Leveraging advanced analytics and AI engine, IPM predicts and prevents network performance issues before they happen, building Capacity growth models and forecasting user behaviour and traffic load on the network, giving proper recommendations that keep network performance on track with this growth and user behaviour changes.
Figure 1: KATANA Platform Modules
IPM Analytics is the heart of KATANA IPM module, and it offers the below different uses cases.
Figure 2: IPM Main Functionalities sub-modules
Anomaly Detection:employs a machine-learning algorithm to understand the patterns of Key Performance Indicators (KPIs), making comparisons and autonomously recognizing deviations. The resulting scores are presented for each instance, facilitating straightforward identification of anomalies and their deviation from the typical cluster or common behavioral patterns within the network.
Forecasting Analysis: iPM encompasses various forecasting techniques within a unified interface, granting users the capability to analyze anticipated future trends in network usage for any counter and Key Performance Indicator (KPI).
Figure 3: DL Traffic Volume Forecast
Capacity Management: As networks expand and experience heightened traffic, there is often a decline in network performance. To prevent this deterioration, iPM Capacity management function becomes crucial to enhance performance and restore it to its initial levels. iPM is Addressing traffic shifts requires the strategic rebalancing of network traffic, ensuring even utilization across the network, thereby deferring capital expenditures on new equipment.
Worst Cell List: The Worst Cell List Report, an integral component of our iPM capabilities, is robustly supported by ranking conditions tied to specific periods for designated Key Performance Indicators (KPIs). This functionality empowers users to assess nodes with the poorest performance through detailed maps and charts.
Worst Degraded List:This module, seamlessly integrated into our iPM suite, efficiently troubleshoots and compiles a list of nodes with degraded performance over a specified period. It conducts in-depth analyses through maps and charts, facilitating immediate examination at the work area for detailed troubleshooting
Figure 4: Creation Criteria for WDL
Benchmark: After implementing an optimization action, users have the flexibility to initiate a benchmark across a set of Key Performance Indicators (KPIs). This benchmarking can be conducted on a Day-to-Day, Week-to-Week, or Month-to-Month basis, allowing for comprehensive performance evaluation over various timeframes.
Swap & Acceptance: In Swap Projects, users are required to compare Key Performance Indicators (KPIs) before and after the swap. iPM provides users with the convenient option to effortlessly compare the performance of vendors, facilitating a streamlined assessment of the impact of the swap on network performance.
In the world of mobile network testing, efficiency and accuracy are crucial for optimizing network performance. The “Eagle Eye” feature in INOS is a powerful tool that enables users to analyze large data sets in logfiles, extracting valuable insights through geofencing. This feature facilitates data-driven decision-making for network optimization.
Unveiling “Eagle Eye”
The “Eagle Eye” feature in INOS allows users to effortlessly search through logfiles containing millions of samples, making it a reality. Unlike the traditional method that required meticulous effort in recalling file names, dates, and locations, “Eagle Eye” offers an intuitive solution. Users can define their area of interest using geofencing, and the feature retrieves all relevant data within that area, saving both time and effort.
Figure 1
Optimizing Search Results
“Eagle Eye” offers various settings to streamline the search process, allowing users to optimize and narrow down their results efficiently. These settings include:
Time Aggregation: Users can pick time granularity (hourly, weekly, or monthly) to analyze data over specific periods, aiding trend identification.
Distance Aggregation: Users can set the desired distance aggregation for a detailed location-based network performance analysis.
Operator MNC/MCC: Users can filter their search results based on specific Mobile Network Code (MNC) and Mobile Country Code (MCC).
KPI Selection: Users can choose the Key Performance Indicators (KPIs) they wish to extract from the logfiles.
Date Range and Time Clustering: Users can set a date range and cluster data in time to better understand network performance changes in specific periods.
Interpreting the Results
Once the search parameters are set, “Eagle Eye” presents users with an interactive map accompanied by a timeline.
Figure-2
“Eagle Eye” offers a user-friendly visual representation for easy navigation, providing insights into network performance across locations and timeframes. It also includes benchmark tables and histogram charts for comparative analysis and trend identification.
Use Case Scenarios
The application of “Eagle Eye” in INOS extends to various use cases like;
Pre-action Network Assessment
Performance Benchmarking
Team Performance Assessment
Investment Impact analysis
Figure-3
Figure-4
Conclusion
“Eagle Eye” in INOS is a game-changer for mobile network testing, with geofencing, result optimization, and visual representations that empower efficient insights extraction. It enhances decision-making, network performance, and operational excellence in mobile network testing.
In a world where mobile communication is focused on the use of apps and data, does the quality of a voice call still matter? And is it worth communications service providers (CSPs) spending effort on improving it?
In this blog post Amr Ashraf, Digis Squared’s RAN and Software Solution Architect and Trainer argues that “Yes, it absolutely is! Voice quality, and particularly silence within calls – can you still hear me? – is one of the most tangible aspects of network quality for end users.”
Read on for insights into Digis Squared’s AI-centered voice call quality testing capabilities, using INOS.
If it’s important, we call.
Does voice quality still matter? “Yes! As voice technologies continue to evolve, and call costs drop, it continues to be important to ensure that the quality and clarity of voice calls is maintained,” explains Amr Ashraf, RAN and Software Solution Architect and Trainer.
