LTEAsia remains one of the best, if not the best event, to meet with most of the region’s CxOs. Its not too big, and not too small. Numbers appeared slightly down on 2 years ago when I last attended, but the quality of attendees remains as high.
I was chairing the Signaling Focus day, the Keynotes on Day 1, the 5G Summit, and judging the Innovation Accelerator. And also running TADMeetup Singapore, which is described in this TADSummit weblog. My review of LTE Asia is limited to the streams where I was involved. When the slides are available I’ll update this weblog with some of the highlights. My overall take is NFV/SDN wasted much time in presentations and discussions, when we have much more clear and present problems: VoLTE interop (because of standardization failures), curtailing the diameter signaling explosion (because of standardization failures), solving the root cause of standardization failures, and focusing on specific service opportunities in for example roaming and enterprise services.
Signaling Focus Day
I opened the day with a frank assessment of where we are in diameter signaling, setting out the challenges, cutting through the marketing drivel, and focusing on specific opportunities, see slides below.
Asia is now the top market for diameter signaling, given the growth of LTE (APAC now has the #2, 3 and 4 markets) and a lead in VoLTE over Europe and the US. But the actual size of the diameter signaling market by 2018 is a significant question, it could be in the range of $1-2B, it needs to deliver on the marketing drivel if its to achieve the upper figure.
The failures in standardization (diameter signaling explosion) remain unaccounted for and un-remedied. The market is dominated by Oracle (thanks to Tekelec and Acme Packet acquisitions), with Ericsson and Huawei joint global second place, but Huawei and Oracle are much closer in APAC.
There are clear opportunities in zero rating, QoS for enterprise (but it needs to be end to end) and network optimization. The only problem is its all getting wrapped up in the crazy NFV/SDN BS. I’ve discussed the Industrial Embarrassment of NFV marketing previously. At this conference and also at Service Delivery Innovation Summit there are too many vague NFV presentations, rehashing old ground.
After my opening pitch poor Jason Emery from Oracle had to follow with the usual Oracle pitch, but he did very well in defending his position, dodging some of the weaker aspects of the Oracle position (for example turbo button and OSPs (Online Service Providers) paying for QoS) and thought on his feet. The following panel session was excellent and included Jason from Oracle, Sean from Ulticom, Bill from Sonus, and Jorgen from Tieto. There was a dynamic and insightful discussion that got the audience engaged. We only got through 1.5 of the 6 questions I prepared. The general agreement was that the diameter signaling explosion can be avoided, but not with the current standards, so urgent change is required. Then on the second question we only dealt with the VoLTE impact on diameter signaling not the impact of CA (Carrier Aggregation) and HetNets. With VoLTE there was much specific discussion on the implementation options within the specification (which is considered too loose) and now VoWiFi to VoLTE is managed in practice. The VoLTE implementation issues were also a discussion point at the IMS World Forum in May and little progress appears to be made.
Jim Machi from Dialogic gave an interesting presentation on the phases of diameter signaling deployment and how for tier 2 telcos they may consider consolidating the DSC (Diameter Signalling Controller) and NFV orchestrator, see slides at end. We wrapped up the day with a number of discussion tables focused on specific topics such as: the impact of VoLTE on diameter signaling, and diameter signaling and IPX roaming. These proved useful in proving a much deeper informal discussion within a more focused group, so were well worthwhile especially at the end of the day.
Overall, I think we need less focus on NFV, and much more specific focus on VoLTE implementation, emergency steps to avoid the diameter signaling mess, and specific discussion on real diameter signaling opportunities not wishful thinking marketing slides. That is specific roaming and enterprise opportunities, so telcos can leave with clear opportunities and the industry can focus rather than jabbering about NFV and SDN.
In the evening we had TADMeetup Singapore, which is reviewed in this TADSummit weblog. It was a joint event with MoMo Singapore, we had 85+ registered attendees turn up plus lots of gate crashers. You can see the pictures of the event at the TADMeetup Singapore site, it was a packed event focused on telecom application development – the stuff necessary to maintain telcos revenues.
Day One Keynotes and 5G Summit
As always LTE Asia has a strong line up of keynotes with Aileen Chia, Deputy Director General (Telecoms and Post), IDA; Kuan Moon Yuen. CEO Consumer Singapore, SingTel; Jin-Hyo Park, SVP, Head of Network R&D Center, SK Telecom; Herfini Haryono, CTO, Indosat, Indonesia; Patrick Scodeller, COO, M1 Limited; and Richard Tan, Deputy CEO, PT Smartfren, Indonesia.
