ACCUVER | leading wireless test and measurement solutions

contact us

RESOURCES: At Accuver, we’re driven to help our customers and theirs be the first to reach new frontiers of
wireless performance, innovation, value and trust.

LIBRARY

Bridging Lab Emulation and Field Probing for Repeatable NTN Validation

The wireless communication landscape is undergoing a radical shift as Non-Terrestrial Networks (NTN) move from theory to reality. We have entered an era of \"coverage without boundaries,\" driven by an unprecedented scale of satellite deployment. As of early 2026, the number of active Low Earth Orbit (LEO) satellites has surged past 12,500, creating a dense celestial network designed to eliminate global coverage gaps. This orbital infrastructure now supports over 10 million satellite broadband subscribers worldwide, while the emerging \"Direct-to-Device\" (D2D) market is projected to connect a potential audience of over 2 billion people through global cellular integration. However, the leap to space comes with unique technical hurdles. The extreme altitudes of Low Earth Orbit (LEO) introduce massive propagation delays, while the high velocity of satellites creates significant Doppler shifts that can disrupt traditional mobile protocols. Ensuring seamless connectivity between these satellites and ground devices requires a rigorous, dual-phase testing strategy―beginning with high-fidelity simulation in a controlled lab environment and concluding with extensive optimization under real-world field conditions. Accuver is proud to introduce a complete end-to-end NTN test suite designed to empower UE manufacturers, base station developers, and network operators to navigate this high-stakes frontier with confidence. NTN Satellite Link Simulation & Analysis In the lab, the primary challenge lies in replicating the physical realities of space without the prohibitive expense of an actual launch. XCAT-SPACE, Accuver’s 3GPP-based channel emulator, is specifically engineered to simulate wireless channels for satellite and aerospace. By connecting an NTN terminal and a base station simulator to XCAT-SPACE, users can recreate a \"NTN channel environment\" that accurately reflects Doppler shift effects and large-scale time delays.XCAT-SPACE enables controlled and repeatable NTN validation, allowing developers to test their systems against thousands of orbital scenarios in a fraction of the time. This is achieved by precisely simulating LEO-based satellite link conditions―including dynamic propagation delays, complex Doppler effects, orbital mobility, and multi-satellite visibility―ensuring every aspect of the satellite link is rigorously accounted for.The system provides dynamic satellite mobility based on actual orbits. Users can manually configure altitude, trajectory, and terminal movement paths, or achieve even higher precision by automatically importing TLE (Two-Line Element) orbit data. This flexibility allows for comprehensive 5G NR NTN terminal testing and D2D LTE end-to-end validation. For those requiring a perfectly isolated environment, pairing XCAT-SPACE with Accuver’s XCAT-SmartShield or XCAT-Shield Box provides a high-fidelity Over-the-Air (OTA) testing setup. These highly controlled conditions closely approximate wired connections, minimizing external RF uncertainty while preserving essential OTA characteristics. By combining satellite channel emulation with controlled OTA testing, Lab validation supports accurate and repeatable evaluation of NTN devices prior to Field verification. To translate this simulation into meaningful intelligence, XCAP, a post-processing solution, delivers specialized NTN analysis. Since satellite base stations are constantly in motion, XCAP uses 3D map visualization and automated TLE (Two-Line Element) data to correlate RF KPIs with precise satellite positions, azimuth, and elevation at the moment of measurement. This provides a clear performance baseline for 5G NR NTN and D2D solutions across major chipsets. Managing these complex testing cycles is streamlined by XCAP-AMS, a web-based automation platform. By centralizing system control and test reporting into an intuitive web UI, XCAP-AMS eliminates location dependency and reduces operational complexity. Together, this ecosystem―from XCAT-SPACE’s high-fidelity emulation to XCAP’s orbital insights and XCAP-AMS’s centralized management―provides a seamless, repeatable path for NTN device evaluation prior to field verification. 24/7 Unmanned Satellite D2C Monitoring Solution Field verification is the essential step in validating NTN performance under real orbital and network conditions. Accuver enables this through a 24/7 unmanned satellite D2C monitoring solution , designed to capture RF performance, service availability, and mobility behavior across distributed locations using live LEO satellites. At the heart of field operations are specialized measurement probes tailored for mission-critical NTN environments. For deployments in regions where terrestrial networks are unavailable―such as mountainous areas, deserts, or polar zones―the XR-Emb provides a rugged, IP54-rated solution. It ensures stable, automated testing (Ping, PS Call, SMS) in extreme outdoor conditions, supporting real-time status monitoring and failover handling for emergency or remote operations. For multi-UE monitoring, the XR-Pu6 enables continuous 24/7 benchmarking of up to six devices simultaneously. This ensures that operators never miss critical testing opportunities caused by unpredictable satellite pass times, providing a robust verification of satellite coverage and D2C (Direct-to-Cell) performance in the field. These distributed field units are centrally managed by XCAL-Manager, a server platform that eliminates the need for on-site engineers in remote or inaccessible areas. Operators can remotely distribute measurement schedules, monitor satellite links in real-time, and receive automated alarms for network anomalies, significantly reducing OPEX. All measurement data from these multiple sites is then aggregated via XCAP-AMS for consistent data handling and unified platform management. For large-scale or geographically distributed deployments, XCAP serves as the centralized engine for deep-dive analysis. This post-processing solution correlates RF KPIs with real-time satellite dynamics―including azimuth, elevation, Doppler shifts, and free space path loss―derived from TLE-based orbit information. This end-to-end field ecosystem ensures that NTN services are rigorously monitored and optimized in the real world, from the remote edge to the central cloud. Unified NTN Validation Workflow Accuver’s NTN solution provides:ㆍSatellite link simulation, controlled OTA testing, and baseline analysis in the LabㆍContinuous 24/7 unmanned D2C monitoring in the FieldㆍCentralized operation, data aggregation, and integrated analysisㆍConsistent KPI correlation across Lab and Field environments Together, these capabilities support a single, structured NTN validation workflow from simulation to live satellite operation. Conclusion NTN validation requires both controlled simulation and real-satellite verification. Accuver delivers a structured, end-to-end NTN validation solution that enables reliable deployment, efficient troubleshooting, and continuous optimization of satellite-based networks.

