WO2022030713A1 - Configuring resources in a self-organizing network - Google Patents

Configuring resources in a self-organizing network Download PDF

Info

Publication number
WO2022030713A1
WO2022030713A1 PCT/KR2021/001500 KR2021001500W WO2022030713A1 WO 2022030713 A1 WO2022030713 A1 WO 2022030713A1 KR 2021001500 W KR2021001500 W KR 2021001500W WO 2022030713 A1 WO2022030713 A1 WO 2022030713A1
Authority
WO
WIPO (PCT)
Prior art keywords
network element
nrt
handover
network
load
Prior art date
Application number
PCT/KR2021/001500
Other languages
French (fr)
Inventor
Shangbin Wu
Joan PUJOL ROIG
Yue Wang
Original Assignee
Samsung Electronics Co., Ltd.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Samsung Electronics Co., Ltd. filed Critical Samsung Electronics Co., Ltd.
Publication of WO2022030713A1 publication Critical patent/WO2022030713A1/en

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W36/00Hand-off or reselection arrangements
    • H04W36/0005Control or signalling for completing the hand-off
    • H04W36/0055Transmission or use of information for re-establishing the radio link
    • H04W36/0061Transmission or use of information for re-establishing the radio link of neighbour cell information
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W36/00Hand-off or reselection arrangements
    • H04W36/16Performing reselection for specific purposes
    • H04W36/165Performing reselection for specific purposes for reducing network power consumption
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W24/00Supervisory, monitoring or testing arrangements
    • H04W24/02Arrangements for optimising operational condition
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W36/00Hand-off or reselection arrangements
    • H04W36/0005Control or signalling for completing the hand-off
    • H04W36/0083Determination of parameters used for hand-off, e.g. generation or modification of neighbour cell lists
    • H04W36/00837Determination of triggering parameters for hand-off
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W84/00Network topologies
    • H04W84/18Self-organising networks, e.g. ad-hoc networks or sensor networks
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02DCLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00Reducing energy consumption in communication networks
    • Y02D30/70Reducing energy consumption in communication networks in wireless communication networks