“The phone call, the most basic and original capability of the mobile phone service, is also the most tangible for end users. Despite the huge range of apps we have on our phones, more often than not, it’s a voice call that’s used for communicating the most important, most sensitive and most urgent information.”
If we can’t clearly hear and understand what is being spoken on a call, or in a voice note, whichever app or method is used to connect or send the audio, then the customer’s perception is always that the network coverage or capacity is of low quality.
“If you find yourself saying ‘Can you still hear me? Are you still there?’, or thinking ‘What did they say?’, then the assumption is that the ‘fault’ is a poor quality service from the CSP”, says Amr. “Of all the aspects of a mobile network, voice quality is the strongest and most obvious indicator to an end-user of the quality of the service. Customers’ expectations of voice quality remain high, whichever technology the digitalised sound is transmitted over.”
Voice technologies today
Today, whilst some traditional mobile voice calls are still carried over legacy circuit-switched networks, calls made over 4G and higher, and for all app-based solutions, these digitalised sounds are transmitted as Voice over IP (VoIP), Voice over LTE (VoLTE) and Voice over WiFi (VoWiFi), all of which enable cost-effective ways to transport voice. Having a single solution that can assess voice quality across all technologies, in an automated and efficient way is vital – and in the web of complex multi-system networks, that AI-centred voice call quality testing solution must also work with solutions from all vendors.
INOS
Digis Squared’s INOS AI tool is a vendor agnostic, multi-network-technology solution delivering automated assessment, testing and optimisation of networks, across all technologies.
INOS & voice call quality testing
“If one of our clients – a CSP, MNO, MVNO or regulator – wants to better understand voice call quality on a specific mobile network, then we use our INOS AI tool to analyse the data,” shares Amr.
Image 1: voice codec rate. INOS analysis, for a specific drive test, on one mobile network, in Cairo
Notes on Image 1: For a voice call to be transmitted over the mobile network, it must first be digitized and compressed. Various standardized compression technologies, or codecs, are used to efficiently transmit the data. This image shows data collected during voice calls during a drive test, and the key on the left shows the codec used.
Enhanced Voice Rate (EVS) and Adaptive Multi-Rate (AMR) are audio compression formats used in the transmission of voice calls – EVS is a super-wide coding standard developed for VoLTE, and AMR is the older standard developed for GSM and UMTS (3G), sometimes called HD+.
Image 1 shows the variation in codec and compression rate utilised during test voice calls, made during a drive test. The changes in codec and compression rate are caused by changes in network coverage and capacity during the coverage, and will have resulted in fluctuations in call quality.
“Using INOS, we can simulate a customer call using the voice quality test to produce unbiased, industry-recognized audio quality scores,” explains Amr. “This test can reveal a great deal about your customers’ experience, as well as the quality of service being provided by your carrier. It also takes minimal preparation to undertake.”
Vital to this test is POLQA – Perceptual Objective Listening Quality Analysis – the global standard for benchmarking voice quality of fixed, mobile and IP-based networks. Standardized by the ITU in 2011, it is used for voice quality analysis of VoIP, HD Voice, 3G, 4G/VoLTE and 5G networks.
“So, whilst drive testing with the INOS kit, we set up a voice call and then use our own hardware solution to inject a POLQA reference audio into the voice call from one side of the call, and from the other side, we record the call, and then compare it using the POLQA algorithm.”
“Given that the POLQA reference audio is 6 seconds long, to analyse this data, we must split our call into audio files that are each only 6 seconds long. To ensure very precise splitting of the audio file, we leverage our AI engine to find the beginning of specific words in audio files. This way, we can ensure that we are aligning the analysis with natural speech patterns, and achieve a more realistic analysis of the data.”
Image 2: MOS score with 6 second sampling. INOS analysis, for a specific drive test, on one mobile network, in Cairo.
Notes on image 2: using data from the same drive test shown in image 1, now the data has been analyzed by INOS, and split into 6 second chunks, aligned with the start of specific spoken words in the audio file.
In telecoms, the Mean Opinion Score (MOS) is a numerical measure of the overall ranking of the quality of voice and video sessions. In image 2 above, we can see that on this journey, only a small minority of sections score the minimum 1 MOS (in black), and most of the call is green (MOS 3 and 4).
Image 3: MOS Score per call. INOS data, for a specific drive test, on one mobile network, in Cairo.
Notes on image 3: again using the same data as above, here the data is averaged out for specific calls, rather than 6 seconds chunks of a call shown in image 3.
INOS & silence within a call
“Silence within a call is a major problem with mobile phone conversations, and significantly impacts the customers’ perception of call quality. We’re all familiar with having to say ‘Can you still hear me? Are you there?’ whilst one of the people on the call is travelling in a car or bus,” continues Amr.
“To measure this, the Digis Squared team utilize our in-house AI capability within INOS to detect silence in voice calls, and analyse the percentage of silence.”
Image 4: Silence per call in seconds, INOS data, for a specific drive test, on one mobile network, in Cairo.
Notes on image 4: this analysis identifies areas where silence during the call was detected. Green indicates no silence, and in yellow, red and black are increasing amounts of detected silence.
INOS delivers automated voice quality reports, with customised KPIs, and actionable insights.
Amr concludes, “All our INOS reports can be fully customised, and are generated within 15 minutes of receipt of the data file, sent directly from the test devices in the field, over the air. What our clients find most useful is that not only are the reports conveniently formatted for immediate use, they also, thanks to our AI engine, clearly identify issues and provide actions which can be taken to address those issues. Data is no use without analysis, and the AI capabilities we have developed within INOS ensure that the analysis is fast, efficient and actionable.”