Jan 19, 2026
XCAP-Based Connectivity Characterization for D2C and NTN Environments

Analyzing communication performance in Direct-to-Cell (D2C) and NTN environments is inherently complex due to continuously changing satellite orbit, geometry, and RF conditions throughout a satellite pass. Interpreting UE measurement logs solely at the KPI level is insufficient to explain why connectivity is established, maintained, or lost. Effective analysis requires correlation of measurement data with satellite motion and observation conditions along common time and spatial axes.XCAP provides an analysis environment that correlates UE measurement data with satellite orbit, geometry, and Doppler behavior, enabling condition-based interpretation of connectivity behavior in NTN environments. This approach allows complex satellite connectivity phenomena to be analyzed beyond standalone KPI outcomes, focusing instead on the conditions under which they occur.1. Satellite KPI VisualizationXCAP automatically reconstructs satellite position, orbit, and observation KPIs for past measurements based on recorded satellite communication data. This enables examination of satellite environment conditions and communication quality at the exact time of measurement within a unified analysis view.Such visualization allows RF KPIs to be interpreted not as isolated performance values, but in the context of the satellite’s elevation and azimuth at the time of observation. As a result, identical KPI values can be distinguished based on differing satellite geometry conditions, supporting more accurate interpretation of connectivity behavior.- 3D Satellite Map : Visualizes time-based satellite orbit movement together with communication measurement data- Polar Phase Map : Displays satellite trajectory using Elevation and Azimuth for observation condition analysis- RF KPI Summary : Provides RSRP, SINR, Handover, and other major RF KPIs based on measurement logs- Satellite KPI Summary : Shows changes in tracking KPIs observed during the measurement period2. 5G NTN Satellite Test AnalysisXCAP analyzes UE connectivity behavior as satellites enter and move through the field of view, identifying which satellite the UE was connected to at each moment and under what conditions connectivity was retained or lost. By correlating log data with time-varying satellite geometry, the analysis characterizes connectivity behavior under realistic D2C and satellite mobility scenarios.Connectivity events such as service loss, RSRP degradation, and handover are aligned with satellite trajectory, enabling recurring event patterns to be identified along specific orbital segments rather than treated as isolated occurrences. This supports interpretation of UE connectivity behavior as a continuous process across a satellite pass. - Classification of RRC Measurement Report events by time, location, and event type - Analysis of handover delay and mobility-related conditions - Evaluation of satellite link retention time - Examination of link variability caused by satellite movement or environmental factors3. Doppler Effect AnalysisLEO satellites introduce significant Doppler frequency shifts due to high orbital velocity. XCAP provides dedicated analysis capabilities to evaluate Doppler variation and its impact on connectivity stability.Doppler variation is analyzed as a function of satellite geometry, enabling differentiation between Doppler-induced effects and RF performance changes driven by propagation or environmental factors. Alignment of measured KPIs with TLE-based orbital models further allows comparison between expected and observed performance trends along a satellite pass. - Analysis of Doppler frequency shift during satellite communication - Correlation of signal variation with satellite Elevation and Azimuth - Presentation of satellite observation KPIs aligned with Doppler behavior - KPI prediction and comparison using TLE-based orbital models - GPS-based orbit mapping for geometry-to-KPI correlationConclusionXCAP treats UE connectivity behavior in satellite-based networks as a function of orbital motion, geometry, and Doppler conditions rather than isolated KPI outcomes. This enables more precise interpretation of NTN and Direct-to-Cell connectivity characteristics, supporting condition-based understanding of connectivity behavior in dynamic satellite environments.

Nov 24, 2025
Transforming In-Building Benchmarking: Automated Multi-Terminal Analysis Powered by XCAP