Definitions

  • the disclosure relates to configuring resources in self-organizing networks.
  • a self-organizing network is a cost-effective known way to adapt and configure network environments by allowing network entities and components to self-configure in an autonomous manner. This reduces the burden of centralized planning and manual intervention, resulting in an overall reduction in operation cost (OPEX).
  • Traditional SONs rely on pre-defined threshold-based policies that monitor hand-picked parameters chosen by operators. These policies are reactive, i.e. reconfigurations are only triggered after incidents have occurred, and rely on a set of heuristically chosen parameters, which results in network deployments that lack flexibility and agility.
  • Traditional SON approaches find limited applicability on 5G and beyond 5G networks, due to the very heterogeneous set of scenarios envisioned for these type technologies.
  • proactive policies that leverage cutting-edge artificial intelligence (AI) to have a zero-touch autonomous network are starting to be used.
  • Configuring resources in self-organizing networks is required for function enhancement.
  • Embodiments can provide a novel signalling procedure that may support both general and artificial intelligence (AI) based network resources orchestration for function enhancement.
  • functions that embodiments can enhance include, but are not limited to, load balancing and energy saving.
  • the signalling can be an extension of current signalling, but may provide parameters that may not be directly accessible to the base station, such as power parameters of remote units (RUs).
  • the proposed signalling may leverage the existing Automatic Neighbour (ANR) function.
  • ANR is a SON feature for automatically configurating neighbouring cells.
  • ANR can use a Neighbour Relation Table (NRT), typically stored by a base station, which may comprise entries for the neighbouring cells of the base station. Each entry in the NRT can contain information regarding one of the neighbouring cells.
  • NRT Neighbour Relation Table
  • a resource policy module either at the O&M function, at the gNB, or at both, can use the extended signalling to estimate loads and power consumptions to decide resource management and orchestration policies. This can directly impact and improve load balancing and energy saving.
  • Figure 1 schematically illustrates first and second network elements of an example SON
  • Figure 2 schematically illustrates an example of the first and second network elements exchanging data relating to an extended NRT
  • Figures 3a - 3c are example architectures of network elements using the extended NRT
  • Figure 4 details an example extended NRT
  • Figures 5a - 5b schematically illustrates an example of a cell provided by a network area being divided into a plurality of logical areas
  • Figure 6 is a block diagram of a post-handover load estimator that can be included in a network element
  • Figure 7 schematically illustrates operation of a decision-making module that can be included in a network element.
  • Embodiments of the disclosure can provide a solution to allow AI and machine learning (ML) as well as general approaches to be leveraged at a 5G nodeB (gNB) by providing novel signalling to access performance-relevant metrics.
  • embodiments can operate a novel signalling and information exchange procedure that allows the 5G nodeB (gNB) to gather information that is currently inaccessible according to the relevant existing standards. This information can be exposed to a third module entity which is in charge of configurating and orchestrating the available resources.
  • the novel signaling procedure can utilize an extended Neighbor Relation Table that provides extended information for resource management and orchestration.
  • the extended NRT can capture not only standard/general entries, which are directly accessible by the gNB (in accordance with a relevant standard), but also additional/non-standard measurements, such as those provided by a network operator, which are not directly accessible through the gNB according to the standard.
  • a computer-implemented method of configuring a network element in a Self-organizing Network comprising: receiving, by a first network element from a second network element, at least one attribute data relating to the second network element included in a Neighbour Relation Table, NRT, field and at least one additional attribute data relating to the second network element; updating by the first network element, an extended NRT, wherein the extended NRT comprises an NRT comprising the received at least one attribute data and the received at least one additional attribute data, and using the extended NRT to configure the first and/or the second network element.
  • the additional network element attribute data may comprise data not included in a general NRT field, such as a measurement not normally directly accessible by a network element such as a base station/gNB, e.g. it may comprise a measurement provided by a network operator.
  • the additional network element attribute data may be unknown to the network element, but known to an upper-level network management function.
  • the additional network element may be obtained from an upper-level network management function of the SON, such as an O&M function.
  • the NRT may comprise a standard NRT, such as used by an ANR function of the SON.
  • the step of using the extended NRT to configure the first and/or the second network element may comprise determining whether or not to handover a user device serviced by a cell provided by the second network element (e.g. base station) to a cell provided by the first network element depending on the at least one additional attribute data.
  • a cell provided by the second network element e.g. base station
  • the at least one additional attribute data may comprise a power parameter relating to the second network element, a load of the second network element, throughput of at least part of a cell provided by the second network element and/or one or more barred access class of a cell provided by the second network element.
  • the step of using the extended NRT to configure the first and/or the second network element can comprise: estimating a post-handover load of a cell provided by the first network element if at least one user device is handed over from a cell provided by the second network element to the cell provided by the first network element, and performing the handover of the at least one user device if it is estimated that the handover will reduce energy consumption of the second network element (e.g. by the second network element switching to a stand-by/de-activated state).
  • Cell coverage provided by a said network element may be logically divided into a plurality of areas and each said network element may store data comprising user device throughput and location information for each of its plurality of areas.
  • the plurality of areas may be arranged radially or concentrically, for example, around a physical location the network element.
  • the method may comprise: obtaining location information of a said user device, and determining in which of the plurality of areas the user device will be located following the handover.
  • the user device throughput of a said area can be computed based on average throughput of all user devices currently within the area.
  • the user throughput of the cell provided by the first network element can be computed using a formula (or mathematical equivalent):
  • L 1 represents a current load of the user device in the cell provided by the second network element
  • T 1 represents a throughput of the area in which the user device is currently located in the cell provided by the second network element
  • T 2 represents a throughput of the determined area in the cell provided by the first network element where the user device will be located following the handover.
  • the method may estimate energy consumption at the first network element (e.g. a base station providing a target cell) before and/or after the handover based on the estimated load and the power parameters in the extended NRT.
  • the first network element e.g. a base station providing a target cell
  • the step of using the extended NRT to configure the first and/or the second network element can comprise deactivating one of the first and the second network element to save energy.
  • the method may comprise: storing historical data representing load status of a cell provided by a said network element over a plurality of time periods and user devices throughput in the cell over the plurality of time periods; using the historical data to predict a number of active cells required to fulfil current or future user device demand in a multi-cell area covered by a plurality of the network elements, and activating and/or de-activating the network elements based on the prediction.
  • a Machine Learning algorithm such as deep reinforcement learning, can be used for the predicting the number of the active cells required to fulfil the current or future user device demand.
  • the method may comprise: using the extended NRT to estimate load at the first network element after a handover and using the extended NRT to estimate load at the second network element after the handover, wherein the handover is part of a network element load-balancing operation that is triggered when load at a network element exceeds or meets a threshold; determining whether the estimated load at the first network element and/or the estimated load at the second network element would trigger a further handover due to exceeding or meeting the threshold, and not performing the handover if it is determined that the estimated load(s) would trigger the further handover.
  • the at least one additional attribute data may comprise data representing a class of service (e.g. emergency/priority or non-emergency/priority) of a user device serviced by a cell provided by the second network element.
  • the step of using the extended NRT to configure the first and/or the second network element may comprise determining whether or not to handover the user device serviced by the cell provided by the second network element to a cell provided by the first network element depending on the class of service of the user device.
  • the network elements may be configured to operate an Automatic Neighbour Relation (ANR) function that can be used for the configuration.
  • ANR Automatic Neighbour Relation
  • the SON may be implemented on a mobile radio access network, such as a 5G or beyond network, etc.
  • a computer-implemented method of configuring a network element in a Self-organizing Network comprising: at a second network element, obtaining data representing at least one attribute of the second network element included in a Neighbour Relation Table, NRT, field and obtaining data representing at least one additional attribute of the second network element, and transmitting, by the second network element to a first network element, the at least one attribute data and the at least one additional attribute data for use in an enhanced NRT for configuring a said network element.
  • a system operating a Self-organizing Network comprising a first network element and a second network element, wherein the first and/or the second network element is/are configured to execute a method substantially as described herein.
  • a network element of a Self-organizing Network comprising: a communications interface configured to communicate with at least one other network element, and a processor configured to execute a method substantially as described herein.
  • the network element may comprise a decision-making module configured to select an orchestration and/or management policy based on the additional attribute data included in the extended NRT.
  • apparatus comprising a processor configured to at least partially execute one or more methods substantially as described herein.
  • the apparatus may comprise a mobile computing device, such as a smartphone.
  • a self-organizing network is a cost-effective known way to adapt and configure network environments by allowing network entities and components to self-configure in an autonomous manner. This reduces the burden of centralized planning and manual intervention, resulting in an overall reduction in operation cost (OPEX).
  • Traditional SONs rely on pre-defined threshold-based policies that monitor hand-picked parameters chosen by operators. These policies are reactive, i.e. reconfigurations are only triggered after incidents have occurred, and rely on a set of heuristically chosen parameters, which results in network deployments that lack flexibility and agility.
  • Traditional SON approaches find limited applicability on 5G and beyond 5G networks, due to the very heterogeneous set of scenarios envisioned for these type technologies.
  • proactive policies that leverage cutting-edge artificial intelligence (AI) to have a zero-touch autonomous network are starting to be used.
  • gNB 5G base station/nodeB
  • DU data unit
  • RU remote units
  • power model the load information of its data unit
  • KPI key performance indicator
  • a centralized Operations and Maintenance (O&M) or Management and Orchestration (MANO) function may manage and orchestrate the network cells.
  • This managing function needs to know the capabilities of the current active cells such that it can estimate the loads, power consumptions, and other KPIs to take management decisions accordingly.
  • some metrics are not directly known or accessible for such operations.
  • T. Uchino, K. Kai, T. Toeda, H. Takahashi, 'Specifications of NR Higher Layer in 5G' NTT Docomo technical journal, vol. 20, no. 3, Jan. 2019, for instance, describes C-plane functions and U-plane functions that can manage/configure network resources.
  • Cell load information can be exchanged. However, this is not sufficient to estimate the incurring additional load when a user is handed over to another cell.
  • a cell may migrate its attached users to another cell (Cell B) in order to enter stand-by mode and reduce the network power consumption . The migrated users will impose additional load to the target Cell B.
  • Cell A does not know the capability of Cell B or an estimate of the load the to-be-migrated users will impose on Cell B, then Cell B may be overloaded by these additional users.
  • Figure 1 schematically illustrates an example of network elements/nodes of a SON 100 that can be configured to execute embodiments.
  • Embodiments may use any suitable SON implementation.
  • SON has been codified in 3GPP Release 8 and subsequent specifications in a series of standards including 36.902, as well as public white papers outlining use cases.
  • a first network element 102A of the SON 100 comprises a processor 104A, a memory 106A and a communications interface 108A
  • a second network element 102B comprises a processor 104B, a memory 106B and a communications interface 108B.
  • Each of the network elements can comprise any suitable apparatus, such as a base station (e.g a gNB in a 5G implementation) or a server configured to perform O&M functions, and may be operated by a service/network provider entity or the like.
  • a base station e.g a gNB in a 5G implementation
  • a server configured to perform O&M functions
  • Other components and features of the network elements will be well-known to the skilled person and need not be described herein in detail. Two network elements are shown in the example for simplicity, but it will be understood that embodiments may use any reasonable number/combination of network elements.
  • the SON may be implemented in a 5G (or beyond/later/higher, e.g. 6
  • the network elements 102 are configured to operate a novel signaling procedure, which utilizes an extended NRT to provide extended information for resource management and orchestration.
  • the extended neighbor relation table can comprise not only general entries of a general NRT (as defined by a relevant standard, such as 3GPP TS 32.511, 'Automatic Neighbour Relation (ANR) management; Concepts and requirements', V16.0.0, Jan. 2020 (where the NRT is called a 'neighbor cell relation table')), which are directly accessible by a certain type of network element, such as a gNB, but also other data/measurements, such as ones provided by a network operator or the like, which are generally not directly accessible through the gNB.
  • a gNB a certain type of network element
  • other data/measurements such as ones provided by a network operator or the like
  • the extended NRT can be a separate and different message from a standard NRT.
  • the information contained in the extended NRT can be used by one or more network element to automatically perform improved resource management and/or orchestration, which can involve configuring at least one attribute of one or more local/remote network element, that is not possible using generally-configured network elements.
  • a gNB can have an Automatic Neighbor Relation (ANR) function that can be used for configuration.
  • ANR Automatic Neighbor Relation
