This definition of circuitry applies to all uses of the term in this application, including in any claims. As a further example, as used in this application, the term circuitry also covers an implementation of only a hardware circuit or processor (or multiple processors) or a portion of a hardware circuit or processor and its (or their) accompanying software and / or firmware. For example, and where applicable to a particular claim element, the term circuitry also covers a baseband integrated circuit or processor integrated circuit for a mobile device, or a similar integrated circuit in a server, cellular network device, or other computing or network device.

[0192] Embodiments of the present disclosure may be implemented by computer software executable by a data processor of a mobile device (such as in a processor entity), or by hardware, or by a combination of software and hardware. Computer software or programs (also referred to as program products) include software routines, applets and / or macros that can be stored in a data storage medium readable by any device, and they include program instructions for performing specific tasks. A computer program product may include one or more computer executable components that are configured to perform embodiments when the program is run. One or more computer executable components may be at least one software code or a portion thereof.

[0193] Furthermore, it should be noted in this regard that any block of the logic flow in the Figures may represent program steps, or interconnected logic circuits, blocks and functions, or a combination of program steps and logic circuits, blocks and functions. The software may be stored on physical media such as memory chips or memory blocks implemented within a processor, magnetic media such as hard disks or floppy disks, and optical media such as, for example, DVDs and their data variants, CDs. Physical media are non-transitory media.

[0194] The memory may be of any type suitable for the local technical environment and may be implemented using any suitable data storage technology, such as semiconductor-based memory devices, magnetic memory devices and systems, optical memory devices and systems, fixed memory, and removable memory. The data processor may be of any type suitable for the local technical environment and may include, as non-limiting examples, one or more of the following: a general-purpose computer, a special-purpose computer, a microprocessor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an FPGA, a gate-level circuit, and a processor based on a multi-core processor architecture.

[0195] Embodiments of the present disclosure may be practiced in various components such as integrated circuit modules. The design of integrated circuits is generally a highly automated process. Complex and powerful software tools are available to convert a logic level design into a semiconductor circuit design that is ready to be etched and formed on a semiconductor substrate.

[0196] The protection scope of various embodiments of the present disclosure is defined by the independent claims. Embodiments and features described in this specification that do not belong to the scope of the independent claims (if any) should be interpreted as examples that help understand the various embodiments of the present disclosure.

[0197] The foregoing description has provided a complete and informative description of exemplary embodiments of the present disclosure by way of non-limiting examples. However, various modifications and adaptations will be apparent to those skilled in the relevant art when the above description is read in conjunction with the accompanying drawings and the appended claims. However, all such and similar modifications of the teachings of the present disclosure will still fall within the scope of the present invention as defined in the appended claims. Indeed, there are additional embodiments that include a combination of one or more embodiments with any other embodiments previously discussed.

Claims

1. An apparatus comprising components for: receiving, at a source node, measurements from a plurality of user equipments; providing, from the source node to a respective one of a plurality of target nodes, a handover request for each of the plurality of user equipment, the handover request comprising a request to measure an occasional load; Determining to execute a handover of a subsequent user equipment to a subsequent target node; determining a predicted contingent load based on the received measurements and the measured contingent load; using the predicted episodic load in mobility decisions; as well as The predicted contingent load is provided to the subsequent target node in a handover request for the subsequent user equipment.

2. An apparatus according to claim 1, wherein the component for determining the predicted sporadic load comprises at least one machine learning model, which, when executed, is configured to: determine the predicted sporadic load based on the measurement value of the subsequent user equipment.

3. An apparatus according to claim 2, wherein the machine learning model, when executed, is configured to: determine the predicted contingent load based on the identity of the target node.

4. The apparatus according to any one of claims 2 and 3 comprises a component for providing the received measurement values ​​to a host node for use in training the machine learning model.

5. The apparatus of claim 4, comprising a component for receiving an indication of the measured sporadic load from each of the plurality of target nodes, and providing the indication of the measured sporadic load to the host node for use in training the machine learning model.

6. The apparatus according to any one of claims 2 and 3 comprises a component for receiving an indication of the measured sporadic load from each of the plurality of target nodes for use in training the machine learning model at the source node.

7. Apparatus according to claim 5 or claim 6, comprising means for receiving from each of said plurality of target nodes an indication of a capacity measured at said occasional load.

8. The apparatus according to any one of claims 1 to 7, comprising means for determining the predicted contingent load further based on an identifier of at least one of the subsequent target node and the source node.

9. The device according to any one of claims 1 to 8, wherein the occasional load comprises: The absolute amount of resources allocated to the user equipment in at least one of the downlink and the uplink, or the ratio of resources allocated to the user equipment in at least one of the uplink and the downlink to all resources scheduled in a given time period.

10. An apparatus comprising components for: At the target node, receiving a handover request for the user equipment from the source node, wherein or the handover request includes a request to measure an occasional load; and If so, providing an indication of the measured episodic load to at least one of the source node and the host node for use in training the machine learning model; Alternatively, the handover request comprises an indication of a predicted occasional load associated with the user equipment.

11. An apparatus comprising means for: receiving, from a source node and one of a plurality of target nodes, at least one indication of an occasional load measured for a user equipment at a corresponding one of the plurality of target nodes and a measurement value for the user equipment; and The at least one indication and the measurement value are used when training a machine learning model, and the machine learning model, when executed, is configured to: determine a predicted sporadic load based on measurements of subsequent user devices.

12. A method comprising: receiving, at a source node, measurements from a plurality of user equipments; providing, from the source node to a respective one of a plurality of target nodes, a handover request for each of the plurality of user equipment, the handover request comprising a request to measure an occasional load; Determining to execute a handover of a subsequent user equipment to a subsequent target node; determining a predicted contingent load based on the received measurements and the measured contingent load; using the predicted episodic load in mobility decisions, and The predicted contingent load is provided to the subsequent target node in a handover request for the subsequent user equipment.

13. An apparatus comprising: 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, cause the apparatus to at least: receiving, at a source node, measurements from a plurality of user equipments; providing, from the source node to a respective one of a plurality of target nodes, a handover request for each of the plurality of user equipment, the handover request comprising a request to measure an occasional load; Determining to execute a handover of a subsequent user equipment to a subsequent target node; determining a predicted contingent load based on the received measurements and the measured contingent load; using the predicted episodic load in mobility decisions, and The predicted contingent load is provided to the subsequent target node in a handover request for the subsequent user equipment.

14. A computer readable medium comprising program instructions for causing an apparatus to at least perform the following operations: receiving, at a source node, measurements from a plurality of user equipments; providing, from the source node to a respective one of a plurality of target nodes, a handover request for each of the plurality of user equipment, the handover request comprising a request to measure an occasional load; Determining to execute a handover of a subsequent user equipment to a subsequent target node; determining a predicted contingent load based on the received measurements and the measured contingent load; using the predicted episodic load in mobility decisions, and The predicted contingent load is provided to the subsequent target node in a handover request for the subsequent user equipment.