Thursday 24 June 2021

O-RAN Introduction for Beginners


Having been writing about Open RAN for a while, I thought it was important to make simple beginners tutorials for O-RAN. As my full time job* is with a company that is heavily involved in Open RAN and O-RAN, I had an insiders view for doing this project. 

I am making a series of videos for Parallel Wireless to help the industry become familiar with the technology and terminology. The playlist is embedded below:

Four of these are ready and more will be added as and when I get some time. Here is a summary of the videos available. Some of them also have a corresponding blog that I am linking below.

  1. Introduction to O-RAN Philosophy: This explains the basics of O-RAN and how O-RAN is transforming the mobile networks industry towards open, intelligent, virtualized and fully interoperable RAN.
  2. Introduction to O-RAN Timeline and Releases: This part looks at important timelines from the O-RAN Alliance, understand the O-RAN Software Community (OSC) and the role it plays in O-RAN, and finally, looks at the O-RAN Open-Source Software releases.
  3. Introduction to O-RAN Architecture: This part looks at how the basic OpenRAN architecture is evolving into the O-RAN Alliance based Intelligent, Virtualized and Fully Interoperable RAN. It starts with a high-level ORAN architecture and then delves into details of Service Management and Orchestration (SMO), Non-Real-Time (Non-RT) RAN Intelligent Controller (RIC), Near-RT RIC and O-Cloud.
  4. O-RAN Technical Steering Committee (TSC) & Workgroups: This part looks at O-RAN Technical Steering Committee (TSC) & Workgroups (WGs). The O-RAN TSC decides or gives guidance on O-RAN technical topics and approves O-RAN specifications prior to the Board approval and publication. The TSC consists of Member representatives and the technical workgroup co-chairs, representing both Members and Contributors. Within the TSC, there are 10 work groups, 4 focus groups, Open-Source Community and Minimum Viable Plan Committee. These have all been discussed within the video.
  5. O-RAN Workgroup1: Task Groups and Deliverables: This part looks at O-RAN Workgroup#1 (WG1), its task groups and sub-task groups and finally the deliverables produced by WG1.
  6. O-RAN Alliance Workgroup 2 and Workgroup 3: Specifications and Other Deliverables: This part looks at O-RAN Workgroup#2 (WG2) and Workgroup#3 (WG3) deliverables.

I am hoping that I will be able to do a few more parts and add a lot more information to the basics so a handy resource is available for anyone interested. Feel free to add links, suggestions, etc. in the comments below. 

Related Posts:

*Full Disclosure: I work for Parallel Wireless as a Senior Director, Technology & Innovation Strategy. This blog is maintained in my personal capacity and expresses my own views, not the views of my employer or anyone else. Anyone who knows me well would know this.

Monday 21 June 2021

3GPP Standards on Edge Computing

A sub-set of 3GPP Market Representation Partners hosted a 2-part webinar series in April 2021 looking at edge computing for industry verticals and on-going standardisation work in 3GPP. The first part write-up is available here. The webinar was attended by a mix of organisations from both verticals and the telecommunication industry, helping to share a common understanding on edge computing. 

The webinar brought together top experts at the 3GPP plenary level, SA2 (Architecture) and SA6 (application enablement and critical communication applications) for a deep-dive into how 5G and related standards can help harmonise and enable technologies like edge computing and artificial intelligence to work together much more efficiently. 

The webinar was co-chaired by Georg Mayer, 3GPP SA Chairman and Stephanie Parker, Trust-IT and Vice-chair of the 5G-IA Pre-Standardisation WG with the John Favaro, Trust-IT and member of the 5G PPP Automotive Working Group. 

The webinar was attended by a mix of organisations from both verticals and the telecommunication industry, helping to share a common understanding on edge computing.

This video embedded below is the recording of the webinar on edge computing held on Thursday 22 April 2021 part 2 - 3GPP Standards on Edge Computing as an educational deep dive to help industry verticals gain a better understanding of an evolving landscape. It gives key insights into 3GPP standardisation work on edge computing with an overview of the main activities taking place within SA (System Aspects and Architecture). Presentations and panel discussions zoom in on the network layer with SA2 Architecture and on the application layer for vertical enablement with SA6 Application Enablement and Critical Communication Applications. The panel discussion with SA TSG, SA2 and SA6 chairmen sheds light on the role of artificial intelligence from both the network and application perspectives, underscoring the vital importance of industry verticals in the standardisation process to meet their specific requirements in 3GPP as a truly global initiative.