  • an intelligent orchestration policy module in either the O&M function, the gNB, or both (i.e. centralized, distributed, or mixed), which will be responsible for managing/configuring resources.
  • instructions can be executed by such a module to process the information contained in the extended NRT to optimize functions such as load balancing and/or energy saving.
  • Figure 2 schematically illustrates the first network element 102A sending 202 a request to the second network element 102B for the extended NRT and the second network element responding 204 with its entries for the table.
  • the first node can then update its extended NRT with the received entries.
  • the extended NRT can also comprise data relating to the first network element (typically the same data fields as obtained for the second network element that the first network elements has obtained/generated in a similar manner) and/or data relating to at least one further network element that is communication with it. All or some of the data in the extended NRT may be processed (by the first network element and/or a remote processor/network element) to perform configuration of one or more of the network elements.
  • Figures 3a - 3c respectively illustrate three embodiments based on different self-organized architectures.
  • Figure 3a shows a gNB centralized architecture with an O&M orchestration policy decision-making module 302A that is external to the gNBs.
  • the O&M policy module is in charge of providing the entries of the extended NRT for the different gNBs and can acquire extended NRTs from neighboring gNBs (only one gNB 303A and one extended NRT 304A are shown in the diagram for simplicity) via the ANR function. Once the extended NRTs are obtained from all the gNBs, the O&M orchestration policy module can make resource configuration/management decisions based on the data.
  • FIG. 3b shows an inter gNB architecture, where each gNB has an individual/internal orchestration policy decision-making module 302B (only one gNB 303B and one extended NRT 304B are shown in the diagram for simplicity).
  • the different gNBs exchange their information in order to fill theirs extended NRTs.
  • the orchestration policy module of each gNB maps the local extended NRT to resource orchestration actions.
  • Figure 3c shows a gNB and O&M mixed/hybrid architecture.
  • the extended NRT data can be exchanged via both an external O&M policy decision-making module 302C and an intra gNB 305C.
  • This type of architecture can follow a master-slaver decision structure (O&M being the master, and intra gNB being the slave, or vice versa), or a joint decision making architecture.
  • Resource management decisions will be made by both the external module 302C and the internal policy modules 305C being executed by each gNB 303C (again, only one gNB 303C and one extended NRT 304C are shown in the diagram for simplicity).
  • the decision-making making module 302A - 302C/305C in any of the above embodiments may be implemented using AI or any other type of O&M policy module.
  • FIG. 4 shows an example of an extended NRT.
  • the extended NRT is an extension of an ordinary/general NRT. It includes information which is not accessible by cells/nodes, but can be provided by the O&M function of the SON.
  • the example comprises traditional O&M NRT fields 402, 404, 406, i.e., IDs of neighbors (e.g., 1, 2, 3,...), neighbor relation attributes No HO, No X2.
  • It also includes a set of non-general parameters in the form of extended attributes 408, 410, 412, which, in the example, comprise of data/entries that can be obtained by network elements/cells, e.g. load and area throughput, as well as data/entries that cannot be directly measured by network elements/cells.
  • Examples include power parameters, mainly of the Remote Unit, but can include a power parameter of the base station or other data unit parameters.
  • the Slope and Intercept values can be obtained via measurement by operators or vendors. These values also rely on how the site is deployed and therefore are not directly accessible in the site.
  • the load and throughput can be measured periodically by the network element.
  • the power parameters may be obtained using a power model of the amplifier of the network element/base station, e.g. computing an indication of how much power will be required to fulfil a certain percentage of the maximum (100%) of the cell's load.
  • the network operator/network management function may send a command to a network element to read the element/node's statistics.
  • the network element can reply with its statistics and the network operator/network management function can send a response comprising the node/element's properties.
  • the extended NRT can comprise additional or alternative data/entries that can be used for different types of configuration purposes, such as service enhancement.
  • ARB Access Class Barring
  • Its function is to bar access of UEs from certain services, in order to maintain KPIs of the network. For example, if the cell is overloaded with emergency services, or another high-priority service, then the cell can bar UEs of other non-emergency services so that the cell can reserve all resources for those emergencies.
  • a similar situation can occur in small cells when resources are limited.
  • base stations can exchange information indicating which classes of UEs (e.g. non-emergency/low priority) are barred in each cell. Then, when performing handover and/or load balancing, the source base station/cell can avoid handing over UEs of these classes to the target base station/cell.
  • a network element may logically divide its/each cell coverage into a number of areas. Examples are illustrated in Figures 5a and 5b. In the example of Figure 5a there are four generally radial areas/segments 502A - 502D arranged around the base station/gNB 500. In the example of Figure 5b there are three generally concentric areas/rings 504A - 504C arranged around the gNB 500. It will be understood that other numbers and arrangements of areas can be used by alternative embodiments. The number and type of logical areas may be determined by software or a user upon configuration or deployment, for example.
  • each of the network elements can maintain a database, or other data structure, for storing the (normalized) throughput and location information of its areas.
  • An example of the area throughput information is presented as a table below:
  • Each area can be given an identifier.
  • the throughput of an area can be computed using the throughput of users within the area. One way to do this is by computing the average throughput of users. Also, the throughput can be normalized in terms of the number of layers (data streams in embodiments that use multiple-input, multiple-output (MIMO)).
  • MIMO multiple-input, multiple-output
  • the geographical coordinate of the center of the area can also be included in the data.
  • a cell/gNB can estimate the user throughput after migrating a user to another cell, if provided the user's location information.In some embodiments the first network element 102A will send a message to the second network element 102B to request its area throughput information. The second network element will respond with its area throughput information.
  • the area throughput information can be obtained by estimating the throughput of users/devices whose locations are within an area.
  • One or more of the network elements may be provided with a post-handover load estimator (e.g. as a software module).
  • the estimator will use the users' information to estimate the users' load if they are handed over to another cell/node/network element.
  • the estimator 600 is schematically illustrated in Figure 6.
  • the information used may include the pre-handover load 602, the current cell area throughput 604, the target cell area throughput 606, and location information of the user 608.
  • location information can be obtained via the Location Management Function (LMF).
  • LMF Location Management Function
  • the serving cell can determine in which area of the target cell the user will be located after the potential handover (e.g. by determining which of the areas of the target cell is closest, in terms of distance/location, to the area in which the user device is currently located) and also estimate the throughput after handover.
  • a goal can be to estimate the post-handover load.
  • the post-handover load can be estimated using information including the current load, the current cell area throughput and the target cell area throughput.
  • An example of post handover load estimation will now be described, where the impact of moving a user currently in Cell 1 to area 2 of Cell 2 is computed. Assuming that the user in Cell 1 has causes load L 1 and is located in area 1. The throughput of area 1 of Cell 1 is T 1 . Let the throughput of area 2 of Cell2 be T 2 . Then, the estimated post handover load in Cell2 can be expressed as:
  • the energy consumption before and after user handover can also be estimated.
  • the module can decide whether the action of user handover can reduce energy consumption. For instance, after estimating the load before and the load after potential handover, the module can know the power consumption before and the power consumption after the potential handover and can then decide whether or not to perform the potential handover, e.g. only perform the handover if the power consumption after the potential handover would be reduced compared to the power consumption before the potential handover.
  • the decision making module can derive the power consumptions of different cells under different loads. Then, after the decision making module estimates the post-handover loads of cells, it can compute a handover allocation such that power consumption is minimised/reduced.
  • Another usage of post handover load estimation is to avoid a ping-pong effect during load balancing.
  • a cell may handover a user to another cell in order to reduce load imbalance.
  • the migrated user can impose a load on the target cell such that load imbalance increases. In this case, this user has to be handed over back to the original cell and the ping-pong effect occurs.
  • the estimated post handover load a cell can estimate the imbalance in advance to avoid this ping-pong effect.
  • the imbalance can be estimated, for example, by comparing the difference between the highest load and the lowest load of different cells. To give an example, CellA has load 20% and CellB had load 10%.
  • Load balancing may be triggered if the highest load and the lowest load differ by a value larger than or equal to 10%, for instance. Thus, the load balancing would be trigged and CellA will seek to handover one or more users to CellB. Without knowing post-handover loads of CellB, after handover, CellA's load reduces to 12%, but the new load of CellB becomes 22%. The 8% drop in CellA may cause a high load increase in CellB because channel conditions are different. Now, load balancing will be triggered again (immediately after the previous load-balancing operation has been completed) and CellB moves users back to CellA and we go back to the origin and repeat. However, if the estimated post-handover loads of CellB are known at the decision module the decision-making module will not trigger this load balancing in the first place.
  • an intelligent decision-making module e.g. a software module
  • each network element which keeps track of historical data and network events, and can select resource orchestration and management policies accordingly.
  • the module can employ users' information, as well as network information, to infer future user behavior as well as network resources consumption and can determine MANO policies based on these.
  • the module may obtain a snapshot of the current network and users' metrics, processes it, and store both the metrics and the processed data locally.
  • the current user information includes, but it is not limited to, physical layer metrics (e.g. channel estimation, throughput), users' and cell load, and the extended NRT. Based on these, the module can undertake MANO policies, which can range from energy-saving solutions, to load balancing and/or service enhancement.
  • FIG. 7 schematically illustrates how the decision-making module 700 can be used to determine optimal energy solution policies.
  • the module obtains the current cell load status 702 as well as the actual user devices' throughputs 704 (e.g. channel, Physical Resource Block usage). Furthermore, the module has recorded the historical values of these metrics over previous decision periods, e.g. the last 10 decision periods. Based on the metrics trends, using traditional forecasting algorithms, or more advanced machine learning solutions, the module can decide that for the expected future resource demand, a multi-cell covered area can reduce the number of active cells 706 from x to y, where y ⁇ x, by selectively activating or de-activating the relevant network elements. Thus, the overall network power consumption is reduced while the users' and networks' services are not disrupted.
  • an ML algorithm such as deep reinforcement learning can be used by the decision-making module 700.
  • Examples of the inputs and outputs of the ML algorithm are illustrated in Figure 7.
  • This learning process can include (or be similar to) giving several input samples to a deep neural network (DNN) so that the DNN generates different outputs (e.g. indicating which cell is on/off).
  • DNN deep neural network
  • the DNN can be told that each action it takes has a certain reward; for example, a high reward when consuming less power and vice versa, provided that KPIs are met.
  • the DNN will stabilize and converge to provide the optimal action for each input. If the current condition matches a historical condition then the cells will be activated/deactivated as previously done. Therefore, in principle, it is possible to can build a table storing the preferred action for each condition. However, when there are too many conditions and actions, the table will be too large and in this case, a DNN can be used instead of a table.
  • Embodiments can be performed by means of instructions being executed by a processor of one or more of the network elements.
  • a developer may make a design choice regarding how, and on which network element, the instructions are executed.
  • at least some of the steps described herein may be re-ordered or omitted. One or more additional steps may be performed in some cases.
  • the steps are described as being performed in sequence, in alternative embodiments at least some of them may be performed concurrently and/or by different ones of the network elements, or even at other/remote computing devices or a cloud service.
  • embodiments can be implemented using any suitable software, programming language, data editors, etc, and may be represented/stored/processed using any suitable data structures and formats.
  • Such an apparatus and/or system may be configured to perform a method according to any aspect, embodiment, example or claim disclosed herein.
  • Such an apparatus may comprise one or more elements, for example one or more of receivers, transmitters, transceivers, processors, controllers, modules, units, and the like, each element configured to perform one or more corresponding processes, operations and/or method steps for implementing the techniques described herein.
  • an operation/function of X may be performed by a module configured to perform X (or an X-module).
  • the one or more elements may be implemented in the form of hardware, software, or any combination of hardware and software.
  • examples of the disclosure may be implemented in the form of hardware, software or any combination of hardware and software. Any such software may be stored in the form of volatile or non-volatile storage, for example a storage device like a ROM, whether erasable or rewritable or not, or in the form of memory such as, for example, RAM, memory chips, device or integrated circuits or on an optically or magnetically readable medium such as, for example, a CD, DVD, magnetic disk or magnetic tape or the like.
  • volatile or non-volatile storage for example a storage device like a ROM, whether erasable or rewritable or not
  • memory such as, for example, RAM, memory chips, device or integrated circuits or on an optically or magnetically readable medium such as, for example, a CD, DVD, magnetic disk or magnetic tape or the like.
  • the storage devices and storage media are embodiments of machine-readable storage that are suitable for storing a program or programs comprising instructions that, when executed, implement certain examples of the disclosure. Accordingly, certain example provide a program comprising code for implementing a method, apparatus or system according to any example, embodiment, aspect and/or claim disclosed herein, and/or a machine-readable storage storing such a program. Still further, such programs may be conveyed electronically via any medium, for example a communication signal carried over a wired or wireless connection.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

