Abstract: According to an aspect, a data inversion circuit configured to perform DBI-AC encoding using a PAM 4 signal may comprise a data generation unit configured to generate input data based on the PAM 4 signal, a channel comprising N data lines, a first auxiliary signal generation unit configured to generate a first auxiliary signal that determines whether to perform a first encoding on the input data based on the number of each of a plurality of data symbols included in the input data, a first data encoding unit configured to generate intermediate data by performing the first encoding on the input data based on the first auxiliary signal, a second auxiliary signal generation unit configured to generate a second auxiliary signal that determines whether to perform a third encoding on the intermediate data by analyzing the relationship between a plurality of data symbols at a current time point and a plurality of data symbols at a previous time point included in the intermediate data and a second data encoding unit c

Abstract: Disclosed is a method to demodulate messages according to two different modulation schemes in 5G and 6G, and thereby identifying which message elements are likely faulted. The two modulation schemes are QAM in which the signal is a sum of two orthogonal amplitude-modulated “branch” signals, and classical amplitude-phase modulation in which each message element's raw signal is both amplitude and phase modulated. The two schemes have similar information density but different noise sensitivities. Therefore, a receiver can compare the demodulated message using one modulation scheme to the same message demodulated according to the other modulation scheme, and flag any message elements that demodulate differently. In addition, one modulation scheme may be more effective than the other depending on conditions.

Abstract: A user equipment is provided with puncturing information indicating a resource to which downlink data is punctured among time-frequency resources to which the downlink data is allocated. The user equipment may decode the downlink data received in the time-frequency resource on the basis of the puncturing information. The downlink data may be mapped to the time-frequency resource by a combined method of a time-first resource mapping method and a frequency-first resource mapping method, or by a distributed resource mapping method.

Abstract: This application provides a reference signal generation method, a reference signal detection method, and a communications apparatus, so that a terminal device or a network device can generate a reference signal by using a pseudo-random sequence initial factor cinit provided in embodiments of this application. Compared with a solution in the current technology, the generation manner can support a relatively large quantity of reference signal sequences, to better meet requirements of a plurality of 5G scenarios. The method may include: obtaining a reference signal sequence based on a pseudo-random sequence initial factor cinit; and mapping the sequence to one or more OFDM symbols, where the pseudo-random sequence initial factor cinit is related to a parameter d, d max(log2(nID,max+1)?10,0) or d=max(log2 (nID,max+1)?12,0), max represents that a larger value is selected from two values, and nID,max represents a maximum value of a reference signal sequence ID.

Abstract: Transmission quality is improved in an environment in which direct waves dominate in a transmission method for transmitting a plurality of modulated signals from a plurality of antennas at the same time. All data symbols used in data transmission of a modulated signal are precoded by hopping between precoding matrices so that the precoding matrix used to precode each data symbol and the precoding matrices used to precode data symbols that are adjacent to the data symbol in the frequency domain and the time domain all differ. A modulated signal with such data symbols arranged therein is transmitted.

Abstract: Presented are a method and a device for receiving a PPDU in a wireless LAN system. Particularly, a receiving STA receives a PPDU from a transmitting STA through a broadband, and decodes the PPDU. The PPDU includes a legacy preamble and first and second signal fields. The legacy preamble and the first and second signal fields are generated on the basis of a first phase rotation value. The first phase rotation value is acquired on the basis of a first preamble puncturing pattern of the broadband. The first preamble puncturing pattern includes the pattern in which 40 MHz or 80 MHz band is punctured in the broadband, when the broadband is a 320 MHz band. The first phase rotation value is [1 ?1 ?1 ?1 1 ?1 ?1 ?1 ?1 1 1 1 ?1 1 1 1].

Abstract: Certain aspects of the present disclosure provide techniques for time division multiplexing (TDMing) user equipment (UE) data using guard interval (GI)-based waveforms. An example method generally includes communicating with a device using a frame format, the frame format comprising a first symbol, a second symbol, and a third symbol, wherein: the first symbol comprises a first data portion for a first UE and a first GI corresponding to the first UE; the third symbol comprises a second data portion for a second UE and a second GI corresponding to the second UE, wherein the third symbol is later in time than the first symbol; and the second symbol comprises a gap or a reference signal (RS), a third GI corresponding to the first UE, and a fourth GI corresponding to the second UE, wherein the second symbol is between the first and third symbol in time.