In-building benchmarking often relies on data collected from multiple operators and mixed terminal environments. When measurements lack consistent metadata or synchronization, post-processing becomes slow, manual, and error-proneㅡmaking it difficult to generate unified and repeatable building-wide KPI results, particularly in multi-operator scenarios. Customer Challenge Traditional in-building workflows force engineers to manually align logs, classify floors, and prepare operator comparison reports. Multi-device sessionsㅡeach with different measurement modes, timestamps, or metadataㅡfurther increase complexity and reduce efficiency. Producing a consistent, building-level analysis therefore demands significant manual effort and is difficult to scale. XCAP Solution XCAP eliminates these inefficiencies by automating core in-building analysis tasks. The platform consolidates logs from diverse terminals, aligns them to a unified building structure, and generates standardized outputs within minutes. With automated classification, comparison, and KPI processing, XCAP delivers a scalable, repeatable, and operator-agnostic indoor testing workflow. Key Capabilities1. Get In-building Data Imports floor plans and in-building statistics from another log model and applies them to sessions that lack in-building metadata.- Supports mixed terminal environments- Time Offset option aligns logs with minor timestamp differences- Same-time validation ensures data integrity 2. Automated Floor Classification Automatically detects and separates measurements by floor, reducing manual review and accelerating analysis for multi-story buildings. 3. Multi-Operator Delta Comparison Generates reference-based delta coverage and differential KPI charts, enabling fast and accurate operator-to-operator evaluation on a unified layout. 4. Spatial Binning Creates virtual analysis grids for zone-level KPI evaluation (RSRP, RSRQ, SINR), improving readability and precision for location-based comparisons. 5. Unified KPI Statistics Provides building-level and floor-level summaries, including:- Technology distribution (5G SA/NSA, LTE)- Signal strength and quality indicators- Floor-specific coverage visualization Conclusion XCAP streamlines in-building benchmarking by reducing post-processing time, standardizing reporting across diverse logs, and improving the accuracy of multi-operator comparison. With automated data integration and consistent, repeatable reporting, engineering teams can identify coverage gaps and performance issues more quicklyㅡallowing them to focus on network optimization rather than manual log alignment.

Nov 24, 2025
Connectivity, and the Future of Vehicle Safety: From Euro NCAP Incentives to Global Mass Rollout

The automotive industry is entering a new phase in which connectivity is no longer viewed as an optional feature but as a critical enabler of road safety. With Euro NCAP’s decision to include connectivity in its safety rating framework starting in the near future, connected vehicle technologies are gaining formal recognition as part of a vehicle’s overall safety performance. This upcoming change is expected to accelerate the adoption of Vehicle-to-Everything (V2X) technologies, placing them at the center of global automotive strategies. Background: Connectivity and Vision ZeroThe European Union’s Vision Zero initiative, which aims to eliminate road fatalities by 2050, has made connectivity a cornerstone of its strategy. By enabling vehicles to share and receive real-time hazard information, connectivity reduces the likelihood of accidents caused by unforeseen road conditions, limited visibility, or delayed human reactions. Euro NCAP, a widely influential vehicle safety assessment program, has announced that connectivity will be incorporated into its star rating system. While not yet enforced, this policy change provides a clear signal to automakers: future ratings will reward vehicles equipped with connectivity features, and consumer demand will follow.Industry Implications OEMs1. Incentive to integrate connectivity features to achieve higher Euro NCAP ratings.2. Easier internal justification for investment since higher ratings influence sales.3. Need to ensure interoperability and participation in shared data ecosystems. Regulators and Road Authorities1. Gain access to real-time hazard and traffic data from connected vehicles.2. Enhanced ability to manage infrastructure safety (e.g., roadworks, weather, wrong-way driving). Customers1. Receive earlier warnings of hazards such as slippery roads, obstacles, or congestion.2. Benefit from safer vehicles without relying on paid services or third-party apps. Industry Ecosystem1. Establishes a data-sharing culture where safety information flows across brands.2. Accelerates market readiness for advanced applications (e.g., emergency vehicle alerts, end-of-queue warnings, sensor data fusion). This announcement sets the stage for rapid adoption. Historically, once Euro NCAP introduces new assessment categories, they quickly become industry standards. Connectivity is expected to follow the same trajectory, from incentive, to widespread adoption, to potential regulation.Accuver’s Role in the Connectivity EcosystemAccuver contributes to the evolving connectivity and V2X ecosystem by providing comprehensive, end-to-end solutions that bridge today’s testing needs with tomorrow’s deployment challenges:- V2N Solutions: Developed in collaboration with leading automotive electronics companies, Accuver enables secure and scalable Vehicle-to-Network integration for connected services.- V2X Modules: Provide reliable, standards-compliant connectivity for vehicles, enabling seamless communication across V2V, V2I, and V2N use cases.- V2X Test Solutions: Cover both physical layer and application layer validation, ensuring interoperability and performance under both real-world and lab conditions.- On-Board Units (OBUs): Extend connectivity to vehicles without integrated V2X modules and support Vulnerable Road Users (VRUs), ensuring broader ecosystem inclusion.- eCall Testing Automation: Automates the manual eCall validation process, delivering higher efficiency, repeatability, and accuracy in ensuring compliance and functional reliability.By offering these solutions, Accuver helps automakers, regulators, and technology providers validate, deploy, and scale connected safety systems. This ensures that connectivity is not only a rating requirement but also a reliable, interoperable, and future-proof component of the road safety ecosystem. ConclusionConnectivity has not yet been fully implemented into Euro NCAP’s rating system, but its planned inclusion signals a clear direction for the industry. V2X technologies are poised to become a core component of vehicle safety, with profound implications for automakers, regulators, and consumers. Accuver’s role is to accelerate this transition. Through V2X modules, OBUs, and comprehensive testing platforms, Accuver equips the industry with the tools required to validate technologies today and prepare for their integration into global safety frameworks tomorrow. By enabling reliable, interoperable, and scalable deployments, Accuver turns policy announcements into measurable road safety outcomes, fully aligned with the long-term goals of Vision Zero.