A system and method of configuring a network element in a Self-organizing Network, SON. Embodiments comprise receiving, by a first network element from a second network element, at least one attribute data relating to the second network element included in a Neighbour Relation Table, NRT, field and at least one additional attribute data relating to the second network element. Embodiments further comprise updating, by the first network element, an extended NRT, wherein the extended NRT comprises the received at least one attribute data and the received at least one additional attribute data, and using the extended NRT to configure at least one of the first network element or the second network element.

Description

CONFIGURING RESOURCES IN A SELF-ORGANIZING NETWORK
The disclosure relates to configuring resources in self-organizing networks.
A self-organizing network (SON) is a cost-effective known way to adapt and configure network environments by allowing network entities and components to self-configure in an autonomous manner. This reduces the burden of centralized planning and manual intervention, resulting in an overall reduction in operation cost (OPEX). Traditional SONs rely on pre-defined threshold-based policies that monitor hand-picked parameters chosen by operators. These policies are reactive, i.e. reconfigurations are only triggered after incidents have occurred, and rely on a set of heuristically chosen parameters, which results in network deployments that lack flexibility and agility. Traditional SON approaches find limited applicability on 5G and beyond 5G networks, due to the very heterogeneous set of scenarios envisioned for these type technologies. In order to overcome the limitations of traditional SONs, proactive policies that leverage cutting-edge artificial intelligence (AI) to have a zero-touch autonomous network are starting to be used.
Configuring resources in self-organizing networks is required for function enhancement.
Embodiments can provide a novel signalling procedure that may support both general and artificial intelligence (AI) based network resources orchestration for function enhancement. Examples of functions that embodiments can enhance include, but are not limited to, load balancing and energy saving. In embodiments the signalling can be an extension of current signalling, but may provide parameters that may not be directly accessible to the base station, such as power parameters of remote units (RUs). The proposed signalling may leverage the existing Automatic Neighbour (ANR) function. ANR is a SON feature for automatically configurating neighbouring cells. ANR can use a Neighbour Relation Table (NRT), typically stored by a base station, which may comprise entries for the neighbouring cells of the base station. Each entry in the NRT can contain information regarding one of the neighbouring cells. In some embodiments a resource policy module, either at the O&M function, at the gNB, or at both, can use the extended signalling to estimate loads and power consumptions to decide resource management and orchestration policies. This can directly impact and improve load balancing and energy saving.
For a better understanding of the disclosure, and to show how embodiments of the same may be carried into effect, reference will now be made, by way of example, to the accompanying diagrammatic drawings in which:
Figure 1 schematically illustrates first and second network elements of an example SON;
Figure 2 schematically illustrates an example of the first and second network elements exchanging data relating to an extended NRT;
Figures 3a - 3c are example architectures of network elements using the extended NRT;
Figure 4 details an example extended NRT;
Figures 5a - 5b schematically illustrates an example of a cell provided by a network area being divided into a plurality of logical areas;
Figure 6 is a block diagram of a post-handover load estimator that can be included in a network element, and
Figure 7 schematically illustrates operation of a decision-making module that can be included in a network element.
Embodiments of the disclosure can provide a solution to allow AI and machine learning (ML) as well as general approaches to be leveraged at a 5G nodeB (gNB) by providing novel signalling to access performance-relevant metrics. In particular, embodiments can operate a novel signalling and information exchange procedure that allows the 5G nodeB (gNB) to gather information that is currently inaccessible according to the relevant existing standards. This information can be exposed to a third module entity which is in charge of configurating and orchestrating the available resources. In embodiments the novel signaling procedure can utilize an extended Neighbor Relation Table that provides extended information for resource management and orchestration. The extended NRT can capture not only standard/general entries, which are directly accessible by the gNB (in accordance with a relevant standard), but also additional/non-standard measurements, such as those provided by a network operator, which are not directly accessible through the gNB according to the standard.
According to a first aspect of the disclosure there is provided a computer-implemented method of configuring a network element in a Self-organizing Network, SON, the method comprising: receiving, by a first network element from a second network element, at least one attribute data relating to the second network element included in a Neighbour Relation Table, NRT, field and at least one additional attribute data relating to the second network element; updating by the first network element, an extended NRT, wherein the extended NRT comprises an NRT comprising the received at least one attribute data and the received at least one additional attribute data, and using the extended NRT to configure the first and/or the second network element.
The additional network element attribute data may comprise data not included in a general NRT field, such as a measurement not normally directly accessible by a network element such as a base station/gNB, e.g. it may comprise a measurement provided by a network operator.
The additional network element attribute data may be unknown to the network element, but known to an upper-level network management function. Thus, the additional network element may be obtained from an upper-level network management function of the SON, such as an O&M function.
The NRT may comprise a standard NRT, such as used by an ANR function of the SON.
The step of using the extended NRT to configure the first and/or the second network element may comprise determining whether or not to handover a user device serviced by a cell provided by the second network element (e.g. base station) to a cell provided by the first network element depending on the at least one additional attribute data.
The at least one additional attribute data may comprise a power parameter relating to the second network element, a load of the second network element, throughput of at least part of a cell provided by the second network element and/or one or more barred access class of a cell provided by the second network element.
The step of using the extended NRT to configure the first and/or the second network element can comprise: estimating a post-handover load of a cell provided by the first network element if at least one user device is handed over from a cell provided by the second network element to the cell provided by the first network element, and performing the handover of the at least one user device if it is estimated that the handover will reduce energy consumption of the second network element (e.g. by the second network element switching to a stand-by/de-activated state).
Cell coverage provided by a said network element may be logically divided into a plurality of areas and each said network element may store data comprising user device throughput and location information for each of its plurality of areas. The plurality of areas may be arranged radially or concentrically, for example, around a physical location the network element.
The method may comprise: obtaining location information of a said user device, and determining in which of the plurality of areas the user device will be located following the handover.
The user device throughput of a said area can be computed based on average throughput of all user devices currently within the area. The user throughput of the cell provided by the first network element can be computed using a formula (or mathematical equivalent):
Figure PCTKR2021001500-appb-img-000001
where: L 1 represents a current load of the user device in the cell provided by the second network element; T 1 represents a throughput of the area in which the user device is currently located in the cell provided by the second network element, and T 2 represents a throughput of the determined area in the cell provided by the first network element where the user device will be located following the handover.
The method may estimate energy consumption at the first network element (e.g. a base station providing a target cell) before and/or after the handover based on the estimated load and the power parameters in the extended NRT.
The step of using the extended NRT to configure the first and/or the second network element can comprise deactivating one of the first and the second network element to save energy.
The method may comprise: storing historical data representing load status of a cell provided by a said network element over a plurality of time periods and user devices throughput in the cell over the plurality of time periods; using the historical data to predict a number of active cells required to fulfil current or future user device demand in a multi-cell area covered by a plurality of the network elements, and activating and/or de-activating the network elements based on the prediction.
In some embodiments a Machine Learning algorithm, such as deep reinforcement learning, can be used for the predicting the number of the active cells required to fulfil the current or future user device demand.
The method may comprise: using the extended NRT to estimate load at the first network element after a handover and using the extended NRT to estimate load at the second network element after the handover, wherein the handover is part of a network element load-balancing operation that is triggered when load at a network element exceeds or meets a threshold; determining whether the estimated load at the first network element and/or the estimated load at the second network element would trigger a further handover due to exceeding or meeting the threshold, and not performing the handover if it is determined that the estimated load(s) would trigger the further handover.
The at least one additional attribute data may comprise data representing a class of service (e.g. emergency/priority or non-emergency/priority) of a user device serviced by a cell provided by the second network element. The step of using the extended NRT to configure the first and/or the second network element may comprise determining whether or not to handover the user device serviced by the cell provided by the second network element to a cell provided by the first network element depending on the class of service of the user device.
The network elements may be configured to operate an Automatic Neighbour Relation (ANR) function that can be used for the configuration.
The SON may be implemented on a mobile radio access network, such as a 5G or beyond network, etc.
According to another aspect of the disclosure there is provided a computer-implemented method of configuring a network element in a Self-organizing Network, SON, the method comprising: at a second network element, obtaining data representing at least one attribute of the second network element included in a Neighbour Relation Table, NRT, field and obtaining data representing at least one additional attribute of the second network element, and transmitting, by the second network element to a first network element, the at least one attribute data and the at least one additional attribute data for use in an enhanced NRT for configuring a said network element.
According to another aspect of the disclosure there is provided a system operating a Self-organizing Network, SON, the system comprising a first network element and a second network element, wherein the first and/or the second network element is/are configured to execute a method substantially as described herein.
According to yet another aspect of the disclosure there is provided a network element of a Self-organizing Network, SON, the network element comprising: a communications interface configured to communicate with at least one other network element, and a processor configured to execute a method substantially as described herein.