Abstract: A UE in a wireless communication system is disclosed. The UE includes a communication module and a processor for controlling the communication module. The processor determines a first cyclic shift (CS) value based on hybrid automatic repeat request acknowledgment (HARQ-ACK) information representing a response to a downlink channel having been received from a base station, determines a CS offset based on request information representing a request to be transmitted from the UE to the base station, determines a second CS value representing a degree of cyclic-shifting a base sequence to be used in a physical uplink control channel (PUCCH) based on the first CS value and the CS offset, and transmits the PUCCH for simultaneous transmission of the request information and the HARQ-ACK information using a sequence generated by cyclic-shifting the base sequence based on the second CS value.

Abstract: A communication method of a terminal with another terminal through a sidelink comprises: determining, by a first user device, one or more radio resource control (RRC) configurations comprising information of a demodulation reference signal (DMRS) for vehicle-to-everything (V2X) communication between user devices; determining, based on the one or more RRC configurations, a Physical Sidelink Control Channel (PSCCH) DMRS for a PSCCH; determining, based on a cyclic redundancy code of the PSCCH, an identifier; determining, based on the identifier, an initialization value associated with a Physical Sidelink Shared Channel (PSSCH) DMRS for a PSSCH; and via one or more orthogonal frequency division multiplexing (OFDM) symbols, transmitting, by the first user device to a second user device, the PSCCH DMRS and the PSSCH DMRS.

Abstract: A method for reference signal generation and a communications device are provided. The method includes: in a case that indication information is received from a second communications device, performing first modulation and then discrete Fourier transform modulation on at least one of target-reference-signal sequences to generate a target reference signal, where the first modulation includes any one of the following: ?/2-BPSK modulation and 8-PSK modulation, and the indication information is used to indicate that a first communications device uses the target reference signal for transmission.

Abstract: Methods, systems, and devices for wireless communications are described. A transmitting device may encode a set of bits to transmit to a receiving device based on a repetition factor. The transmitting device may map, based on the repetition factor, the set of encoded bits to a subset of subcarriers such as adjacent subcarriers of a set of subcarriers. The transmitting device may generate a signal including the set of encoded bits based on the mapping, and transmit the generated signal to the receiving device. The receiving device may receive a modulated signal from the transmitting device, and identify, based on a repetition factor, a subset of subcarriers including adjacent subcarriers of a set of subcarriers associated with the modulated signal. The receiving device may average the subset of subcarriers including the adjacent subcarriers, and demodulate the modulated signal in accordance with the averaged subset of subcarriers including the adjacent subcarriers.

Abstract: According to an embodiment, an operation method of a terminal in a wireless communication system comprises the steps of: generating a first waveform from a signal related to a physical sidelink control channel (PSCCH) and generating a second waveform from a signal related to a physical sidelink shared channel (PSSCH); generating a third waveform on the basis of a peak to average power ratio (PAPR) and a constellation shift and transmitting the second waveform and the third waveform, wherein the constellation shift is applied only to the first waveform.

Abstract: A terminal and a base station in a wireless communication system and a method performed by the terminal and the base station are disclosed. According to an embodiment, the method performed by the terminal in the wireless communication system comprises: receiving a physical downlink control channel (PDCCH), the PDCCH including downlink control information (DCI) for scheduling one or more physical downlink shared channels (PDSCHs); receiving the PDSCHs according to the DCI; and transmitting a hybrid automatic repeat request acknowledgement (HARQ-ACK/NACK) codebook for the PDSCHs.

Abstract: A communication device including one or more processors configured to determine a first signal component of a received modulated signal; determine a second signal component of the received modulated signal; generate a phase shift of the first signal component; generate a phase shift of the second signal component; compare the phase shift of the first signal component and the phase shift of the second signal component with each other; and generate a plurality of constellation points, wherein each of the plurality of constellation point is based on the determination of the first signal component, the determination of the second signal component, and the comparison.