Oct 22, 2025
[Case Study] XCAL-VQML Deployed at Major Network Vendor’s R&D Center

The rapid growth of video-based communication, ranging from streaming services to enterprise conferencing, has made video quality assurance a critical requirement for operators and telecommunication equipment vendors. Unlike traditional KPIs such as throughput or latency, video quality is inherently subjective and influenced by multiple factors at device, service platform, and network levels. - How does the network affect actual video quality as perceived by end users?Traditional testing approaches, which rely on physical cameras or manual MOS evaluations, made it difficult to:- Eliminate environmental variables such as lighting or device handling.As a result, the R&D team struggled to build confidence in their ability to benchmark video service quality under real-world conditions accurately.XCAL-VQML OverviewAccuver provided the XCAL-VQML solution, an AI-based video quality assessment framework designed to deliver accurate MOS values without requiring original reference videos.Key capabilities included:AI-powered MOS Prediction:Supports widely used services such as YouTube, WhatsApp, WeChat, Skype, Microsoft Teams, etc., reflecting actual end-user experiences.Controlled Measurement Environment:OutcomeThe deployment produced tangible results:- Repeatable MOS values were achieved, proving that XCAL-VQML could reliably measure how the network impacts perceived video quality.These results demonstrate that XCAL-VQML produces reliable and reproducible video quality scores, even under varying device or application conditions. - Understand the real impact of network conditions on video quality.- Accelerate innovation in video communication in the mobile network.ConclusionFor this usage, the deployment of XCAL-VQML transformed a critical challenge into a breakthrough success. What began as doubts about how to measure video quality ended with a proven framework for reliable, scalable testing.

Sep 25, 2025
[Case Study] XCAL-Ranger: Continuous 24/7 Measurement Across Train, NTN, and Private 5G

With the rapid evolution of 5G, private networks in industrial environments and NTN communication, the demand for automated, robust, and scalable monitoring solutions is increasing worldwide. Ensuring reliable connectivity is now a critical requirement not only for consumer satisfaction but also for mission-critical applications across transportation, satellite, and smart manufacturing sectors. XCAL-Ranger addresses these emerging needs by providing continuous, 24/7 monitoring and control in even the most challenging environments. 1. Continuous Network Quality Monitoring in Transportation (Rail, Metro, Maritime)Transportation systems such as trains, subways, and ships require stable connectivity to ensure both passenger satisfaction and operational safety. XCAL-Ranger enables continuous and automated measurement of network quality by being mounted directly onto moving vehicles. Challenges- Harsh environments such as tunnels, underground stations, and offshore conditions can impact measurement stability- Maintaining reliable installation and stable power supply on moving vehicle- Data synchronization and remote monitoring may require robust Backhaul connectivityApplication- Mounted in trains or subways, it collects service quality data while moving through tunnels, stations, and high-speed tracks- Installed on ships or ferries, it monitors coastal and offshore coverage, detecting areas of weak or unstable connectivity- Data can be integrated into passenger information systems, safety applications, or QoS dashboards for operatorsBenefits- Identify coverage gaps and service blind spots in real-time- Support network optimization projects by correlating measurement results with passenger complaints- Supports compliance with regulatory standards while enhancing passenger experience- Ensures robust, stable, and 24/7 monitoring and control even in harsh and mobile environments2. Automated Testing of Direct-to-Device (D2D) Satellite Communications with LEODirect-to-Device (D2D) satellite communication with Low Earth Orbit (LEO) constellations enables smartphones and IoT devices to connect directly to satellites. Such communication windows are short and dynamic, requiring automated, scheduled testing. Challenges- Satellite pass times are short and require precise scheduling for meaningful testing- Environmental factors such as weather or urban obstructions can interfere with satellite signal acquisition- Logging and analyzing intermittent test data requires advanced automation and coordinationApplication- XCAL-Ranger is configured to periodically attempt satellite connections whenever a LEO satellite passes overhead- Automated tests include signal acquisition, call attempts, and data session initiation with direct-to-device satellites- Logs are uploaded remotely for performance evaluation of LEO network readinessBenefits- Provides continuous validation of satellite coverage in test locations without human intervention- Enables R&D; teams and operators to benchmark satellite service performance- Reduces missed test opportunities caused by unpredictable satellite pass timings- Guarantees robust, stable, and 24/7 automated testing of satellite communication, even under unpredictable conditions3. Assurance of Private 5G Networks in Smart FactoriesPrivate 5G networks are increasingly deployed in industrial environments to support automation, robotics, and mission-critical communications. Ensuring stable network performance in such dynamic environments are essential. Challenges- Factory environments are subject to heavy interference from machinery and structural obstacles- Maintaining continuous monitoring across large and complex sites can be resource-intensive- Security and privacy concerns may arise when integrating monitoring solutions into operational networksApplication- XCAL-Ranger is deployed across a smart factory floor to run continuous performance checks- Automated test scripts validate connectivity, latency, and application service performance for AGVs, robots, and wireless sensors- Device status and performance data are centrally monitored through the XCAL-Manager platformBenefits- Guarantees that 5G-enabled devices operate seamlessly in production environments- Reduces downtime by proactively detecting network degradation- Provides auditable network performance reports for enterprise customers and regulators- Delivers robust, stable, and 24/7 monitoring and control, ensuring reliability in demanding industrial settings XCAL-Ranger transforms traditional, labor-intensive network testing into a fully automated, scalable, and reliable monitoring system. Whether in transportation systems, satellitecommunications, or smart factories, XCAL-Ranger ensures robust, stable, and around-the-clock monitoring and control even in the most challenging environments making next-generation connectivity truly dependable.