The network element may comprise a decision-making module configured to select an orchestration and/or management policy based on the additional attribute data included in the extended NRT.
According to another aspect of the disclosure there is provided apparatus comprising a processor configured to at least partially execute one or more methods substantially as described herein. The apparatus may comprise a mobile computing device, such as a smartphone.
According to another aspect of the disclosure there is provided computer readable medium (or circuitry) storing a computer program to operate methods substantially as described herein.
According to the disclosure, there is provided a method and apparatus as set forth in the appended claims. Other features of the disclosure will be apparent form the dependent claims, and the description which follows.
A self-organizing network (SON) is a cost-effective known way to adapt and configure network environments by allowing network entities and components to self-configure in an autonomous manner. This reduces the burden of centralized planning and manual intervention, resulting in an overall reduction in operation cost (OPEX). Traditional SONs rely on pre-defined threshold-based policies that monitor hand-picked parameters chosen by operators. These policies are reactive, i.e. reconfigurations are only triggered after incidents have occurred, and rely on a set of heuristically chosen parameters, which results in network deployments that lack flexibility and agility. Traditional SON approaches find limited applicability on 5G and beyond 5G networks, due to the very heterogeneous set of scenarios envisioned for these type technologies. In order to overcome the limitations of traditional SONs, proactive policies that leverage cutting-edge artificial intelligence (AI) to have a zero-touch autonomous network are starting to be used.
In current network configuration, there are 5G base station/nodeB (gNB) related metrics that are unknown to the gNB, but known to upper-level network management functions, e.g. the load information of its data unit (DU), the configuration of remote units (RUs), and the power model. These are not directly accessible by the gNB but are useful for management and resource orchestration. For example, in an RU having an energy-saving feature, the expected amount of power consumption is a key performance indicator (KPI) which can only be obtained if a predefined power model exists and is accessible by the RU.
A centralized Operations and Maintenance (O&M) or Management and Orchestration (MANO) function may manage and orchestrate the network cells. This managing function needs to know the capabilities of the current active cells such that it can estimate the loads, power consumptions, and other KPIs to take management decisions accordingly. However, as mentioned above, some metrics are not directly known or accessible for such operations. T. Uchino, K. Kai, T. Toeda, H. Takahashi, 'Specifications of NR Higher Layer in 5G' NTT Docomo technical journal, vol. 20, no. 3, Jan. 2019, for instance, describes C-plane functions and U-plane functions that can manage/configure network resources.
In existing SON settings, cell load information can be exchanged. However, this is not sufficient to estimate the incurring additional load when a user is handed over to another cell. In a known energy saving feature, a cell (Cell A) may migrate its attached users to another cell (Cell B) in order to enter stand-by mode and reduce the network power consumption . The migrated users will impose additional load to the target Cell B. During the handover, if Cell A does not know the capability of Cell B or an estimate of the load the to-be-migrated users will impose on Cell B, then Cell B may be overloaded by these additional users.
Figure 1 schematically illustrates an example of network elements/nodes of a SON 100 that can be configured to execute embodiments. Embodiments may use any suitable SON implementation. For example, SON has been codified in 3GPP Release 8 and subsequent specifications in a series of standards including 36.902, as well as public white papers outlining use cases.
A first network element 102A of the SON 100 comprises a processor 104A, a memory 106A and a communications interface 108A, and a second network element 102B comprises a processor 104B, a memory 106B and a communications interface 108B. Each of the network elements can comprise any suitable apparatus, such as a base station (e.g a gNB in a 5G implementation) or a server configured to perform O&M functions, and may be operated by a service/network provider entity or the like. Other components and features of the network elements will be well-known to the skilled person and need not be described herein in detail. Two network elements are shown in the example for simplicity, but it will be understood that embodiments may use any reasonable number/combination of network elements. The SON may be implemented in a 5G (or beyond/later/higher, e.g. 6G) cellular radio/wireless network.
In some embodiments the network elements 102 are configured to operate a novel signaling procedure, which utilizes an extended NRT to provide extended information for resource management and orchestration. The extended neighbor relation table can comprise not only general entries of a general NRT (as defined by a relevant standard, such as 3GPP TS 32.511, 'Automatic Neighbour Relation (ANR) management; Concepts and requirements', V16.0.0, Jan. 2020 (where the NRT is called a 'neighbor cell relation table')), which are directly accessible by a certain type of network element, such as a gNB, but also other data/measurements, such as ones provided by a network operator or the like, which are generally not directly accessible through the gNB. The extended NRT can be a separate and different message from a standard NRT. The information contained in the extended NRT can be used by one or more network element to automatically perform improved resource management and/or orchestration, which can involve configuring at least one attribute of one or more local/remote network element, that is not possible using generally-configured network elements.
A gNB can have an Automatic Neighbor Relation (ANR) function that can be used for configuration. In some SONs there is an intelligent orchestration policy module in either the O&M function, the gNB, or both (i.e. centralized, distributed, or mixed), which will be responsible for managing/configuring resources. In embodiments instructions can be executed by such a module to process the information contained in the extended NRT to optimize functions such as load balancing and/or energy saving.
Figure 2 schematically illustrates the first network element 102A sending 202 a request to the second network element 102B for the extended NRT and the second network element responding 204 with its entries for the table. The first node can then update its extended NRT with the received entries. The extended NRT can also comprise data relating to the first network element (typically the same data fields as obtained for the second network element that the first network elements has obtained/generated in a similar manner) and/or data relating to at least one further network element that is communication with it. All or some of the data in the extended NRT may be processed (by the first network element and/or a remote processor/network element) to perform configuration of one or more of the network elements.
Figures 3a - 3c respectively illustrate three embodiments based on different self-organized architectures.
Figure 3a shows a gNB centralized architecture with an O&M orchestration policy decision-making module 302A that is external to the gNBs. In this type of architecture, the O&M policy module is in charge of providing the entries of the extended NRT for the different gNBs and can acquire extended NRTs from neighboring gNBs (only one gNB 303A and one extended NRT 304A are shown in the diagram for simplicity) via the ANR function. Once the extended NRTs are obtained from all the gNBs, the O&M orchestration policy module can make resource configuration/management decisions based on the data.
Figure 3b shows an inter gNB architecture, where each gNB has an individual/internal orchestration policy decision-making module 302B (only one gNB 303B and one extended NRT 304B are shown in the diagram for simplicity). In a distributed manner, the different gNBs exchange their information in order to fill theirs extended NRTs. Once all the individual extended NRTs are obtained, the orchestration policy module of each gNB maps the local extended NRT to resource orchestration actions.
Figure 3c shows a gNB and O&M mixed/hybrid architecture. The extended NRT data can be exchanged via both an external O&M policy decision-making module 302C and an intra gNB 305C. This type of architecture can follow a master-slaver decision structure (O&M being the master, and intra gNB being the slave, or vice versa), or a joint decision making architecture. Resource management decisions will be made by both the external module 302C and the internal policy modules 305C being executed by each gNB 303C (again, only one gNB 303C and one extended NRT 304C are shown in the diagram for simplicity).
The decision-making making module 302A - 302C/305C in any of the above embodiments may be implemented using AI or any other type of O&M policy module.
Figure 4 shows an example of an extended NRT. The extended NRT is an extension of an ordinary/general NRT. It includes information which is not accessible by cells/nodes, but can be provided by the O&M function of the SON. The example comprises traditional O&M NRT fields 402, 404, 406, i.e., IDs of neighbors (e.g., 1, 2, 3,...), neighbor relation attributes No HO, No X2. It also includes a set of non-general parameters in the form of extended attributes 408, 410, 412, which, in the example, comprise of data/entries that can be obtained by network elements/cells, e.g. load and area throughput, as well as data/entries that cannot be directly measured by network elements/cells. Examples include power parameters, mainly of the Remote Unit, but can include a power parameter of the base station or other data unit parameters. One example method of modelling power consumption is a linear model, i.e. Pwr = Slope*Load + Intercept. The Slope and Intercept values can be obtained via measurement by operators or vendors. These values also rely on how the site is deployed and therefore are not directly accessible in the site.
For instance, the load and throughput can be measured periodically by the network element. The power parameters may be obtained using a power model of the amplifier of the network element/base station, e.g. computing an indication of how much power will be required to fulfil a certain percentage of the maximum (100%) of the cell's load. In some cases, the network operator/network management function may send a command to a network element to read the element/node's statistics. The network element can reply with its statistics and the network operator/network management function can send a response comprising the node/element's properties. These additional data can be advantageously used for configuring network elements in accordance with the examples given herein. It will also be understood that in alternative embodiments the extended NRT can comprise additional or alternative data/entries that can be used for different types of configuration purposes, such as service enhancement. For instance, there is a function in LTE called Access Class Barring (ACB). Its function is to bar access of UEs from certain services, in order to maintain KPIs of the network. For example, if the cell is overloaded with emergency services, or another high-priority service, then the cell can bar UEs of other non-emergency services so that the cell can reserve all resources for those emergencies. A similar situation can occur in small cells when resources are limited. Thus, in some embodiments, base stations can exchange information indicating which classes of UEs (e.g. non-emergency/low priority) are barred in each cell. Then, when performing handover and/or load balancing, the source base station/cell can avoid handing over UEs of these classes to the target base station/cell.
The area throughput data/entry in the example extended NRT of Figure 4 can be used for performing user handover between cells in an advantageous manner. A network element may logically divide its/each cell coverage into a number of areas. Examples are illustrated in Figures 5a and 5b. In the example of Figure 5a there are four generally radial areas/segments 502A - 502D arranged around the base station/gNB 500. In the example of Figure 5b there are three generally concentric areas/rings 504A - 504C arranged around the gNB 500. It will be understood that other numbers and arrangements of areas can be used by alternative embodiments. The number and type of logical areas may be determined by software or a user upon configuration or deployment, for example.