Abstract: A computer-implemented method includes tasks performed by one or more processors in a management network: (a) receiving, from an end-user network, data associated with a plurality of agent instances servicing incoming communications for the contact center, wherein the contact center, the end-user network, and the management network are associated with the enterprise; (b) determining, based on a specification of a logical directive, an operation to be performed in relation to the plurality of agent instances, the specification of the logical directive having at least one condition that, if satisfied by the received data associated with the plurality of agent instances servicing the incoming communications for the contact center, defines the operation to be performed; and (c) causing the determined operation to be performed, wherein the operation includes sending a communication via the end-user network to at least one agent instance in the plurality of agent instances and/or to at least one supervisor instance

Abstract: Provided are a beam recovery method and apparatus, a terminal and a storage medium. The method includes: in a beam recovery process, in response to determining processing of a current stage in the beam recovery process being unsuitable to be implemented on an active bandwidth part (BWP), determining a target BWP of the current stage according to the active BWP and a switch mapping relationship determined for the current stage, switching to the target BWP to perform the processing of the current stage, and continuing to perform the beam recovery process until implementing a beam recovery.

Abstract: Generally described, systems and methods are provided for monitoring and detecting causes of failures of network paths. The system collects performance information from a plurality of nodes and links in a network, aggregates the collected performance information across paths in the network, processes the aggregated performance information for detecting failures on the paths, analyzes each of the detected failures to determine at least one root cause, and initiates a remedial workflow for the at least one root cause determined. In some aspects, processing the aggregated information may include performing a statistical regression analysis or otherwise solving a set of equations for the performance indications on each of a plurality of paths. In another aspect, the system may also include an interface which makes available for display one or more of the network topology, the collected and aggregated performance information, and indications of the detected failures in the topology.

Abstract: Techniques for connectivity issue remediation are provided. A first link trace message is transmitted from a source end point to a destination end point. A first topology graph for a network is generated based on the first link trace message, and a presence of a connectivity issue in the network is detected. A second link trace message is transmitted from the source end point to the destination end point. A second topology graph for the network is automatically generated based on second link trace message, and a component in the network that caused the connectivity issue is identified based on comparing the first and second topology graphs.

Abstract: Partners in an entity increasingly need to obtain information about data relating to communications of a terminal. A configuration method and a configuration device configure a terminal device able to be connected to a slice of a communication network. The slice is instantiated so as to route data between the terminal device and a service device in accordance with at least one routing criterion. The method is implemented by management equipment for managing the terminal device and able to communicate with the terminal device. The method includes storing a configuration context of the terminal device in memory, and transmitting, to the terminal device, a configuration message including the identifier of a proxy connection device and the identifier of the network slice.

Abstract: Disclosed is a server including a plurality of information processing modules so as to form a unified platform for an operating system deployed on the server, each module having an elementary software, BIOS, provided to use configuration data when the module is started up, the server also being provided to allow at least one configuration datum to be modified, and being characterized in that it further includes a management controller, BMC, provided to store configuration data in an associated memory, and to transmit, when the at least one configuration datum is modified, such data to management controllers associated with the other modules, and in that the elementary software is provided to recover the configuration data from the management controller when the module is started up.

Abstract: Embodiments relate to methods, systems, and computer program products for path management in a processing system. In a method, in response to receiving a request for adding a target controlling unit into a processing system, a plurality of network nodes in the processing system are divided into a group of subnets based on a topology of the plurality of network nodes, the plurality of network nodes being connected to at least one controlling unit in the processing system. A workload estimation is determined, the workload estimation representing a workload to be caused by the target controlling unit to the processing system. A target subnet is selected from the group of subnets for connecting the target controlling unit into the processing system based on the workload estimation. With these embodiments, the target subnet may be selected in an automatic way such that the performance of the processing system may be increased.

Abstract: An apparatus comprises a memory and a processor communicatively coupled to one another. The memory may be configured to store multiple configuration commands. The processor may be configured to establish a communication session between a first network component and a second network component based at least in part upon the configuration commands. Further, the processor may be configured to determine whether the communication session between the first network component and the second network component is interrupted, generate multiple space backup configuration commands indicating to back up the configuration commands to a space server in response to determining that the communication session is interrupted, transfer the space backup configuration commands to the space server, and reestablish the communication session between the first network component and the second network component via the space server based at least in part upon the space backup configuration commands.

Abstract: Disclosed is a cloud services exchange platform that provides for multi-tenant orchestration with a flexible underlay facilitating tenant move with minimal disruption. Multi-tenant node engines are adapted to orchestrate the instantiation, hosting, and/or provisioning of services to one or more endpoints on behalf of a customer. Multiple dataplanes are associated with the cloud services exchange platform.