PDF of presentations as follows:

Global5G has a summary with main takeaways and poll findings here. The following is from there:

Main Takeaways

  1. 5G will help technologies like edge computing and artificial intelligence to harmonise and enable them to work together much more efficiently.
  2. 3GPP Release 17 is foundational for edge computing but more will come in future releases given its importance in mobile communications and as we gradually move beyond 5G. The webinar was therefore a timely deep-dive into today's landscape. 
  3. Artificial Intelligence and edge computing can both serve as building blocks but in different ways: 
    • Network layer perspectives: AI can further optimise edge computing applications.
    • Application layer persepctives: Edge computing can be a building block for AI, e.g. offloading limited capabilities from the device to the network.
  4. Global initiatives like 3GPP can help reduce regional fragmentation, drive convergence and enable network-compliant rollouts that benefit the ecosystem around the world.
  5. As a global initiative, 3GPP is well placed to build on its strong relationships and collaborations with ETSI MEC and GSMA. 
  6. It is absolutely essential that industry verticals get involved in 3GPP working groups, which is where key activities take place and where their requirements should be channelled. It is also important that verticals understand how their seemingly specific requirements could be relevant to other sectors. Being part of 3GPP is a complex but highly rewarding experience. It does not need to be a life-long commitment.

Poll Findings - Participant Viewpoints

Do you participate in standardization on edge computing?

Interestingly most respondents do not take part in any standardisation initiatives. Hence the webinar series was an opportunity to highlight the many activities taking place and encourage participants to get involved. Those that do take part mostly contribute to 3GPP and other forums (29%) like ETSI (SDO) and industry associations like 5GAA and 5G-ACIA as some of the early movers on edge computing. Beyond 3GPP, a smaller number of respondents (11%) contribute to ETSI and other forums such as 5GAA and GSMA and the same amount (11%) are involved in other forums.

How important do you think coordination on edge computing standardisation is?

Coordination on edge computing standardisation needs to be prioritised with 65% of respondents saying it's vital and another 33% saying it's quite important. Only 1 respondent said it's not needed. An important output via the 5G-IA Pre-Standardisation WG and supported by panellists and organisers (5G-IA, 5GAA, 5G-ACIA and PSCE) would be a user-friendly guide on edge computing standardisation to help stakeholders navigate the landscape. 

Do you see a need for new areas of standardisation for edge computing?

Findings from this poll are particularly interesting as we have a close split between those that think more standardisation work is needed (47%) and those that don't know (43%) with just 10% saying it's not needed. Webinar organisers have come up with two possible explanations. On the one hand, we may be looking at a fragmented landscape that would benefit from more unification, also from an architecture perspective. On the other hand, organisations looking at the landscape may simply be overwhelmed by the dverse activities taking place. They may also have new applications sitting on top of the network but are not sure if they need to be standardised. Practical guidance could go a long way in clarifying this uncertainty. 

Again, a quick guide on edge computing standardisation could be a useful output, highlighting also the good cooperation already taking place as an important step in the right direction. 

You can see Part 1 of this webinar here.

Related Posts

Saturday 19 June 2021

Edge Computing - Industry Vertical Viewpoints


A sub-set of 3GPP Market Representation Partners hosted a 2-part webinar series in April 2021 looking at edge computing for industry verticals and on-going standardisation work in 3GPP. The webinar was attended by a mix of organisations from both verticals and the telecommunication industry, helping to share a common understanding on edge computing. 

The first webinar brought together experts from the 5G Automotive Association (5GAA), the 5G Alliance for Connected Industry and Automation (5G-ACIA), Edge Gallery, ETSI Multi-access edge computing (MEC) and the Automotive Edge Computing Consortium (AECC) to highlight opportunities and updates on how diverse market sectors can benefit from offloading data at the edge of the network. Further insights came from interactive discussions and polling with participants. This webinar is part of a 5G user webinar and workshop series designed for industry verticals co-hosted by 5G-IA, 5GAA, 5G-ACIA and PSCE as Market Representation Partners of 3GPP.

This video embedded below is the recording of the webinar on Tuesday 20 April on edge computing - part one, giving an educational deep dive on industry vertical viewpoints. 5GAA (5G Automotive Association) gives an overview of its white paper, use cases and upcoming trials for Cellular-V2X in the automotive sector. Edge Gallery shows how it is supporting the Industrial Internet of Things with its 5G open-source solutions and application development support. ETSI MEC explain its common and extensible application enabling platform for new business opportunities. 5G-ACIA (5G Alliance for Connected Industry and Automation) describes new work on the applicability of 5G industrual edge computing within the associaton. The Automotive Edge Computing Consortium (AECC) brings insights into how it is driving data to the edge.