Sep 24, 2025
A Comparative and Measurement-Based Study on Real-Time Network KPI Extraction Methods for 5G…

Turkcell, Turkiye’s leading telecom provider, has advanced its 5G research using Accuver’s XCAL to evaluate real-time KPI extraction from 5G networks. This study highlights XCAL’s accuracy, fast refresh rate, and stable performance in measuring key indicators like RSRP, RSRQ, and SINR.XCAL’s advanced features enabled precise and reliable analysis, offering a key competitive advantage in 5G performance evaluation. The full paper is available on arXiv for those seeking more insights.We sincerely thank Turkcell, one of our key customers, for their continued research efforts in advancing 5G innovation with XCAL.

Jun 11, 2025
[White Papers] AI-powered RCA

Table of contents1. Introduction 2. What is AI RCA? 3. Technology applied to model. 4. Learning and evaluation 5. XCAP-Cloud with AI Powered RCA6. Use Cases 7. Future Directions *Under R&D collaboration with Korean MNOs IntroductionWith the rapid development of wireless communication technology, ultra-high data transmission speeds and connections to various devices increase, the communication environment is becoming more diverse and complex. Accurate and rapid response to communication system failures resulting from this diversity and complexity is essential. To address these market demands, we provide automated wireless network optimization testing solutions as well as logic-based RCA solutions that identify the causes of various defects that occur in mobile communication networks and provide appropriate solutions.Our logic-based RCA solution utilizes wireless network communication protocol transmission/reception information and terminal status information to accurately identify and resolve problems through structured data and rule-based analysis. However, in this advanced communication system, parameter settings for each analysis rule are complex, and it is difficult to consider characteristics of the field situation, limiting rapid response.To overcome these limitations, we developed a machine learning-based RCA solution. By utilizing the latest machine learning technology to learn subtle differences in the network environment hidden in the vast amount of data collected from mobile communication networks, large-scale data can be quickly analyzed and diagnosed based on the data without relying on individual subjectivity. This is expected to contribute to improving the stability of communication systems. What is AI Powered RCA? Our solution is a machine learning-based RCA solution that utilizes our automated testing solution to perform root cause labeling using raw data obtained from network access failure and service interruption log samples. The training dataset contains the network\'s signal level and quality indicators, as well as network quality indicator metrics such as data throughput, latency, and packet loss rate for each layer.Figure1. AI RCA concept diagramThis training dataset consists of approximately 1 million log data, including network issues that occurred in various environments during field testing. This data is collected through a variety of methods, including field testing, simulation testing, and laboratory testing, to reflect the complexity of network problems. To address the complexity of the problem presented by strong correlations between key indicators, our model minimizes similarities between data characteristics and learns each root cause individually to enable accurate classification and understanding.Additionally, machine learning has the characteristic of being capable of continuous learning and improvement. This means our solutions can continuously optimize and improve models to respond to new challenges that arise during real-world operations. This offers great advantages in maintaining reliability in a rapidly changing communications environment.Additionally, AI\'s automated decision-making capabilities help quickly process and diagnose large amounts of data. Our model supports efficient and accurate communication system problem solving while minimizing individual subjectivity through data-based judgment.These technical advantages allow our solutions to leverage the powerful analytical power of machine learning to help improve network reliability and improve availability and performance. Figure2. Examples of data set Technology applied to AI Model We adopted a gradient boosting ensemble model using XGBoost and developed a powerful tool for effective root cause diagnosis. XGBoost is applicable to both classification and regression problems and is particularly characterized by excellent performance for a variety of data sizes and types.Figure3. XGBoost concept diagram XGBoost features provide excellent performance for a variety of data sizes and types. This is very useful as we deal with large amounts of network issue data from a variety of environments. The model learns useful patterns from large amounts of data and can effectively identify root causes. Additionally, XGBoost uses parallel processing and optimized data structures to provide fast learning and prediction speeds. This is a big advantage of XGBoost\'s fast learning and prediction capabilities to quickly respond to various problems that occur in large-scale networks.For root cause diagnosis, prioritizing each characteristic and creating each XGBoost model based on this increases the interpretability of the model and provides a clear understanding of the characteristics of each root cause. Through these feature priorities, the model learns the importance of features for each root cause and enables effective response to the complexity of the problem.In this way, our model using XGBoost combines strong prediction ability with fast response speed to achieve efficient root cause diagnosis. Leaning and evaluation process Figure 4 shows the learning and evaluation process. First, the data is purified through data preprocessing. Duplicate data is removed, and outliers and missing data are replaced with the most prevalent value in each data set. Lastly, for data imbalance, down-sampling