Find out more about INOS
INOS can be implemented as a public or private cloud, or on-premise solution, and is also available as a “Radio Testing as-a-service” model. Its extensive AI-analysis and remote OTA capabilities ensure speedy and accurate assessment of all aspects of network testing: SSV, in-building and drive testing, network optimisation and competitor benchmarking, across all vendors, network capabilities and technologies, including 5G, private networks and OpenRAN.
Image credits: all images, except first, copyright Digis Squared. First image (woman on phone) Gabriel Gonzalez
About Digis Squared
Managed Services, System Integration & Consulting. We transform telecom networks, deploy new technologies, and manage vendors, for network operators, service providers and regulators. Apply our vendor-agnostic expertise, automated AI-led tools and processes to transform your technical and commercial capabilities. We work with agility, deep experience, and our in-house cognitive tools to optimise and manage multi-vendor networks across all technologies. Headquartered in the UK, Digis Squared has offices in Angola, Egypt and UAE.
As part of the recent successful commercial launch of Africell Angola, the Digis Squared AI-tool INOS, powered by Intel® Xeon® Scalable Processors, was used as a key part of the network cluster testing and acceptance process. Digis Squared’s Key Account Manager for Africell, Ahmed Ma’moon, shares more.
“Digis Squared manages the entire end to end Managed Services for the new Africell Angola mobile network*. Ahead of the commercial launch on 7th April, and working closely alongside our partners, we ensured that the network was tested robustly before launch.” [*Read more about that, here.]
“Whilst the capabilities of our vendor-agnostic tool, INOS, are extensive, in the pre-launch phase of the Africell Angola project, its major role was in field optimization following the SSV (Single Site Verification) and sites acceptance phase. The team used INOS devices out in the field in vehicles to collect network performance data for all live sites. We were able to optimise our resources too – thanks to the ability to remotely update scripts, we didn’t need to send engineers out into the field; the INOS kit can be driven to a specific location and along a predefined route by anyone, and the data is automatically uploaded into the cloud immediately.”
What is a mobile network cluster?
Mobile network coverage is often drawn as a honeycomb-like pattern of neatly meshing hexagons.
In reality, the coverage is not neatly tessellated hexagons, but very irregular shapes, due to the landscape, buildings and other features, and coverage from adjacent cells may overlap, or there may be some thin gaps. Mobile network planning engineers allocate different frequency bands (also called channels) to neighbouring cells – this helps to minimise interference even when coverage areas overlap slightly. The group of cells on different bands is known as a cluster.
INOS for Cluster Optimisation
Ahmed continues, “Working from Digis Squared’s offices, our engineers were able to control and update scripts remotely, push revised routes to drivers, and review data live in the cloud during the tests. During the pre-launch phase we ran field measurements using INOS for Luanda province clusters and sub-clusters, undertook the analysis to identify coverage issues, implemented optimisation changes live on the network, and then re-tested and benchmarked the results against the initial data.”
Above: Example INOS dashboard for field measurements used for analysis and optimization
“INOS’ AI-capabilities ensure that analysis of vast amounts of data is completed very rapidly – within 15 minutes of uploading data – so we were able to assess the results, implement fixes and re-run the tests very swiftly.”
Above: Sample coverage analysis algorithms in INOS
Comprehensive pre-launch testing to optimise for post-launch excellence
“INOS was a vital tool for us in the pre-launch field optimization activities for Africell Angola, to ensure best network coverage and performance, and excellent user experience after launch. INOS helped to speed up the field optimization process for all Luanda clusters, and complete the work in advance of the scheduled launch date.”
INOS & Intel
“INOS was a vital tool for us in the pre-launch field optimization activities for Africell Angola, to ensure best network coverage and performance, and excellent user experience after launch. INOS helped to speed up the field optimization process for all Luanda clusters, and complete the work in advance of the scheduled launch date.”
INOS can be implemented as a public or private cloud, or on-premise solution, and is also available as a “Radio Testing as-a-service” model. Its extensive AI-analysis and remote OTA capabilities ensure speedy and accurate assessment of all aspects of network testing: SSV, in-building and drive testing, network optimisation and competitor benchmarking, across all vendors, network capabilities and technologies, including 5G, private networks and OpenRAN.
Yasser Elsabrouty, Digis Squared Chief Business Officer and Co-Founder, shares updates on the Private Network approach Digis Squared is undertaking for clients, and developments in the pipeline. But first, what is a Private Network?
What is a Private Network?
Built specifically for individual businesses or organisations, these dedicated LTE or 5G networks have typically been envisaged for mission-critical or highly secure environments. Increasingly however, businesses are deploying them to ensure robust coverage and capacity, reinforce intellectual property protection, and deliver commercial independence from the major network operators and CSPs within their physical business campus-environment.
Private Networks can be deployed in many different shapes and sizes, using various mix and match combinations of spectrum, applications and other factors.
Deployment
Dedicated, on-premise networks for both radio access network and core network
Hybrid use of some public mobile components plus dedicated on-premise components
Spectrum
Industrial: in some countries, regulators allocate specific licensed spectrum (Germany and Japan for example)
Shared: regulators allocate spectrum which is shared by multiple stakeholders, under license
Public: MNOs or CSPs lease part of their licensed spectrum to an enterprise for a fee
Unlicensed spectrum: assigned by the regulator, non-exclusive, free-to-use
Digis Squared & Private Networks
Digis Squared provides end-to-end System Integration services across multiple technologies including RAN, 4G/5G Core, Security, Messaging and Cloud, covering design, installation, testing, managed services operations.