Bios and PDF presentations as follows:

Global5G has a summary with main takeaways and poll findings here. The following is from there:

Main takeaways

  1. The webinar was an excellent deep-dive into the edge computing landscape highlighting on-going work in automotive, manufacturing and the Industrial Internet of Things, as well as standardisation work in ETSI and open-source approaches. 
  2. It illustrated the value of edge computing with strong signs coming from industry in terms of growing interest and adoption roadmaps. There is an impressive number of initiatives across the globe embracing edge computing, with examples of cooperation globally as seen in 5GAA, 5G-ACIA, AECC and ETSI MEC. 
  3. Industrial automation, digital twins and infrastructure control among the main drivers for growing demand. 
  4. Collaboration on edge computing is essential and will become even more important as applications increasingly move to the edge. Continued discussions are needed to have greater clarity at multiple layers: business and technology, SW and HW. Collaboration can also support efforts to educate consumers and businesses, both key to uptake and achieving network compliant rollout.  
  5. The collaboration underpinning the 3GPP MRP webinar series is an excellent example of how we can intensify joint efforts across the ecosystem working towards convergence and ensuring RoI, e.g. for telecom investments. 

Poll Findings - Participant viewpoints

Where would you position your organisation in terms of implementing edge computing?

Only 16% of respondents already have a commercial strategy in place for edge computing while 26% are starting to develop one. Therefore 42% are expected to have one in short term. 30% are at early learning stage to understand market opportunities and 28% are exploring its potential. 

In which verticals do you expect the first implementations other than automotive?

The automotive sector is an early mover in edge computing, as testified by 5GAA and AECC presentations in the webinar with both having published studies and white papers. 5GAA is planning trials in 2021 in various locations globally so another webinar on this topic in 2022 would be helpful. After automotive, manufacturing is expected to be the next sector to implement edge, as testified by the 5G-ACIA presentation. All three associations are market representation partners of 3GPP, with 5GAA also contributing to standardisation work. In the 5G PPP, 5GCroCo (cross-border automotive use cases) has contributed to standardisation activities of both 5GAA and AECC. Gaming, AR/VR and media is the next sector expected to adopt edge computing. 

What are your top 2 priority requirements for edge computing? 

Low latency is the top requirement for most respondents (33%) followed by interoperability and service continuity (both on 20.5%) with transferring and processing large volumes of data and very high reliability in joint third place (both on 12.8%). It' will be important to see how many of these requirements feature in early deployments as not all of them will be there at first rollout. The poll also shows how requirements combine together, e.g. 2 priority requirements: Low latency + very high reliability; Interoperability + Service continuity; Interoperability + Low latency; 3 requirements: Interoperability + Service continuity + Transferring and processing large volumes of data and 4 requirements: Interoperability + Service continuity + Low latency + Transferring and processing large volumes of data. 

Part 2 of this webinar is available here.

Related Posts

Monday 14 June 2021

A mmWave Special Cell in Open RAN Environment

NR RRC signaling messages exchanged for establishing a 5G radio connection, in particular the NR RRC Reconfiguration and NR RRC CG Config messages, contain a parameter called "SpCellID", which refers to the Special Cell ID. 

The concept of the Special Cell already exists in 3GPP LTE Advanced standards. Here a Special Cell is set of physical cells with same or different carrier frequency and physical cell ID (PCI) that overlap in a certain geographical area and thus, are combined for data transmission to/from UEs located in this area.

This concept now also gains high importance for 5G NR mmWave spectrum and here is why:

Many 5G mmWave radio transmitters can only handle a maximum bandwidth of 100 MHz, but the radio sector shall be covered with total bandwidth of e.g. 600 MHz. To achieve this six mmWave radio transmitters are installed in parallel at the same spot covering the same footprint. 

Each transmitter is identified on the radio interface by its own dedicated NR ARFCN (carrier frequency) and PCI. Thus, from UE point of view the sector is covered with 6 dedicated NR cells that all together form a Special Cell.

When a UE gets radio resources assigned in this 5G sector one of  the 6 cells is the Primary Cell, which NR CGI (Cell Global Identity) is then used as Special Cell ID in layer 3 signaling messages. All other cells act as Secondary Cells.

In an Open RAN environment the F1AP protocol allows perfect analysis of the SpCell resource allocation since it contains the SpCellID as well as all SCellIDs to be setup in the call. 