using Euclidean distance calculation and up-sampling using SMOTE are performed to balance the data.In the Euclidean distance calculation method, the distance of points other than the root cause label is compared one-to-one based on the center point in the distribution of the root cause label. After comparison, points that are judged to be too far from the distribution of the root cause label are removed, leaving only the points that exist as close to the boundary as possible. The SMOTE method synthesizes adjacent minority class samples between majority class samples. This increases the number of samples of minority classes, helping the model learn better and recognize minority classes better.Figure4. Training process and evaluation processAfter performing this preprocessing process, an XGBoost model according to each priority is created and trained and verified. During verification, a grid search technique is used to find the most appropriate hyperparameters for the model. If you input the data set you want to test into the learned RCA solution, you can get results where the root cause label predicted by the model matches the actual root cause label.The evaluation results are as follows: It was derived as the average accuracy value of each model. Total countSuccess countFail count Accuracy 5G-NR PS 27365242433122 88.59% 5G-NR Voice29903 25376 4527 84.86% LTE PS90505 82813 7692 91.50% LTE Voice29501 26958 2543 91.38% Total177274 159390 17884 89.91% Figure5. AI model evaluation results XCAP-Cloud with AI-powered RCA AI-powered RCA is provided by XCAP-Cloud, a cloud-based mobile network analysis solution. We use test equipment to collect data generated from the telecommunications carrier\'s mobile communication network. Collected data is uploaded to the server in log form. Users can define rules to identify specific patterns in logs and send them to the AI model when those patterns are found. Logs must be interpreted before being fed into an AI model. Log interpretation involves understanding the contents of the log and extracting KPIs. Interpreted logs and extracted KPIs are sent to the AI model through grpc. grpc is a protocol for efficient and reliable data transmission. ARCA infers the root cause based on the received data.Inference results can be checked using various visual tools provided by XCAP-Cloud. Visual tools help you intuitively understand inference results. Figure6. XCAP-Cloud The accuracy of AI prediction results extracted in real time from the XCAP-Cloud system equipped with an AI-powered RCA solution was observed to be 97%. Figure7. Real-time AI prediction accuracy Use Cases All issues that may arise in the 5G/LTE environment are categorized and managed through RCA, and each event is labeled to help easily identify the root cause. VoNR Call Setup Failure CaseVoice services using 5G RAN, 5G Core, and IMS are called Voice over New Radio-VoNR. NR UEs can perform voice services directly on the NR network without falling back to the LTE network. VoNR Call Setup Failure can occur for a variety of reasons. In the initial network construction stage, Cell Search failure, PDCCH Decoding error, IMS Registration failure, etc. are the main causes that can cause problems in which the terminal cannot connect to the network or register with the IMS server. This solution can quickly classify the problem by extracting the cause of cells that cause many setup failures. VoNR Call Drop CaseAs a mobile communication system, even if the initial setup is successful and the call is connected normally, a call drop may occur when entering the cell edge or when handover occurs due to RF deterioration or when the settings of the source cell and target cell cannot maintain connectivity. can. In addition, even though information about neighboring cells is searched periodically, when a call must be continued without finding a suitable cell, a large amount of RTP packets is lost, and the network reclaims radio link resources to cause a call drop. can. This solution can help you accurately analyze call drops. By additionally checking the packet data and Layer3 messages provided within XCAP-Cloud, you can quickly identify problems and take appropriate action. Figure8. RCA Workflow NR FTP Low Throughput CaseNR FTP Low Throughput CaseAfter the normal call setup process is performed in 5G NR, you may experience quality issues in data calls such as FTP and HTTP with processing speeds lower than expected.Typically, degraded RF performance may indicate low throughput, while normal RF performance may indicate parameters related to throughput and capacity. It can be caused by various reasons such as UL/DL bandwidth, MCS, Layer, Rank Index, etc.This solution can help with accurate analysis by extracting the cause of low RB allocation that occurs in the network and the resulting low throughput cases. Future Directions We aim to provide intuitive and versatile solutions to diagnose wireless network problems quickly and accurately.A successful AI model must not only achieve high accuracy, precision, and recall, but also ensure reliable prediction performance at a level that actual customers can use as wireless network analysis indicators. To achieve this, AI models must learn the know-how of highly skilled wireless network analysis experts and continuously evolve.Our goal of a wireless network analysis know-how training system is to build a customer-tailored AI model learning infrastructure to provide a system that allows customers to directly discover data sets and upgrade AI models. Through this, concerns about personal information and data leaks can be resolved, and AI models that meet customer needs can be built more effectively.In addition, if only data from the wireless network connection section of the mobile and base station is used, the root cause of the failure due to problems with the upper layer access probe or core probe may be unclear. Therefore, there is a need to develop it into a comprehensive learning model from end to end. We will continue to take on this challenge without stopping.Figure9. Customer-tailored AI model learning system