Abstract: A method of managing a network of connected network devices can include detecting that a device descriptor file for a first network device has been updated and stored to a data store managed by a peer-to-peer network of computing systems; analyzing the updated device descriptor file stored to the data store; analyzing historical information stored in the data store regarding other device descriptor file updates related to the first network device or related to other network devices; performing predictive analytics processing on the results of the analyzing and generating a result; and storing the result of the performing predictive analytics processing to the data store.

Abstract: A sensor device for transmitting a sensor parameterization between sensors. The sensor device includes a first sensor with a sensor parameterization and a second sensor. The first sensor and the second sensor communicate with each other via a communication link. The first sensor sends the sensor parameterization to the second sensor via the communication link. The second sensor receives and accepts the sensor parameterization. The second sensor further provides the first sensor with the energy required to transmit the sensor parameterization via the communication link.

Abstract: Some embodiments provide a method for automatically configuring VPN gateways. The method receives a first configuration for a first VPN gateway located at a first datacenter. The configuration includes configuration data for a first set of VPNs connecting a first set of networks at the first datacenter to other networks at other datacenters. The method automatically modifies the configuration data to generate a second configuration for a second VPN gateway. The method configures the second VPN gateway using the second configuration to setup a second set of VPNs connecting a second set of networks to the other networks at the other datacenters.

Abstract: A communication control system includes a memory and a processor coupled to the memory. The processor is configured to according to an instruction from an external device, perform processing related to changing a network setting on a communication device, after the processing has been executed, determine whether or not communication of the communication device matches a specific communication condition, and notify an error to the external device in a case in which determination has been made that the communication of the communication device does not match the specific communication condition.

Abstract: A data processing method and device, a storage medium and an electronic device are disclosed. The method comprises responsive to determining that an application instance is established, generating, by an orchestrator, a data acquisition request message corresponding to the application instance and sending the data acquisition request message to a next communication node, wherein the next communication node includes one of a network control device and a target network node; acquiring, by the target network node, data according to the data acquisition request message received from the orchestrator or through the network control device, and sending a data acquisition result to a processing device of the application instance; and processing, by the processing device of the application instance, the data acquisition result through a model in the application instance and outputting a processing result.

Abstract: The disclosed technique includes establishing consensus among devices to share performance metrics data of those devices. An example of the devices includes endpoint devices that are owned or managed by a common entity. The devices form a consensus-based network. The devices are each operable to independently connect to an external network. A manager device of the devices dynamically selects which of the devices will operate like a network access device to route traffic between the devices and the external network. The network access device is selected by the manager device based on the performance metrics data to optimize a global performance metric of the consensus-based network.

Abstract: A method for generating a network topology for multiple wireless intelligent devices of a single distributed local network is disclosed. The method includes determining an optimal network topology for the multiple wireless intelligent devices of the single distributed local network. The method further includes generating a plurality of commands to be transmitted to one or more of multiple wireless intelligent devices of the determined optimal network topology. Further, the method includes instructing the plurality of commands to connect the one or more of multiple wireless intelligent devices to generate the determined optimal network topology.

Abstract: The communication network system according to the present disclosure comprises: a life cycle manager (LCM) that transmits a request for virtual network function (VNF) setting to one or more clouds; and a cloud adapter that mediates communication between the LCM and the cloud to which the request is addressed, The cloud adapter creates, based on a response to the request from the cloud to which the request is addressed, a common-format management ID independent of the type of the cloud to which the request is addressed, and transmits the management ID to the LCM. The LCM may transmit a query regarding a processing status of the request to the cloud adapter, including the management ID. The cloud adapter may recognize the cloud to which the query is addressed on the basis of the management ID and query the cloud to which the query is addressed such that the cloud returns the processing status.

Abstract: Systems, methods, and related technologies for classifying a device on a network are described. A method includes capturing device information corresponding to a device on a network. The method inputs unstructured crowdsourced data on the network into a machine learning model to produce structured crowdsourced data. The method classifies the device based on evaluating the device information with the structured crowdsourced data.