In some embodiments each of the network elements can maintain a database, or other data structure, for storing the (normalized) throughput and location information of its areas. An example of the area throughput information is presented as a table below:
Example of area throughput information
Figure PCTKR2021001500-appb-img-000002
Each area can be given an identifier. The throughput of an area can be computed using the throughput of users within the area. One way to do this is by computing the average throughput of users. Also, the throughput can be normalized in terms of the number of layers (data streams in embodiments that use multiple-input, multiple-output (MIMO)). The geographical coordinate of the center of the area can also be included in the data. With this information, a cell/gNB can estimate the user throughput after migrating a user to another cell, if provided the user's location information.In some embodiments the first network element 102A will send a message to the second network element 102B to request its area throughput information. The second network element will respond with its area throughput information. The area throughput information can be obtained by estimating the throughput of users/devices whose locations are within an area.
One or more of the network elements may be provided with a post-handover load estimator (e.g. as a software module). The estimator will use the users' information to estimate the users' load if they are handed over to another cell/node/network element. The estimator 600 is schematically illustrated in Figure 6. The information used may include the pre-handover load 602, the current cell area throughput 604, the target cell area throughput 606, and location information of the user 608.
At the network element that provides the current serving cell, location information can be obtained via the Location Management Function (LMF). Then, using the extended NRT, the serving cell can determine in which area of the target cell the user will be located after the potential handover (e.g. by determining which of the areas of the target cell is closest, in terms of distance/location, to the area in which the user device is currently located) and also estimate the throughput after handover. A goal can be to estimate the post-handover load.
The post-handover load can be estimated using information including the current load, the current cell area throughput and the target cell area throughput. An example of post handover load estimation will now be described, where the impact of moving a user currently in Cell 1 to area 2 of Cell 2 is computed. Assuming that the user in Cell 1 has causes load L 1 and is located in area 1. The throughput of area 1 of Cell 1 is T 1. Let the throughput of area 2 of Cell2 be T 2. Then, the estimated post handover load
Figure PCTKR2021001500-appb-img-000003
in Cell2 can be expressed as:
Figure PCTKR2021001500-appb-img-000004
With the estimated load and the power parameters in the extended NRT, the energy consumption before and after user handover can also be estimated. As a result, the module can decide whether the action of user handover can reduce energy consumption. For instance, after estimating the load before and the load after potential handover, the module can know the power consumption before and the power consumption after the potential handover and can then decide whether or not to perform the potential handover, e.g. only perform the handover if the power consumption after the potential handover would be reduced compared to the power consumption before the potential handover. Once the power parameters of the various cells are known from the data in the extended NRT, the decision making module can derive the power consumptions of different cells under different loads. Then, after the decision making module estimates the post-handover loads of cells, it can compute a handover allocation such that power consumption is minimised/reduced.
Another usage of post handover load estimation is to avoid a ping-pong effect during load balancing. Without the post handover load estimation, a cell may handover a user to another cell in order to reduce load imbalance. The migrated user can impose a load on the target cell such that load imbalance increases. In this case, this user has to be handed over back to the original cell and the ping-pong effect occurs. With the estimated post handover load, a cell can estimate the imbalance in advance to avoid this ping-pong effect. The imbalance can be estimated, for example, by comparing the difference between the highest load and the lowest load of different cells. To give an example, CellA has load 20% and CellB had load 10%. Load balancing may be triggered if the highest load and the lowest load differ by a value larger than or equal to 10%, for instance. Thus, the load balancing would be trigged and CellA will seek to handover one or more users to CellB. Without knowing post-handover loads of CellB, after handover, CellA's load reduces to 12%, but the new load of CellB becomes 22%. The 8% drop in CellA may cause a high load increase in CellB because channel conditions are different. Now, load balancing will be triggered again (immediately after the previous load-balancing operation has been completed) and CellB moves users back to CellA and we go back to the origin and repeat. However, if the estimated post-handover loads of CellB are known at the decision module the decision-making module will not trigger this load balancing in the first place.
In some embodiments an intelligent decision-making module (e.g. a software module) can be embedded in each network element which keeps track of historical data and network events, and can select resource orchestration and management policies accordingly. The module can employ users' information, as well as network information, to infer future user behavior as well as network resources consumption and can determine MANO policies based on these. 
At every sampling period T, the module may obtain a snapshot of the current network and users' metrics, processes it, and store both the metrics and the processed data locally. The current user information includes, but it is not limited to, physical layer metrics (e.g. channel estimation, throughput), users' and cell load, and the extended NRT. Based on these, the module can undertake MANO policies, which can range from energy-saving solutions, to load balancing and/or service enhancement.
The example shown in Figure 7 schematically illustrates how the decision-making module 700 can be used to determine optimal energy solution policies. At the beginning of each decision period, the module obtains the current cell load status 702 as well as the actual user devices' throughputs 704 (e.g. channel, Physical Resource Block usage). Furthermore, the module has recorded the historical values of these metrics over previous decision periods, e.g. the last 10 decision periods. Based on the metrics trends, using traditional forecasting algorithms, or more advanced machine learning solutions, the module can decide that for the expected future resource demand, a multi-cell covered area can reduce the number of active cells 706 from x to y, where y < x, by selectively activating or de-activating the relevant network elements. Thus, the overall network power consumption is reduced while the users' and networks' services are not disrupted.
In some embodiments an ML algorithm, such as deep reinforcement learning can be used by the decision-making module 700. Examples of the inputs and outputs of the ML algorithm are illustrated in Figure 7. This learning process can include (or be similar to) giving several input samples to a deep neural network (DNN) so that the DNN generates different outputs (e.g. indicating which cell is on/off). At the same time, the DNN can be told that each action it takes has a certain reward; for example, a high reward when consuming less power and vice versa, provided that KPIs are met. After numerous iterations, the DNN will stabilize and converge to provide the optimal action for each input. If the current condition matches a historical condition then the cells will be activated/deactivated as previously done. Therefore, in principle, it is possible to can build a table storing the preferred action for each condition. However, when there are too many conditions and actions, the table will be too large and in this case, a DNN can be used instead of a table.
Embodiments can be performed by means of instructions being executed by a processor of one or more of the network elements. A developer may make a design choice regarding how, and on which network element, the instructions are executed. It will also be appreciated that at least some of the steps described herein may be re-ordered or omitted. One or more additional steps may be performed in some cases. Further, although the steps are described as being performed in sequence, in alternative embodiments at least some of them may be performed concurrently and/or by different ones of the network elements, or even at other/remote computing devices or a cloud service. It will also be understood that embodiments can be implemented using any suitable software, programming language, data editors, etc, and may be represented/stored/processed using any suitable data structures and formats.
It is understood that according to an exemplary embodiment, a computer readable medium storing a computer program to operate a method according to the foregoing embodiments is provided.
Attention is directed to any papers and documents which are filed concurrently with or previous to this specification in connection with this application and which are open to public inspection with this specification, and the contents of all such papers and documents are incorporated herein by reference.
All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and/or all of the steps of any method or process so disclosed, may be combined in any combination, except combinations where at least some of such features and/or steps are mutually exclusive.
Each feature disclosed in this specification (including any accompanying claims, abstract and drawings) may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise. Thus, unless expressly stated otherwise, each feature disclosed is one example only of a generic series of equivalent or similar features.
The disclosure is not restricted to the details of the foregoing embodiment(s). The disclosure extends to any novel one, or any novel combination, of the features disclosed in this specification (including any accompanying claims, abstract and drawings), or to any novel one, or any novel combination, of the steps of any method or process so disclosed.
The techniques described herein may be implemented using any suitably configured apparatus and/or system. Such an apparatus and/or system may be configured to perform a method according to any aspect, embodiment, example or claim disclosed herein. Such an apparatus may comprise one or more elements, for example one or more of receivers, transmitters, transceivers, processors, controllers, modules, units, and the like, each element configured to perform one or more corresponding processes, operations and/or method steps for implementing the techniques described herein. For example, an operation/function of X may be performed by a module configured to perform X (or an X-module). The one or more elements may be implemented in the form of hardware, software, or any combination of hardware and software.
It will be appreciated that examples of the disclosure may be implemented in the form of hardware, software or any combination of hardware and software. Any such software may be stored in the form of volatile or non-volatile storage, for example a storage device like a ROM, whether erasable or rewritable or not, or in the form of memory such as, for example, RAM, memory chips, device or integrated circuits or on an optically or magnetically readable medium such as, for example, a CD, DVD, magnetic disk or magnetic tape or the like.
It will be appreciated that the storage devices and storage media are embodiments of machine-readable storage that are suitable for storing a program or programs comprising instructions that, when executed, implement certain examples of the disclosure. Accordingly, certain example provide a program comprising code for implementing a method, apparatus or system according to any example, embodiment, aspect and/or claim disclosed herein, and/or a machine-readable storage storing such a program. Still further, such programs may be conveyed electronically via any medium, for example a communication signal carried over a wired or wireless connection.