If the gNB-DU fails to allocated resources for a particular Secondary Cell this will also be signaled together with a failure cause value on F1AP as illustrated in the figure below. Also radio link failures occurring within the Special Cell will be signaled on F1AP including a cause value that provides deeper insight than  protocol causes seen on X2AP (in case of 5G NSA connections) or NGAP (in case of 5G SA connections). 

(click on image to enlarge)


Thursday 10 June 2021

Nokia veterans Harri Holma and Antti Toskala explain 5G Basics

An online conference on 5G is currently going on. 'Backed by 5G: Technology for Impact', is hosted by Start North as a part of the Aalto University Summer Course 5G Hack the Mall.

It is a two-week online conference which gathers the industry experts, entrepreneurs and policymakers together to discuss, present and question how 5G will affect our society, economy and everyday life. We want you to join the change by offering a chance to learn from the best, to start or strengthen your journey to expertise in 5G. With the support of our partners, you have the possibility to listen and get your questions answered by the global 5G leaders from the leading companies, academia, NGOs and public institutions, which all are daily involved with changing the world and enabling change with the latest technology.

The current conference features couple of Nokia experts who are well known in the industry for their books on the mobile technologies. Dr. Harri Holma, Fellow, Nokia Bell Labs, spoke on "What is Good to Know About 5G Technology Components". His talk is embedded below:


The second talk is by Dr. Antti Toskala, Fellow, Nokia Bell Labs, on Radio Access (5G Physical Layer). His talk is embedded below

Slides are shared to the 5G Summer School participants. If you are keen to get your hands on the slides, please email: hello@startnorth.com. You can watch all the videos from the event on the Start North YouTube channel here.

Related Posts

Monday 7 June 2021

TSDSI's Low Mobility Large Cell (LMLC) Requirements in 5G


Back in November 2020, ITU completed the evaluation for global affirmation of IMT-2020 technologies. Three new technologies were successfully evaluated by ITU and were found to conform with the International Mobile Telecommunications 2020 (IMT-2020) vision and stringent performance requirements. The technologies are: 3GPP 5G-SRIT and 3GPP 5G-RIT submitted by the Third Generation Partnership Project (3GPP), and 5Gi submitted by Telecommunications Standards Development Society India (TSDSI). 

I have explained in earlier videos that 5G-SRIT  and 5G-RIT corresponds to Non-Standalone and Standalone respectively. 5Gi on the other hand is an updated version of 5G-RIT designed mainly to improve rural coverage. 

TSDSI announced this as follows:

TSDSI’s 5G Radio Interface Technology named as “5Gi” has cleared the rigorous processes of  International Telecommunication Union (ITU) and has been approved by the SG5 of ITU as a part of Draft Recommendation M.[IMT-2020.SPECS] in its meeting held on 23rd November 2020.

5Gi, the first  ever Mobile Radio Interface Technology contribution from India to become part of ITU-R’s  IMT recommendation, went through a rigorous evaluation process of the ITU-R working groups over the past 3 years before getting the approval.

This standard is a major breakthrough for bridging the rural-urban digital divide in 5G deployment due to enhanced coverage. It enables connecting majority of India’s villages through towers located at gram panchayats in a cost effective manner. It has found support from several countries as it addresses their regional needs from a 5G standpoint.

The standard will now be circulated by ITU to member states for adoption and approval. Specifications are expected to be published by ITU in early February 2021.

TSDSI thanks its members, the Department of Telecommunications, Govt. of India and its partners for their support over the last four years in helping get this standard reach the final stage in ITU.

In a keynote address presented to the 2020 IEEE 5G World Forum plenary session, Radha Krishna Ganti from TSDSI discusses rural connectivity challenges in India, Low Mobility Large Cell requirements, benefits of implementing LMLC for rural coverage, and internet ecosystem updates. His talk is embedded as follows:

TSDSI explains their 5Gi technology as follows:

TSDSI standard fulfils the requirements of affordable connectivity in rural, remote and sparsely populated areas. Enhanced cell coverage enabled by this standard, will be of great value in countries and regions that rely heavily on mobile technologies for connectivity but cannot afford dense deployment of base stations due to lack of deep fibre penetration,  poor economics and challenges of geographical terrain. The International Telecommunication Union (ITU), a UN body that is setting requirements for IMT 2020 (aka 5G), had earlier adopted the Low-Mobility-Large-Cell (LMLC) use case proposed by TSDSI as a mandatory 5G requirement in 2017. This test case addresses the problem of rural coverage by mandating large cell sizes in a rural terrain and scattered areas in developing as well as developed countries. Several countries supported this as they saw a similar need in their jurisdictions as well. TSDSI successfully introduced an indigenously developed 5G candidate Radio Interface Technology, compatible with 3GPP Technology, at the International Telecommunications Union (ITU) in 2019 for IMT 2020 ratification. The RIT incorporates India-specific technology enhancements that can enable larger coverage for meeting the LMLC requirements. It exploits a new transmit waveform that increases cell range developed by research institutions in India (IIT Hyderabad, CEWiT and IIT Madras) and supported by several Indian companies. It enables low-cost rural coverage and has additional features which enable higher spectrum efficiency and improved latency.

While technically this sounds interesting and as discussed in the talk, would make sense due to a large market like India, there are other solutions that are already possible that probably may make this redundant.

As someone who worked with the rural communities to bring coverage in hard to reach areas, small cells and In-band backhaul was one such solution to improve coverage in not-spot areas. Examples of that here and here. Relays are other option that don't cost much but can bring coverage quickly, at a much lower price.

Typically, in practice, the cells easily reach 10km radius. In theory this distance can be as much as 100km. Last year, Australian operator Telstra and vendor Ericsson announced that they have successfully managed to increase the range of an LTE cell from 100 km to 200 km. So, we can already have large cells with existing 4G/5G cells. 

Facebook connectivity is working on SuperCell concept, a Wide-Area Coverage Solution for Increasing Mobile Connectivity in Rural Communities. Details here. NGMN published a paper on Extreme Long Range Communications for Deep Rural Coverage. Details here.

Finally, we also have 5G Integrated Access and Backhaul (IAB) that can be used for backhauling and solving backhaul issues. They will end up playing a role in rural areas as well as dense urban areas eventually.

Let me know what you think.

Related Posts:

Monday 31 May 2021

5G User Plane Redundancy


We looked at the 5G Enhanced URLLC (eURLLC) earlier. One of the ways to improve reliability is to have redundancy in the user plane. This can use different approaches like: 

  • Duplicating N3
  • Adding a secondary gNB using Dual connectivity
  • Introducing another UPF
  • Two anchor UPFs

In fact they are all built on top of each other so you can decide how critical are your user plane redundancy needs. 

I came across this short video from Mpirical embedded below that covers this topic nicely. In case you want to refresh your 5G Core Network architecture, jump to our old tutorial here.

Related Posts:

Monday 24 May 2021

ITU Standardization Bureau on Machine Learning for 5G


Last year I blogged about Global ITU AI/ML 5G Challenge on the theme “How to apply ITU's ML architecture in 5G networks".  The grand challenge finale happened in December. All the recording and presentations are available here.

Back in October, Bilel Jamoussi from ITU presented a keynote to the 2020 IEEE 5G World Forum plenary session where he addressed the challenges of applying machine learning in networks, ITU’s ML toolkit, and ITU’s AI/ML in 5G Competition. IEEE Tv shared the presentation only in April so the competition part is a bit outdated. It does nevertheless an interesting 20 minute talk.

ITU Recommendation Y.3174, Framework for data handling to enable machine learning in future networks including IMT-2020 is available here.

Related Posts

Monday 17 May 2021

3GPP RAN Plenary Update and Evolution towards 5G-Advanced

(click on image to enlarge)

ETSI recently held a webinar to provide a 3GPP RAN Plenary update by Wanshi Chen, Senior director of technology at Qualcomm Technologies, who was appointed as the RAN Chair not too long back. The webinar video is embedded below. The following is from the 3GPP summary of the webinar:

Wanshi Chen acknowledged that Release 17 - the third release of 5G specifications - has been under pressure due to COVID-19 restrictions, but despite making the move to e-meetings, he reported that the group’s experts have managed to ensure positive progress towards the freeze of the RAN1 physical layer specifications on schedule, by December 2021.

This is to be followed by the Stage 3 freeze (RAN2, RAN3 and RAN4) by March 2022 and the ASN.1 freeze and the performance specifications completion by September 2022 – On the timeline agreed back in December 2019.

This staggered timeline has been made achievable with careful planning and management, demonstrated to the webinar viewers via a complex planning schedule, with a slide showing the array of Plenary & WG meetings and Release landmarks - Interspersed with a series of planned periods of inactivity, to allow delegates some relief from 3GPP discussions.