Abstract: The disclosed systems, structures, and methods are directed to risk assessment of an optical network. A simulation framework includes a risk map engine including a performance prediction engine that generates a simulation of the optical network based at least in part on an input network topology and/or service map. The prediction performance engine runs the simulation to predict, based at least in part on received network telemetry data, direct and indirect impacts on the optical network of a risk factor represented in a what-if scenario. The risk map engine includes a risk assessment engine that determines a risk associated with the risk factor based at least in part on the predicted direct and indirect impacts and on a likelihood of occurrence of the risk factor. The risk assessment engine generates a risk map showing aggregate risks to the optical network from a plurality of risk factors.

Abstract: A method of network viewport data fetching includes receiving a request at a mobile device to launch a network visualization application; causing a GUI to display of a viewport corresponding to a first geographic area and over the first geographic area one or more shapes; determining a latitude and longitude of a first and a second endpoint; determining a distance of a diagonal; determining a midpoint of the diagonal; in response to a distance between the midpoint and either the first or second endpoint being less than or equal to a threshold distance, updating to a second geographic area; or in response to the distance between the midpoint and either the first or second endpoint being greater than the threshold distance, updating the display of the viewport corresponding to a third geographic area and updating the one or more shapes that display the indication of network coverage quality.

Abstract: Methods and apparatus to scale in and/or scale out arbitrary resources managed by a cloud automation system are disclosed. An example apparatus includes processor circuitry; and a non-transitory computer readable medium comprising instructions which, when executed, cause the processor circuitry to: in response to an indication to scale a first component of an application to be deployed: determine an execution plan to scale the first component based on a dependency graph corresponding to a dependency within a blueprint specifying a logical topology of the application; perform a custom action to scale the first component, the custom action identified in a scaling parameter associated with the application; and update operation of a second component based on scaling the first component, the second component dependent on the first component, the update to enable the second component to interact with the first component after the scaling.

Abstract: The present disclosure is directed toward a device and a method for evaluating a wireless link established between an access point and a user equipment. The device and method include determining a level of Quality of Experience of an application intended to be run on the user equipment using a mapping between a parameter representative of the QoE of the application under different wireless transmission conditions and sets of parameters representative of said different transmission conditions of the wireless link.

Abstract: A device may receive, from Internet of Things (IOT) devices, telemetry data in a variety of formats and may convert the telemetry data to converted telemetry data in a first unified format. The device may receive, from one or more of the IoT devices, offline IoT data in a variety of formats and may convert the offline IoT data to converted offline IoT data in a second unified format. The device may receive events from one or more of the IoT devices and may identify anomalies based on the telemetry data and in near-real time relative to receiving the telemetry data. The device may store the converted telemetry data, the converted offline IoT data, the events, and the anomalies, as stored data, in a data structure, and may perform actions based on the stored data.

Abstract: Some embodiments are directed to a data center selection device for cloud-based application deployment. The device may send a request for test data from multiple test applications at multiple data centers and determine separate network statistics for the multiple data centers. A data center may be selected based at least in part on the measured network statistics, where an application may be deployed.

Abstract: In a bandwidth estimation device, a measurement unit measures a measurement time from reception by a receiver of a first packet of a packet train including a plurality of packets that are sequentially transmitted from a transmitter via a communication line to reception by the receiver of a last packet of the packet train, An estimation validity determination unit compares the measurement time with a threshold time predefined based on performance of the receiver or characteristics of the communication line to determine whether a valid estimate of an available bandwidth of the communication line can be made. In response to the estimation validity determination unit determining that a valid estimate of the available bandwidth can be made, an estimation unit estimates the available bandwidth of the communication line based on an amount of data in the packet train and the measurement time.

Abstract: Examples are disclosed herein relating to a method performed by a remote capture server, of tracking a media slice through at least part of a communications network comprising a plurality of network elements, the method comprising: inspecting a first payload of a first data packet observed at an egress capture point of a first network element, inspecting a second payload of a second data packet observed at an ingress capture point of a second network element, comparing the first payload with the second payload and associating the first data packet with the second data packet in response to determining that the first payload and the second payload are the same, wherein the first and second data packets correspond to a first media slice in response to determining they are associated.

Abstract: A method includes receiving, by a computing system, a declarative testing descriptor for active testing of a network slice implemented by first virtualized services in a network; receiving, by the computing system and from an orchestration layer, metadata associated with the network slice; determining, by the computing system and based on the declarative testing description and the metadata, active testing configuration for testing the network slice; and starting an active test according to the active testing configuration and determining service level violations based on a result of the active test.