Yasser Elsabrouty, Digis Squared Chief Business Officer and Co-Founder shares insights into the Private Network activity of the team.
“At Digis Squared, we are providing end to end Consultancy and System Integration services to build private networks to meet our customers’ needs. We work very flexibly with our clients – some know exactly what they need, and will ask Digis Squared to manage the System Integration, deployment and operational aspects of their pre-defined project. Others ask us to define all end to end elements: the deployment model, Spectrum, Radio details, Core network, orchestration and applications. Our teams are able to work with considerable flexibility, and according to the customer’s requirements and use cases to ensure they have and optimised Private Network which meets their needs.”
“We also offer predefined mix and match ‘off the shelf’ models that can be used as a starting point, and then adapt and deploy for the specific, customised and bespoke needs of each customer. We can take into account their specific requirements, including the size of the facility, devices deployed, which machines need to communicate with which departments, and other considerations. Some clients are looking at Private Networks to resolve specific coverage issues, or explore latency management for time-sensitive networks (TSNs). Whatever the scope of the Private Network, project our teams are enabling customers to scale up their businesses, serving more customers and satisfying the ever-increasing demand for private networks.ֿ”
“Using this mix-and-match method, we are developing profiles for different kinds of customers that can easily be implemented as needed, depending on size, area, number of sensors and cameras, or any other parameter. This approach will save time in deploying systems for customers, serving more customers while ensuring top-quality, optimised installations.”
If you would like to arrange a dedicated time to talk with the team, please get in touch, sales@digissquared.com
Digis Squared, independent telecoms expertise.
We transform telecom networks, deploy new technologies, and manage vendors, for operators, service providers and regulators.
Apply our expertise, automated AI-led tools and processes to transform your technical and commercial capabilities. We work with agility, deep experience, and our in-house cognitive tools to optimise and manage multi-vendor networks across all technologies.
Ahead of MWC22, Digis Squared’s CTO, Abdelrahman Fady, shares insights into ongoing development work on the Digis Squared suite of cognitive network testing and optimisation solutions, and declares tools “OpenRAN Ready”.
“The development teams in our Technology Centres in Cairo and London have been very busy enhancing the existing suite of cognitive tools to ensure that they are “OpenRAN Ready”. In advance of MWC next week, I’m really happy to share some of the key updates underway,” stated Abdelrahman.
Digis-One: technology and vendor agnostic Unified Fault Management
Can now connect with the big four legacy vendors and main OpenRAN vendors
Unifying all alarms from across all network systems and vendors into a single screen
OpenRAN solutions seamlessly integrated to a single view on one screen
iPM: intelligent technology and vendor agnostic network performance management platform
Able to connect to legacy vendors in addition to main OpenRAN vendors, and integrate their different performance files into a single unified database
Unify and visualize all these KPIs into a single coherent view on one screen, and represent them geographically
Single touch comparison between legacy vendors and OpenRAN vendors performance
INOS: technology and vendor agnostic intelligent network field testing and optimisation solution
New INOS OpenRAN testing and analytics module launched to easily identify gaps and differences in performance, L3 & L2 messages content and formats, network throughputs, and measure quality between OpenRAN and Legacy RAN sites
Forecasting of vulnerable areas after OpenRAN deployment
Automated acceptance report for new Open RAN sites
“If you want to learn about our new “OpenRAN Ready” cognitive network optimisation solution capabilities,” said Abdelrahman, “I will be pleased to meet with you in Barcelona next week, please get in touch!”
If you or your team would like to discover more about our capabilities, please get in touch: use this link or email sales@DigisSquared.com .
Addendum, 22 February 2022
LinkedIn ◦ Following Digis Squared’s earlier “OpenRAN Ready” solution announcement, Hazem Amiry, Regional Sales & Business Development Manager, shared that, “We’ve been working for some time now on an OpenRAN PoC with a very large operator in the Middle East, as the lead contractor on a project with a multi-awarding winning software-enabled OpenRAN solution provider. This project is enabling the team to learn first-hand the benefits of OpenRAN deployment, and ensure we are able to fully optimise our suite of cognitive tools to real-life, complex deployment issues efficiently.”
Digis Squared’s team complete INOS migration from local on-premises deployment to first cloud-based installation enabling “Radio Testing as a Service”, with Intel® Xeon® Gold 6338N Processor, in the Intel Lab.
Thanks to membership of Intel Network Builders, work undertaken in the Intel Lab has enabled the Digis Squared team to run INOS over the Intel® Xeon® Gold 6338N processor, in the first cloud-based installation of INOS. This work is the first step in our assessment of INOS as a cloud-based solution with Intel processors. Further work is planned with the Intel Lab team assessing other enhanced processors and benchmarking performance enhancement.
Yasser Elsabrouty, Digis Squared Chief Business Officer and Co-Founder said, “Thanks to Intel Network Builders and membership of Intel Winners Circle, INOS is now providing Radio Testing automation over the cloud, enabling “Radio Testing as a Service” over private or public cloud. The cognitive testing tool can seamlessly manage large amounts of data in a multi-tenant environment, providing full automation and real-time reporting.”