Claims (15)

  1. A method of configuring a network element in a Self-organizing Network, SON, the method comprising:
    receiving, by a first network element from a second network element, at least one attribute data relating to the second network element included in a Neighbour Relation Table, NRT, field, and at least one additional attribute data relating to the second network element;
    updating by the first network element, an extended Neighbour Relation Table, NRT, wherein the extended NRT comprises the received at least one attribute data and the received at least one additional attribute data, and
    using the extended NRT to configure at least one of the first network element or the second network element.
  2. A method according to claim 1, wherein the at least one additional network element attribute data comprises at least one of data not included in the NRT field; a measurement provided by a network operator, or data obtained from an upper-level network management function of the SON.
  3. A method according to claim 1, wherein the extended NRT is used by an Automatic Neighbour Relation, ANR, function of the SON.
  4. A method according to claim 1 wherein the at least one additional attribute data comprises at least one of a power parameter relating to the second network element, a load of the second network element, throughput of at least part of a cell provided by the second network element, or one or more barred access class of a cell provided by the second network element.
  5. A method according to claim 4, wherein the step of using the extended NRT to configure at least one of the first network element or the second network element comprises determining whether or not to handover a user device serviced by a cell provided by the second network element to a cell provided by the first network element depending on the at least one additional attribute data.
  6. A method according to claim 5, wherein the step of using the extended NRT to configure at least one of the first network element or the second network element comprises:
    estimating a post-handover load of a cell provided by the first network element if at least one user device is handed over from a cell provided by the second network element to the cell provided by the first network element, and
    performing the handover of the at least one user device if it is estimated that the handover will reduce energy consumption of the second network element.
  7. A method according to claim 6, wherein cell coverage provided by at least one of the first network element or the second network element is logically divided into a plurality of areas and each the at least one network element stores data comprises user device throughput and location information for each of its plurality of areas, wherein the plurality of areas are arranged radially or concentrically around a physical location the network element.
  8. A method according to claim 7, comprising:
    obtaining location information of the at least one user device, and
    determining in which of the plurality of areas the at least one user device will be located following the handover.
  9. A method according to claim 7, wherein the user device throughput of an area is computed based on average throughput of all user devices currently within the area and wherein the throughput of the cell provided by the first network element is computed using a formula:
    Figure PCTKR2021001500-appb-img-000005
    where:
    L 1 represents a current load of a user device in the cell provided by the second network element;
    T 1 represents a throughput of the area in which the user device is currently located in the cell provided by the second network element, and
    T 2 represents a throughput of the determined area in the cell provided by the first network element where the user device will be located following the handover.
  10. A method according to claim 6, further comprising estimating energy consumption at the first network element before and/or after the handover based on the load and the power parameter in the extended NRT.
  11. A method according to claim 1, further comprising:
    storing historical data representing load status of a cell provided by at least one of the first network element or the second network element over a plurality of time periods and user devices throughput in the cell over the plurality of time periods;
    using the historical data to predict a number of active cells required to fulfil current or future user device demand in a multi-cell area covered by a plurality of network elements, and
    activating and/or de-activating the network elements based on the prediction.
  12. A method according to claim 1, further comprising:
    using the extended NRT to estimate load at the first network element after a handover and using the extended NRT to estimate load at the second network element after the handover, wherein the handover is part of a network element load-balancing operation that is triggered when load of at least one network element exceeds or meets a threshold;
    determining whether the estimated load at the first network element and/or the estimated load at the second network element would trigger a further handover due to exceeding or meeting the threshold, and
    not performing the handover if it is determined that the estimated load would trigger the further handover.
  13. A method according to claim 1, wherein the at least one additional attribute data comprises data representing a class of service of a user device serviced by a cell provided by the second network element, and
    the step of using the extended NRT to configure the at least one of the first network element or the second network element comprises determining whether or not to handover the user device serviced by the cell provided by the second network element to a cell provided by the first network element depending on the class of service of the user device.
  14. A method of configuring a network element in a Self-organizing Network, SON, the method comprising:
    at a second network element, obtaining data representing at least one attribute of the second network element included in a Neighbour Relation Table, NRT, field and obtaining data representing at least one additional attribute of the second network element, and
    transmitting, by the second network element to a first network element, the at least one attribute data and the at least one additional attribute data for use in an enhanced NRT for configuring at least one of the first network element or the second network element.
  15. A network element of a Self-organizing Network, SON, the network element comprising:
    a communications interface configured to communicate with at least one other network element, and
    a processor configured to:
    receive, from another network element, at least one attribute data relating to the other network element included in a Neighbour Relation Table, NRT, field, and at least one additional attribute data relating to the other network element;
    update an extended Neighbour Relation Table, NRT, wherein the extended NRT comprises the received at least one attribute data and the received at least one additional attribute data; and
    use the extended NRT to configure at least one of the network element or the other network element.
PCT/KR2021/001500 2020-08-06 2021-02-04 Configuring resources in a self-organizing network WO2022030713A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
GB2012225.5 2020-08-06
GB2012225.5A GB2597931A (en) 2020-08-06 2020-08-06 Configuring resources in a self-organizing network