Abstract: A machine selects a candidate data service to be tested for service quality. The machine performs a first-phase test of the candidate data service by providing predetermined input data to the candidate data service and measuring a first quality metric that results. The machine may determine that the candidate data service is to undergo the second-phase test based on the first quality metric, and the machine may accordingly perform the second-phase test of the candidate data service. The second-phase test may include: adding the candidate data service to a pool of five data services processing undetermined input data; routing a portion of the undetermined input data to the candidate data service; and measuring a second quality metric resultant from the candidate data service processing the routed portion. Based on the second quality metric, the machine may record whether the candidate data service is to remain in the pool.

Abstract: A system and a method for eliminating data loss in a virtually aggregated network are described. A first network device may identify inactivity of a central network device present in a communication network. The central network device is responsible for distributing routing information between a plurality of network devices including the first network device. The first network device delays network route calculations until a second network device is elected from the plurality of network devices to perform functions of the central network device. The second network device generates link state information related to the plurality of network devices, and shares the link state information with the plurality of network devices. Upon receiving the link state information, the plurality of network device may resume the route calculations.

Abstract: Device, method, medium and computer program product for voice data transmission. The device includes a processor that executes instructions in memory to: establish a call connection for voice data transmission with another device, wherein the said device parses and identifies the number of the said another device to determine whether the voice data transmission is associated with an international call, and when the voice data transmission is determined to be associated with an international call, the determine hop limit value for the voice data transmission as a first value greater than the default hop limit value; or when the voice data transmission is determined to not be associated with an international call, the hop limit value for the voice data transmission is determined as the default value; and generate packets for sending voice data based on the hop limit value determined when the call connection is established.

Abstract: An example branch gateway (BG) of a local area network (LAN) includes processing circuitry and a memory. The memory includes instructions that cause the BG to determine that a network traffic load of the BG is greater than that of another BG. The instructions further cause the BG to determine that a first network traffic load skew between the BG and the other BG is greater than a first threshold. The instructions further cause the BG to select a first virtual local area network (VLAN) of a set of VLANs of the LAN, wherein network traffic of client devices of the first VLAN is routed through the BG based on a master role of the first VLAN being assigned to the BG. The instructions further cause the BG to transmit a message to the other BG that assigns the master role of the first VLAN to the other BG.

Abstract: A first logical link path and a second logical link path between two endpoints are identified. The first logical link path and the second logical link path comprise an OSI layer 5 link path from source to destination including intermediate network devices. The first logical link path and the second logical link paths include at least a different intermediate network device. A communications session is established between the two endpoints. Data packets for the communications session are communicated using the first logical link path. A metric indicative of a time of receipt of packets on the first logical link path is determined. Based on the metric, presence of a network condition at the communications network that is indicative of a performance degradation of the first logical link path is determined. In response, data packets for the communications session are communicated on the second logical link path.

Abstract: A packet advertisement method is implemented by a first network device. The packet advertisement method includes the first network device sending a first packet to a second network device advertising a route to the second network device. The first network device receives a second packet from the second network device in response to a first packet error. The second packet includes first indication information indicating the first packet error. The first network device determines, based on the first indication information, a route advertisement policy of a feature included in the first packet.

Abstract: According to various embodiments, an electronic device includes at least one processor. The at least one processor may be configured to: confirm a data session establishment request from a first application executed by the at least one processor; confirm at least one first descriptor corresponding to the first application for path selection; confirm a first network interface corresponding to the at least one first confirmed descriptor based on the at least one first confirmed descriptor; establish a first data session corresponding to the at least one first confirmed descriptor; and transmit and/or receive data related to the first application through the first network interface using the first data session.

Abstract: Multiple types of lines are made simultaneously available, including a Wi-Fi link, a cell link and a wired link. A list of running cloud applications is identified by monitoring A quality of each available link for each running cloud application is periodically tested, including measurements of latency, jitter and packet loss. A first link is selected for a first application and a second link is selected for a second application. Data packets related to the first application are transmitted over a first link and data packets related to the second application over the second link.

Abstract: In one embodiment, a device obtains quality of experience labels for an online application accessible via a network. The device also obtains telemetry data from the network that is associated with quality of experience labels. The device makes, based on the quality of experience labels and telemetry data, a determination as to whether the online application is similar to a second online application for which a prediction model is already available. The device trains, based on the determination, a prediction model for the online application to predict a quality of experience metric for the online application.