“Delivering INOS Testing as a Service over the cloud will increase efficiency, convenience and scalability, delivering the instant capability to run thousands of radio network tests from anywhere, anytime, in combination with smart automation, real-time reports and KPI deviation alerts. Digis Squared’s cognitive INOS tool just became a whole lot smarter!”
10nm technology, 32 cores, 64 threads, 3.6GHx max turbo frequency, full specification.
Benefits & observations
Running 25 INOS Radio Field Tests in the Intel Lab, the following enhancements were measured, and benefits observed,
1. Increased cores & threads
The Intel® Xeon® Gold 6338N processor enabled Digis Squared to setup 2 or more parallel INOS containers serving two (o more) different customer accounts. In the field, this extra capability enabled by the Intel® Xeon® Gold 6338N processor would mean that,
More copies of INOS modules can run together in parallel, providing higher processing capability
Lower response time and faster handling for APIs and web requests
Duplicating INOS running modules presents high availability
When assessing response time across all 25 tests, the results show that the Intel® Xeon® Gold 6338N easily handles the volume of data as data payload increases x2.5 over the 25 tests.
2. Max Turbo Frequency: the INOS platform receives high traffic bursts periodically, due to the nature of telecoms. The increased max turbo frequency of the Intel Xeon processor empowers INOS to handle these bursts without any probability of outage.
3. Intel® Turbo Boost Technology 2.0 Frequency: Increases the capability of INOS to receive big sudden bursts of requests, keeping stable progress and high performance (i.e. no delay on data retrieval, no delay on rendering data to maps and tables, and reduced time to prepare reports.)
4. Number of UPI links: INOS consumes a huge volume of processor capability and RAM. To optimise INOS performance, we are looking not just for capacity of the processor, but also how this processor chip interconnects with the rest of the system components. The Intel® Xeon® Gold 6338N presents better integration with various I/O devices reflecting in INOS performance, especially when handling large bursts of input data files.
5. Max memory size: For INOS, more memory means more concurrent users, more software threads running in parallel, and an increase in the number of docker containers running simultaneously. The increased max memory size indicated in this table will deliver at least three or more times the number of INOS containers when using the Intel® Xeon® Gold 6338N.
6. Intel® AES-NI & Intel® Trusted Execution Technology: INOS SW runs on sensitive client data, and this capability will save data from any corruption and violation trials.
Conclusions & next steps
Having successfully completed this first assessment with the Intel Lab, the Digis Squared team are confident in the deployment of INOS as a cloud-based solution utilising Intel® Xeon® Gold processors, delivering optimised performance and enhanced speeds.
Yasser concluded, “Further work is planned with the Intel Lab team assessing other enhanced processors and measuring performance enhancement, and then, mutual testing with Open RAN market leaders!”
If you or your team would like to discover more about our capabilities, please get in touch: use this link or email sales@DigisSquared.com .
The optimisation of capacity management in mobile networks is vital: too little capacity constraints revenue opportunities and impacts customer experience, but idle capacity risks high opex and under-performing investment in assets. Capacity management has always used mathematical modelling techniques to attempt to find the sweet spot, and optimise opportunities and costs. In the past, such predictions were based on historical data, but now AI enhancement of capacity management changes that. The deployment of network virtualization, 5G and network slicing requires the use of cognitive planning; it is vital that capacity planning models are able to assess a step-change in the volume of data points in real-time or near-real-time.
RAN Automation Architect and Data Scientist at Digis Squared, Obeid Allah Ali, describes how AI, automation and advanced analytics are being deployed to gain even greater network capacity planning efficiencies.
What exactly is machine learning, and why is it important?
Machine Learning (ML) is an application of artificial intelligence (AI) that enables computer programs to learn and improve over time because of their interactions with data.
It automates analytics by making predictions using algorithms that learn repeatedly.
Its easy self-learning technique, rather than rule-based programming, has found widespread use in a variety of contexts.
So, whether it’s making life easier with navigation advice based on predicted traffic behaviour, assessing large amounts of medical data to identify new patterns and links, or warning you about market volatility so you can adjust financial decisions, AI and ML technology has permeated many aspects of our daily lives.
The power of prediction machines
In simplified terms, prediction is the process of filling in the missing information. It takes the information you have, often called ‘data,’ and uses it to generate information you don’t have. Most machine learning algorithms are mathematical models that predict outcomes.
How will machine learning impact businesses?
There are two major ways that forecasts will alter the way businesses operate.
At low levels, a prediction machine can relieve humans of predictive activities, resulting in cost savings, and for example removing emotional bias.
A prediction machine could become so accurate and dependable that it alters how a company operates.
How big is the growth in mobile connectivity?
Above: from GSMA “The State of Mobile Internet Connectivity Report 2021” [3], their most recent report
Some further statistics on the growth in mobile data, from the same GSMA report [3],
global data per user reaching more than 6 GB per month – double the data usage for 2018
94% of the world’s population covered by mobile broadband network
By the end of 2020, 51% of the world’s population – just over 4 billion people – were using mobile internet, an increase of 225 million since the end of 2019
And from [4] GSMA Mobile Economy 2021 report,
By the end of 2025, 5G will account for just over a fifth of total mobile connections.
Capacity and performance of mobile networks
The rapid growth of mobile traffic places enormous strain on mobile networks’ ability to provide the necessary capacity and performance.