Jun 05, 2024
[White Papers] A New Standard for Video Quality Assessment, VQML

.center-table { margin-left: auto; margin-right: auto; border-collapse: collapse; } ContentsNecessity of VQMLWhat is VQML? Techniques in VQMLLearning & Evaluation Process of VQMLInput & Output Value Dataset Utilized in VQMLOutstanding Points of VQMLHigh Performance of VQMLVerify Reliability of VQMLNecessity of VQMLThe mobile network market is growing explosively due to the rapid growth of 4G and 5G subscribers and the expansion of service coverage. According to the Ericsson Mobility Report in November 2020, 5G subscribers are expected to exceed 3.5 billion by the end of 2026.In this situation, the growth of mobile video services is no surprise. Figure 1 shows the traffic volume of mobile data by media measured by BI Intelligence. In 2020, video services account for about 75% of the total mobile data usage. Combined with the growth of 5G and the COVID-19 situation we are facing today, this number will grow even more and mobile network operators should be prepared for it.Figure 1. Mobile data traffic usage by mediaThe performance and quality of the video services that users experience are critical aspects of the network operations. In live-streaming services that transmit video over the network in real-time, videos may not be completely transmitted due to various loss issues in the network. This may result in different qualities of received videos between consumers. To prevent such situation, we need to find a way to accurately measure the video quality in real-time and ensure consistence in good quality of video transmission on mobile networks.Innowireless’ VQML is introduced as a deep learning-based, high-level video quality assessment solution. Compared to other traditional evaluation methods, VQML is distinguished as a better solution due to its faster processing speed, cheaper implementation and no need for original reference video.As a new way to measure video quality, VQML lets mobile and broadband network operators meet customer expectations and needs with optimized network operation.What is VQML? The simplest way to measure the quality of a video is to obtain the MOS (Mean Opinion Score) using human judgements. This is also the most accurate way, but it requires too many people and too much time and cannot be proceeded in real-time.In order to resolve these time and cost problems, VQML takes high volume of video data and continuously trains a neural network to predict the video quality score as accurately as the human judgements. Using deep learning, VQML learns patterns of videos and MOS values from a database derived from viewers’ large-scale surveys.VQML predicts the quality of the video as a MOS value within a range of 1 and 5 which corresponds to the actual human perception.The meaning of each score is as follows. Score Quality Perception 5 Excellent 4 Good 3 Fair 2 Poor 1 Bad Figure 3. Video quality assessment in VQMLTechniques in VQMLMethods for Video Quality Assessment (VQA) include Full Reference (FR), Reduced Reference (RR), and No Reference (NR). Figure 4. Video Quality Assessment SolutionsFR, which is currently used by most products, evaluates the quality by comparing the original video with the received video. Although it shows high reliability due to direct comparison between videos, it is difficult for the client to have the original video. It is also difficult to be used on platforms where videos are created and serviced in real-time.For this reason, FR method is not a suitable solution due to its limitations especially in current situations where live-streaming services are commercialized, and video conferences and classes are increasing due to the Covid-19. VQML operates based on the NR method that measures quality only with the received video. Without the need for the original video, NR method can calculate the quality metric of any video in real-time and effectively assess the video quality in areas where FR methods are difficult to apply.Some well-known areas where NR methods are ideal quality assessment tools are CCTV and real-time video platforms. These areas can utilize the NR method to measure the actual perceived quality by identifying all the degradation factors on their own video without a comparison video.NR method mostly measures quality by extracting statistical characteristics of the video with mathematical algorithms. It works based on the KPI designed by researchers, so it may result in large average error compared to the actual quality score perceived by humans.VQML uses large-scale and highly reliable database to repeatedly train itself to continuously improve its prediction of video quality score in order to compensate for the limitations of the NR method.Learning & Evaluation Process of VQMLThe training of VQML uses the KoNViD-1k and YouTube-UGC datasets. The KoNViD-1k dataset is a large database consisting of MOS collected from over 1,000 videos and dozens of viewers. The YouTube-UGC dataset refers to a video database with 4K UHD characteristics. VQML’s deep learning network consists of 2 CNN modules and 1 GRU module.Figure 5. Learning Process of VQMLWhen a video sample from the learning dataset is input into the model, the CNN module processes each frame and the GRU module follows to analyze the sequence of consecutive frames in order to recognize the pattern of the features of the video sample. The extracted features are used to finally predict the quality of the video.After the measurement result comes out, calculate the difference from the actual MOS and modify the deep learning network to reduce this error. By repeating this process, the deep learning network learns to measure the quality value of that video with minimal error.After enough training, VQML can predict a video quality score that is very close to the actual MOS value.Input & Output ValueVQML only requires the video itself to predict the video quality score. When the video is input into the VQML neural network, VQML automatically decodes it and converts it into frames, which are sets of pixel values expressed in RGB, and extracts features from the frames to compute the quality score of the video as the output of the model. VQML can execute this process in speed that is close to real-time.By default, the quality score of the video is output after the entire video is processed through VQML. There is an option provided to adjust the video viewing time in order to predict the quality score for specific segments within the video.Dataset Utilized in VQMLDeep learning method has recently been applied to various applications. A highly acknowledged database is required to ensure reliable performance of such deep learning-based solutions.The KoNViD-1k database (http://database.mmsp-kn.de/konvid-1k-database.html) used for VQML’s learning is a highly reliable database consisting of 1,200 videos evaluated by more than 100 people. It has been cited more than 130 times in papers