Wanshi Chen noted that the efficiency of e-meetings has not been comparable with physical meetings, in terms of getting everything done. To compensate for that, the companies involved have planned two RAN1 meetings in 4Q21 and two meetings for each of the RAN working groups in the 1Q22. He observed: “We will monitor Release 17 RAN progress closely and take the necessary actions to make sure we can get the release completed on time.”

Release 18 Planning

Looking forward to Release 18 and the start of work on 5G-Advanced, Chen outlined the schedule for an online RAN workshop from June 28 – July 2, to define what will be in the release. The workshop will set the scene for email discussions about the endorsed topics for consideration. The work will culminate with Release 18 Package Approval, at the December 2021 Plenary (RAN#94).

The high-level objective of the workshop will be to gather company proposals in three areas:

  • eMBB driven work;
  • Non-eMBB driven functionality;
  • Cross-functionality for both.

Wanshi Chen concluded that during the Release 18 planning process, some capacity must be kept in hand; keeping around 10% of WG effort in reserve, for workload management and to meet late, emerging critical needs from commercial deployments.

The following Q&A topics were covered, along with the time stamps:

  • The effect of the pandemic and eMeeting management schedules and tools (19.25).
  • Balance between commercial needs and societal needs, emergency services, energy efficiency, sustainability (21.20).
  • The importance of the verticals in the second phase of 5G – With 5G-Advanced. How will this Rel-18 workshop compare in scale with the 5G Phoenix workshop in 2015? (23.00)
  • The job of the Chair is to be impartial…but Wanshi guesses that Antennas, MiMo enh., Sidelink, Positioning, xR, AI machine learning…. could come up in Rel-18! (26.15)
  • Will 5G-Advanced have a strong identity & support? (30.05)
  • The potential for hybrid meetings – No clear answers yet, but we have learnt a lot in the past year.(34.35)
  • The link between gathering new requirements and use cases in SA1 and RAN work and RAN1’s role in focusing these needs for radio work. (40.10)
  • Software-ization of the RAN. Do you see more open RAN work coming to 3GPP? (44.18)
  • Machine type communications and IoT – Where is IoT going in 3GPP RAN? (47.01)
  • Some thoughts on Spectrum usage from a 3GPP point of view, is that difficult to fathom for non-experts? (52.00)
  • Can Standards writing become more agile, less linear? (54.00)

If you want to get hold of the slides, you will have to register on BrightTALK here and then download from attachments.

Signals Research Group has a short summary of 3GPP RAN #91 electronic plenary held in late March. It is available to download after registration from here.

xoxoxoxoxoxo Updated later, 07 June 2021 oxoxoxoxoxoxox 

5G-Advanced logo is now available as shown above. Guidelines on how to use the logo is available on 3GPP here.

Related Posts:

Thursday 13 May 2021

Anomaly Detection and other AI Algorithms in RAN Optimization


Yesterday I watched this very inspiring live chat that I would like to recommend to anyone who is interested in how machine learning techniques (aka "AI") can help to optimize and troubleshoot the Radio Access Network.

 [The real contents of in the video starts at approx. 42:00 min] 

My key takeaways from this fireside chat are: 

Verizon Wireless has enough data (100… 500 time series KPIs per cell) that they use to feed anomaly detection ML algorithms and this generates a huge number of alarms, but only a few actionable outputs. 

The “big elephant” (Nick Feamster) is to identify if these alarms indicating real problems that can and have to be fixed or if they just indicate a new behavior of e.g. a new handset or a SW version that was not present in the training phase of the ML algorithm and hence, its pattern is detected as a new “anomaly”. 

For Bryan Larish (Director Wireless AI Innovation, Verizon) the “big open problem” is “that it is not clear what the labels are” and “no standard training sets exist”. 

[For more details watch the video section between 52.00 min and 57:32 min and listen to Bryan’s experience!] 

In most cases Verizon seems to need subject matter experts to classify and label these anomaly alarms due to “the huge diversity” in data pattern. 

According to Bryan only for very few selected use cases it is possible to build an automated loop to fix the issue. Especially the root causes of radio interference are often mechanical or cabling issues that need manual work to get fixed. 

All in all it is my personal impression at the end of the session that anomaly detection is currently a bit overhyped and that the real challenges and problems to be resolved start after anomalies are detected.

Nevertheless, as Bryan summarizes: “ML is a very, very powerful tool.” 

However, strategically he seems not to see a lot of value in anomaly detection by itself, but rather: “Can we use machine learning (results) to change how we build networks in the future?”