To meet demand, communications services providers (CSPs), mobile network operators and their suppliers need a range of options, including more spectrum, new technology, small cells, and traffic offloading to alternate access networks.
To meet commercial business objectives, mobile network operators are under pressure to maximize the utilization of existing resources while avoiding capacity bottlenecks that reduce revenues and negatively influence end-user experience.
Additionally, network operators have to assess risk, contractual SLAs (especially in the context of MVNOS who utilise their network, and corporate contracts), the total cost of ownership, and the impact on customer experience, perception and brand.
Radio Access Network costs are estimated to be 20% of the opex cost of running a network [1]. And the impact of opex on network quality correlates strongly with increased ARPU and reduced churn; when network quality is highest, service providers benefit from a higher average ARPU (+31 %) and lower average churn (-27%) [2].
Finding the perfect balance of capacity, quality, efficiency and cost – not too much, not too little – is complex and dynamic.
Capacity forecasting for mobile networks
The Digis Squared team have developed machine learning algorithms and decoders that can, based on network activity, decode how User Traffic Profiles are changing. With the deployment of 5G and network slicing techniques, modelling network usage patterns and customer behaviour and predicting future demand becomes immediately far more complex – the only way to successfully model this will be with AI.
Detecting a problem
We detect anomalies in cells in the existing network, plus highly utilized cells, using machine learning and a design approach algorithm based on several reported KPIs. We use this information to distinguish what requires immediate attention from what should be monitored for proactive action. Using multivariable modeling techniques, that is, assessing multiple KPIs across each cell, enables us to have a highly nuanced model, optimising all available capacity.
Forecasting
Operators must be able to estimate the required traffic capacity for their mobile networks in this competitive climate to invest in extensions when they are truly needed, and deploy the most cost-effective solution, while maximizing investment and maintaining good network quality. In this phase of the development of the model, we will discover future troublesome cells to guide our approach and actions using predictive models.
AI enhancement of capacity management: what’s next?
Today, we use an open-loop control system to apply our AI methods. However, as predictive model accuracy improves, we anticipate transitioning to a fully automated Self-Organized Network (SON) – enabling closed-loop network management with self-planning, self-configuration, self-optimization, and self-healing – system in the near future.
In conversation with Obeid Allah Ali, RAN Automation Architect and Data Scientist at Digis Squared.
If you or your team would like to discover more about our capabilities, please get in touch: use this link or email sales@DigisSquared.com .
AI-native network slicing for 5G networks. Successful 5G deployments rely on the use and integration of many other new technologies. In this blog post, Tameem Sheble RAN data scientist at Digis Squared, and AI for 5G researcher, takes us on a deep dive through the close integration of 5G with AI, and how network slicing is vital to deliver the 5G vision.
The view from 2015: the 5G vision
New telecom technologies take years to be discussed, agreed and defined in standards. Previous wireless technologies have been designed and architected for one major use case: enabling mobile broadband.
During the development phase of 5G standards, by means of reaching consensus and aligning expectations, the International Telecoms Union (ITU) defined their framework and overall objectives of the future out to 2020 and beyond, in a vision document. As part of this workaround 5G architecture, they considered three distinctive, cutting-edge service verticals – enhanced mobile broadband, massive machine-type communications and ultra-reliable and low-latency communications – and their usage scenarios and opportunities for communications service providers (CSPs).
Above: The original vision for 5G. Diagram from “IMT Vision for 2020 and beyond”, published 2015 [1]
Network Slicing (NS) is one of the 5G key enablers. It’s a technique that CSPs can use to satisfy the different needs and demands of the 5G heterogeneous verticals, as illustrated below, using the same physical network infrastructure.
AI-native network slicing for 5G networks
Network slicing enables the virtual and independent logical separation of physical networks. It’s a technique used to unlock the value of 5G networks, by opening the possibilities for customer-centric services based on the demand on the network while managing cost and complexity. Consequently, vendors and standardization communities consider 5G NS a key paradigm for 5G and beyond mobile network generations.
Although the 5G NS process brings flexibility, it also increases the complexity of network management. The introduction of AI into the 5G architecture, AI-native, is motivated by the vast amount of unexploited data and the inherent complexity and diversity that requires AI to be deployed as an integral part of the overall system design. Although the rising temptation is to rely on AI as a pillar for managing 5G network complexity, in practical terms, AI and 5G are indivisible. They tend to converge from an application perspective, and they become two halves of a whole. AI’s value relies on 5G; for example; critical data-driven decisions need to be communicated with ultra-low latency and high reliability.
AI as a potential solution to network slicing
Let’s turn now to addressing AI in a nutshell for the management of the complex sliced 5G network, a complexity that relates to decision-making towards efficient, dynamic management of resources in real-time. CSPs need to leverage the use of the vast volume of data flowing through the network in a proactive way, by forecasting and exploiting the future system behaviour.
The management lifecycle of a network slice consists of four main phases,
Preparation
Instantiation
Operation
Decommissioning.
Many researchers have proposed AI solutions that underline the first 3 phases, as the decommissioning phase doesn’t involve management decisions. Admission control and network resources orchestration are some of the key slice management functions that need to make slice-level decisions to meet their requirements, while simultaneously maximizing the overall system performance.
Looking at this in more detail, this is achieved by controlling a massive number of parameters as a result of uncovering complex multivariate relationships that are related to each other in time, geolocation, etc. Proposing an AI solution must be done case by case depending on problem formulation and framing, algorithmic requirements, the scarcity and type of data and the operational time dynamics.