around the world since the VQA group at the University of Konstanz, Germany, published it at the IEEE QoMEX 2017 academic conference, and is widely accepted by academia as a database for video quality assessment.Another learning database, the YouTube-UGC dataset, contains user-generated content collected from YouTube, providing videos of different resolutions and formats. Some of them have 4K UHD (Ultra High Definition) characteristics, which are useful resources for research and development activities and are widely sed for quality evaluation and content classification.The KoNViD-150k database (http://database.mmsp-kn.de/konvid-150k-vqa-database.html), released in 2021, also consist of KoNViD-150k-A set (152,265 videos evaluated by 5 people) and KoNViD-150k-B set (1,577 videos evaluated by more than 89 people), ensuring high reliability. The VQA group at the University of Konstanz says this database can be used for efficient video quality tests. Besides the training datasets from KoNViD-1k, VQML utilizes the KoNViD-150k-B set database for objective performance testing.This allows the deep learning network in VQML to have high accuracy and perform video quality prediction that reflect patterns in real-world communication environment.Outstanding Points of VQMLVQML is also unique in the configuration of deep learning networks.VQA solutions based on deep learning often include transfer learning using pre-trained CNN modules.Most solutions utilize a pretrained CNN module followed by a recurrent module such as a LSTM or GRU module. In such structure, the CNN module is pre-trained using the Image Net database (https://www.image-net.org/download.php) consisting of more than 1 million images built by Professor Li-Fei-Fei at Stanford University. The model’s learning process based on the training database only occurs throughout the following recurrent module.In contrast, VQML consists of 2 CNN modules and 1 GRU module. Each CNN module is pre-trained with the ImageNet database and the KonIQ-10k database (http://database.mmsp-kn.de/koniq-10k-database.html) which is a set of over 10,000 images produced by the VQA group of the Konstanz University. Then a GRU module continues to train the model with the features extracted from the 2 previous CNN modules. So the actual learning based on the training data occurs through all 3 modules. In order to mimic a unique characteristic of the human visual system, a temporal pooling layer is added. This layer takes into account how humans perceive an instant degradation in video quality due to a sudden drop in the streamed video much greater than it actually is, and result in a stronger degradation in overall video quality.In conclusion, VQML is able to extract highly detailed contextual as well as temporal characteristics of the assessed videos better than the other VQA solutions.High Performance of VQMLFigure 9 is a verification graph of VQML’s performance based on the KoNViD-1k dataset.Figure 9. Correlation with Ref-MOS and VQMLThis is the correlation between the actual viewers’ MOS and the quality score predicted by VQML. The measurement results are very close to the actual MOS values forming almost a straight line. The graph shows a correlation of about 86.5%, which proves the VQML is a reliable solution.Also, the MAE (Mean Absolute Error) of VQML is 0.235, which is smaller than the average MAE of other products currently commercially available. This means that the output of VQML’s prediction can be as accurate as the human perception on judging the quality of the Video.Verify Reliability of VQMLTo verify the reliability of VQML, tests were conducted using various types of videos. Each video has its own characteristics that represent the video quality in a specific environment. TypeFeatureDramaGeneral screenMovieRelatively dark screenSports Lots of movement and bright lightAnimation Artificial color that stands out Figure 10. Video types and features used in tests After distorting the indicators of each video into several units using the FFmpeg codec, the MOS value of the video to which the distorted indicators were applied was measured. Blockiness Option Change the degree of block generation by adjusting the bitrate with b:v option Unit 10000k / 5000k / 1000k / 500k / 300k/ 200k / 100k BlurOptionUse the boxblur option to adjust the degree of blur in the videoUnit0.0 / 2.5 / 5.0 / 7.5 / 10.0 / 20.0 / 30.0 BrightnessOptionChange brightness by adjusting brightness option of Eq AVOptionsUnit-1.0 / -0.75 / -0.5 / -0.25 / 0 / 0.25 / 0.5 / 0.75 / 1.0 ColorfulnessOptionColor distortion by adjusting the saturation option of Eq AVOptionsUnit0.0 / 0.5 / 1.0 / 1.5 / 2.0 / 2.5 / 3.0ResolutionOptionChange the resolution with the scale optionUnit2160p / 1440p / 1080p / 720p / 480p / 360p / 240p / 144pFigure 11. Encoding options, units and examples applied to video indicatorsThe measurements made using the VQML algorithm are as follows. ?Blockiness : The higher the bitrate, the higher the MOS value.?Blur : The cleaner the video, the higher the MOS value.?Brightness : The lower the brightness, the lower the MOS value. ?Colorfulness : The lower or higher the color distortion intensity, the lower the MOS value. ?Contrast : The lower the contrast, the lower the MOS value. ?Resolution : The higher the resolution, the higher the MOS value. Figure 12. MOS value measurement results for distorted indicator-specific videoColorfulnessOptionColor distortion by adjusting the saturation option of Eq AVOptionsUnit0.0 / 0.5 / 1.0 / 1.5 / 2.0 / 2.5 / 3.0ContrastOptionChange the contrast by adjusting the contrast option of Eq AVOptionsUnit0.0 / 0.5 / 1.0 / 1.5 / 2.0 / 2.5 / 3.0ContrastOptionChange the resolution with the scale optionUnit2160p / 1440p / 1080p / 720p / 480p / 360p / 240p / 144pFigure 11. Encoding options, units and examples applied to video indicatorsThe measurements made using the VQML algorithm are as follows.Blockiness : The higher the bitrate, the higher the MOS value.Blur : The cleaner the video, the higher the MOS value.Brightness : The lower the brightness, the lower the MOS value. Colorfulness : The lower or higher the color distortion intensity, the lower the MOS value. Contrast : The lower the contrast, the lower the MOS value. Resolution : The higher the resolution, the higher the MOS value. Figure 12. MOS value measurement results for distorted indicator-specific videoWhen the MOS value according to the change of each encoding option was compared with the original video, the original recorded the highest score. In other words, as the degree of distortion increases, the video quality decreases, and the MOS value measured by VQML tends to decrease.These results are interpreted as important indicators showing that VQML is a reliable tool for video quality evaluation. As such, VQML, an original video quality assessment solution unique to Innowireless, is expected to be sufficiently competitive in the market.

May 17, 2024

1
2
3
4
5