Publications (1)

Publication Number Publication Date
WO2022030713A1 true WO2022030713A1 (en) 2022-02-10

Family

ID=72425252

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2021/001500 WO2022030713A1 (en) 2020-08-06 2021-02-04 Configuring resources in a self-organizing network

Country Status (2)

Country Link
GB (1) GB2597931A (en)
WO (1) WO2022030713A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2023184441A1 (en) * 2022-03-31 2023-10-05 Oppo广东移动通信有限公司 Primary cell configuration or reconfiguration method and apparatus, device, and medium

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP4346285A1 (en) * 2022-09-23 2024-04-03 Comcast Cable Communications LLC Energy efficiency in radio access network

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170070896A1 (en) * 2014-03-18 2017-03-09 Nec Corporation Control apparatus, base station apparatus, radio terminal, and method for updating neighbour relation table
US20190132777A1 (en) * 2017-05-04 2019-05-02 Ofinno Technologies, Llc Beam-Based Neighbor Relation Information

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2472792A (en) * 2009-08-17 2011-02-23 Nec Corp Measurement reporting in a mobile communications system
US9288690B2 (en) * 2010-05-26 2016-03-15 Qualcomm Incorporated Apparatus for clustering cells using neighbor relations
EP2575391B1 (en) * 2011-09-30 2016-07-20 Telefonaktiebolaget LM Ericsson (publ) Neighbor cell selection based on cell access mode for X2 based handover in a E-UTRAN
US10368253B2 (en) * 2017-07-25 2019-07-30 At&T Intellectual Property I, L.P. System and method for managing dual connectivity with dynamic anchor cell selection
WO2019096399A1 (en) * 2017-11-17 2019-05-23 Nokia Technologies Oy Cell relations optimization

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170070896A1 (en) * 2014-03-18 2017-03-09 Nec Corporation Control apparatus, base station apparatus, radio terminal, and method for updating neighbour relation table
US20190132777A1 (en) * 2017-05-04 2019-05-02 Ofinno Technologies, Llc Beam-Based Neighbor Relation Information

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
ERICSSON: "Correction for Network Sharing - MOCN", 3GPP DRAFT; R3-202099, 3RD GENERATION PARTNERSHIP PROJECT (3GPP), MOBILE COMPETENCE CENTRE ; 650, ROUTE DES LUCIOLES ; F-06921 SOPHIA-ANTIPOLIS CEDEX ; FRANCE, vol. RAN WG3, no. Online Meeting ;20200420 - 20200430, 9 April 2020 (2020-04-09), Mobile Competence Centre ; 650, route des Lucioles ; F-06921 Sophia-Antipolis Cedex ; France , XP051870588 *
LG UPLUS, LG ELECTRONICS: "ANR enhancement for coexistence of SA and NSA deployment", 3GPP DRAFT; R3-203144, 3RD GENERATION PARTNERSHIP PROJECT (3GPP), MOBILE COMPETENCE CENTRE ; 650, ROUTE DES LUCIOLES ; F-06921 SOPHIA-ANTIPOLIS CEDEX ; FRANCE, vol. RAN WG3, no. E-Meeting; 20200601 - 20200612, 21 May 2020 (2020-05-21), Mobile Competence Centre ; 650, route des Lucioles ; F-06921 Sophia-Antipolis Cedex ; France , XP051887628 *
LG UPLUS: "ANR enhancement for coexistence of SA and NSA deployment", 3GPP DRAFT; R3-203168, 3RD GENERATION PARTNERSHIP PROJECT (3GPP), MOBILE COMPETENCE CENTRE ; 650, ROUTE DES LUCIOLES ; F-06921 SOPHIA-ANTIPOLIS CEDEX ; FRANCE, vol. RAN WG3, no. E-Meeting; 20200601 - 20200612, 20 May 2020 (2020-05-20), Mobile Competence Centre ; 650, route des Lucioles ; F-06921 Sophia-Antipolis Cedex ; France , XP051887206 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2023184441A1 (en) * 2022-03-31 2023-10-05 Oppo广东移动通信有限公司 Primary cell configuration or reconfiguration method and apparatus, device, and medium

Also Published As

Publication number Publication date
GB2597931A (en) 2022-02-16
GB202012225D0 (en) 2020-09-16

Similar Documents

Publication Publication Date Title
Fourati et al. Comprehensive survey on self-organizing cellular network approaches applied to 5G networks
CN105308999B (en) A kind of method and network controller of sensitive wireless access network
CN105052191B (en) The method and system of management based on cloud for ad hoc deployed wireless networks
US20240040461A1 (en) Ue, network node and methods for handling mobility information in a communications network
WO2015131677A1 (en) Method and device for constructing virtual cell and selecting cooperative node
CN103650571B (en) Performing measurements in a digital cellular wireless telecommunication network
CN104137595A (en) Self-organizing network function interaction
CN104704869A (en) Mobility robustness optimization based on reference signal strength maps
WO2022030713A1 (en) Configuring resources in a self-organizing network
CN114402654A (en) Apparatus for radio access network data collection
JP2014523159A5 (en)
CN104521270A (en) Self organizing network operation diagnosis function
US11799733B2 (en) Energy usage in a communications network
EP3530069B1 (en) System and method for scalable radio network slicing
EP3432629A1 (en) Partitioning method and apparatus for partitioning a plurality of wireless access points into management clusters
US20160337878A1 (en) Improving network efficiency
Gucciardo et al. A flexible 4G/5G control platform for fingerprint-based indoor localization
WO2018184667A1 (en) Apparatus and method for performing network optimization
Aguilar-Garcia et al. Location-aware self-organizing methods in femtocell networks
Ferrús et al. Data analytics architectural framework for smarter radio resource management in 5G radio access networks
Fourati et al. Self-organizing cellular network approaches applied to 5G networks
CN104412641A (en) Signature enabler for multi-vendor SON coordination
JP2022105306A (en) Method and apparatus for initiating handover (ho) procedure in open-radio access network (o-ran) environment
US11337131B1 (en) Method and apparatus for recommending real-time handover to a target cell in open-radio access network (O-RAN) environment
CN106572475A (en) Access node management method, access network management entity, equipment and access nodes

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21852589

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21852589

Country of ref document: EP

Kind code of ref document: A1