AI for network slices admission control (Phase 1)
Admission control – during the slice preparation phase – is a very critical decision-making control mechanism, it ensures that the requirements of the admitted slices are satisfied. During this control mechanism, a trade-off between resources sharing and KPIs fulfilment needs to be tackled. The decision on how many network slices run simultaneously, and how to share the network infrastructure between those slices, has an impact on the revenues of the CSPs.
The trade-off is further complicated by variables that alter over time, which makes the optimization of revenue based on admitted slices a difficult task. This is where Deep Reinforcement Learning (DRL) approach comes into the picture.
In a nutshell, the DRL algorithm has to learn the arrival pattern of network slices and make, for example, revenue-maximizing decisions based on the current system utilization and the anticipated long-term revenue evolution. Once a network slice is requested, and based on the system current utilization, two separate neural networks are in charge of scoring the two actions (i.e., accepting or rejecting the request), where each score represents the revenue associated with each action. Based on the difference in scores, the action corresponding to the higher revenue is selected; the algorithm interacts with the system and evaluates the accuracy of the forecasted revenue through a loss function. This value is then feedback to the corresponding neural network to perform weight update, so that the algorithm starts converging to a global maximum and performs better in the subsequent request iterations.
AI for network resources orchestration and re-orchestration (Phase 2 and 3)
After the successful admission, slices must be allocated sufficient resources in such a way that the available capacity is used in the most efficient way that minimizes the operational expenses (OPEX). The trade-off here is between under-provisioning that leads to Service Level Agreement (SLA) violation, and over-dimensioning thus wasting resources.
CSPs need to be proactive by forecasting, at a slice level, the future capacity needed, based on previous traffic demand, and consequently timely reallocate resources when and where needed. This is where the Convolutional Neural Network (CNN) architecture comes into the picture for time-series forecasting.
Legacy state-of-the-art traffic time-series forecasting models focus on forecasting the future demand that minimizes some symmetric loss (e.g., mean absolute error), that treats both under- and over-prediction equally. But this type of legacy approach doesn’t consider the risk of under-provisioning and SLA violation – it is useless for 5G deployment!
Researchers argue that a practical AI-native resource orchestration solution has to forecast the minimum provisioned capacity that prevents SLA violation. The balance between over-dimensioning and under-provisioning is therefore controlled by the CSPs, by introducing a customized loss function that overcomes the drawbacks of the “vanilla symmetric losses”.
The AI literature proposes the use of 3-dimensional CNN architecture over the recurrent neural network (RNN) architecture – which is considered the state-of-the-art algorithm for forecasting time-series data – in order to exploit and uncover spatial and temporal traffic relationships.
The future for AI in 5G and beyond
Whilst general AI limitations are now well known – trustworthiness, generalization and interpretability – exploiting AI to assess and manage complex decisions is vital for the smooth operation of 5G networks. And as network technologies continue to grow in complexity and capability, AI will clearly be necessary as a pillar technology for future-generation zero-touch mobile networks
In conversation with Tameem Sheble RAN data scientist at Digis Squared, and AI for 5G researcher.
If you or your team would like to discover more about our capabilities, please get in touch: use this link or email sales@DigisSquared.com .
With an ever-growing volume of wireless network traffic produced inside buildings, network design and performance must be evaluated from within buildings. In this blog post, we assess the growing need for indoor coverage and the impact of the pandemic, plus Digis Squared RAN and Software Solution Architect and Trainer, Amr Ashraf, describes how in-building coverage testing without an engineer on-site can be undertaken, with INOS.
The ever-growing importance of in-building coverage
Often quoted research (1) estimated that “approximately 80 percent of wireless data traffic originates or terminates within a building”. Anecdotally, that figure is far higher now. Lockdowns and work-at-home mandates of the Covid-19 pandemic, plus the growing need to digitally maintain contact with friends and family are sure to have driven this even higher.
The pandemic has generated, and increased, specific needs for wireless connectivity indoors,
• Switch to working at home • Increase in voice traffic and video conferencing/communication, gaming and streaming traffic as we stay connected online at home to friends and family • Apps handling proximity detection/tracking and alerts about infected contacts • Tele-medicine: urgent care assessments and consultations, updating families unable to visit, remote assessments and advice, maintaining safe care-homes for the elderly and hospice patients
Even aside from the pandemic, the explosion in social media and mobile-centric content generation and consumption has dramatically increased the volume of mobile data consumed indoors.
But if indoor coverage is poor, then this impacts both operator revenue, and, perhaps more critically, brand loyalty and churn, as the need to connect now, indoors is far higher than any remaining loyalty consumers (and businesses) have for an operator’s brand.
Testing wireless connectivity inside buildings
Digis Squared RAN and Software Solution Architect and Trainer Amr Ashraf shares insights into the challenges and solutions for testing indoor coverage.
“Indoor network testing presents its own set of challenges, not encountered when undertaking traditional outdoor drive-testing. These indoor challenges include everything from gaining physical access to the site, to collecting as much relevant benchmarking data as possible in a single pass, and determining whether solutions provide data uploads to the cloud and data processing in a timeframe that enables a technician to test and troubleshoot network issues in one visit – if there is poor network coverage indoors, this may impact the speed at which we can assess the results!”
“Indoor testing today utilises smartphone and tablet applications, with all equipment packed discretely into a backpack-based test solution for indoor network testing. This approach has led to the number of walk-testing options for interior settings significantly expanding in recent years. Then, with detailed plans or architecture drawings of the building, and an efficient walking route planned out, a team member can be tasked with wearing the back-pack, starting the app, and walking through the route.”
“As mobile network operators and communications service providers have concentrated more and more on in-building coverage, they often encounter a problem: they are unable to gather all the measurements they need in a single test walk.”
“Critically, it’s no longer necessary for the person walking the route inside the building to be an engineer. The technical assessment can be undertaken by skilled staff, remotely, ensuring your scarce engineering resource can be deployed efficiently across many projects. When an issue is detected during the building walk-through, the network can be optimised remotely – and because the INOS testing and analysis takes just 15 minutes from receipt of data, our aim is to ensure that we can re-test and re-walk the improved area as part of a single visit to the building.”
“One of our clients described testing a distributed antenna system at a major convention centre that served four wireless operators using three different wireless technologies across multiple channels, for a total of about 20 different operator/technology/band combinations, each of which required a separate measurement. A complex configuration, but one which is quite common in large business-focused buildings.”
“The indoor network testing for this project was carried out with INOS using the Digis Squared proprietary backpack-based In-Building Test Suite. In contrast to user-equipment-based backpack testing systems, which are typically restricted by the number of devices and technologies that can be tested concurrently, the INOS solution depends on a scanning receiver intended for multi-technology networks. That is to say, we are not constrained, there is no technical limit on the number of devices we can use in the testing.”
Undertaking an in-building survey
“The INOS backpack is a multi-technology integrated solution for testing and measuring multi-device mobile networks. Whether it’s for conducting an indoor or outdoor walk or cycle test, or an outdoor drive test, the INOS backpack offers a small design for portability and simple movement. Data interaction is accomplished by using a WiFi hotspot to link an Android tablet (as a controller unit) to test terminals. A powerful solution for portable multi-network benchmarking, supporting up to 20 test terminals and a scanner for testing and measuring simultaneously.”
“The measurements are transferred to the cloud for additional data management and processing, and the testing is undertaken according to the test plans given by the controller unit.”
“We use an Android tablet to operate all of the testing equipment in the backpack, connected via Wi-Fi to the test phones, which are also integrated into the backpack. This configuration gives the technician complete control over the devices, enabling them to add pinpoints throughout the in-building walk as data is collected, or repeat sections immediately after dynamic network optimisations are implemented.”
Part of the INOS interface showing the controlling tablet view, with information about the connected testing devices and their status.
Case study
“Recently, a global Tier One mobile operator used the Digis Squared INOS backpack testing technology to investigate networks in Cairo. They wanted to undertake benchmarking on their own network, as well as those of their main rivals, both inside buildings and outside. Data speeds, latency, and web browsing durations were among the main performance parameters they tested with INOS, as were dropped calls and RSSI signal levels. Once captured, the INOS data collected was sent over the air to the INOS cloud-based platform for immediate automated analysis and presentation via an analytics dashboard.”
INOS data captured during in-building testing inside the “Mall of Arabia”, in Cairo, Egypt
INOS advantage compared with traditional approaches
“One of the primary advantages INOS delivers is our very quick analysis and reporting capability. After just a few minutes of testing, we can practically immediately provide a comprehensive report with all KPIs.”
“The vast majority of network coverage-related complaints occur indoors, traditionally necessitating an engineer to visit the customer’s house or office to undertake a network evaluation – this legacy approach results in high operational costs, and scheduling delays in identifying the issue.”
“Let’s compare that with the INOS solution. Anyone can be tasked with capturing data with INOS, no technical knowledge is needed to carry the backpack around the building or location of interest. It is not necessary to divert a skilled engineer out in to the field to capture data – on some projects we’ve tasked Uber drivers with taking an INOS bag around a pre-defined route, and returning it to us, or asked a member of the admin team to cycle a route with the INOS backpack. The INOS system can even be utilised to submit a self-service complaint to skilled RF optimization specialists in the office, who can then undertake an initial assessment remotely using the INOS kit controller and web application. And of course, as only one person is needed to take the bag in a car, or walk it around a building, the solution is also Covid-19 safe.”
“INOS also enables operators and suppliers to capture data in the field remotely, analyse the data, determine which issues can be solved remotely, and then efficiently schedule and resolve problems which can only be addressed in the field .”
“If you want to know more, we’re always happy to chat through what we can do to help you. Meet us at MWC22 or let’s fix up a call online.”
In conversation with Amr Ashraf, Digis Squared RAN and Software Solution Architect.
If you or your team would like to discover more about our capabilities, please get in touch: use this link or email sales@DigisSquared.com .
Key Advantages of INOS
Tablet: From the Android picture gallery, users can quickly import tiny to huge floor plans (of any form of structure).
Floor plans and data are kept in the cloud and may be shared with co-workers.
Ease of use, testing, and interior navigation can all be undertaken by non-technical personnel.
In real time, test data is uploaded to the INOS Cloud server.
Post-analysis: results can be mapped onto indoor floor layout, with a web-based dashboard.
All-in-one mobile solution with device, network, and service benchmarking capabilities.
From the standpoint of subscribers, it provides extensive network performance statistics.
INOS is used as the test device, allowing for a single investment to be used for multiple purposes.
Test procedures, data processing and analysis can be fully automated, resulting in increased overall efficiency, and optimised consistency.