Some of the take aways from the presentations were 5G is likely to be traditionally licensed, while the approach to TV Whitespaces will be unlicensed and with 180MHz of spectrum that opens the door to some interesting business opportunities. Kuan Moon showed it took 32 months for 3G to pass 2G subscriptions and 17 months for LTE to pass 3G. Its likely 2G will be switched off in 2017. LTE-A subscribers consume 35% more data than LTE subs. Jin-Hyp from SK Telecom share their trials of 5G achieving 3.6 Gbps. Pipes will be fatter and the latencies lower. Though the vision of the services consuming the capacity is similar to the slides shown on what would be consuming the data for 3G, 3.5G, 3.75G, 4G (LTE), 4.xG (LTE-A / CA).
One of the many aspects of weak thinking is the 50B IoT number is quoted like its all mobile network connected, most are not. Most are fitbits, watches, etc. which are all connected by bluetooth or WiFi to a mobile or fixed device. So it really should be called the internet of peripherals (IoP). Similarly on connected cars, cars are already already connected by mobile phones for several decades. The objection then raised is for the manufacturers to know the status of the engine. Has anyone noticed how reliable cars are these days? Just doing the oil change provides an opportunity for the car to be checked quite adequately.
We appear to think the rich and famous lifestyle will be endowed upon us all. It will in some ways, but generally the connected homes (which already exist in the pads of the rich and famous), and the high end cars (which are already connected) are not that likely to be the same for the rest of us, because of simple economics in what we can afford. Have you noticed food, education, healthcare prices are all increasing faster that your income? The PDI (Personal Disposal Income) for most people in developed markets is not increasing if you take out the fat-cat data. Income equality is diminishing, even nice to haves like a full broadcast TV package will be squeezed out to just internet access to help cover the children’s education, are are already seeing signs of this in the US.
Let’s face it self-driving cars can not be driven in many parts of the country, like my driveway, can’t be taken out in snow, could drive straight over a gaping pothole, and can not tell the difference between a pedestrian and police officer waving on the road. The list of practical limitations goes on, but I love how by 2018 it will all be solved and we’ll all have self-driving cars coz Google says so. I anticipate still driving my current car then. Let’s stop the wishful thinking and focus on what can make a difference in the next 2 years as we have some critical problems to solve as an industry.
The keynote panel was refreshing, fun, and insightful. Joseph Waring, now with the GSMA, did a nice article on the opening panel. We opened with a round of rapid fire questions. Some of the insights are:
- In developed markets 5G will happen before sometime between 2018 and 2020, while in emerging markets well after 2020;
- Carrier aggregation is more important than hetnets given their deployment challenges;
- 3G networks will still be around after 5G networks are rolled out;
- All the capacity being offered by LTE-A and 5G will not put them in competition with fixed networks in developed markets, while in emerging amrketings it will;
- Mixed views on NFV/SDN with a general consensus (though put diplomatically) the vendor marketing is unlikely to be delivered;
- Voice and SMS will remain chargeable services was a consensus view. I guess it becomes an accounting issue once everything is a subscription bundle.
I noticed recently, which quite shocked me, I’m down to less than 2 mobile calls connected per week (lots more attempts), while Vonage, Skype, Hangouts, and other OSPs eat up the lion’s share of my voice and video minutes. When I view my mobile bill as just a mobile internet access bill, its quite expensive.
The 5G Summit ran in the afternoon. Some of the highlights for me were Itsuma Tanaka of NTT DoCoMo’s vision of the 5G core (really a next gen core beyond EPC) with a clear indication that perhaps its time to move away from the 3GPP standards; David Williams from AsiaDigitalMojo focused on what 5G could mean to the user experience; and Rod Randall of SIRIS Capital’s Cloud Vision see slide highlights below (as well as his very practical view on the steps and challenges towards NFV).
We ran an Innovation Accelerator at the end of Day 1, where Tropo, Temasys, Telestax and Joogla pitched; with Rod Randal, Frank Bomers of Accelerasia, and myself as judges. Congrats to Tropo who won, just pipped the others, it was close. The slides are in the deck below.
While in Singapore I was unable to SMS local SG numbers reliably with the US Verizon #. I kept getting, “<Singapore phone number> is a landline #. Reply Y to send all TXT messages to this # as voice messages for 0.25/msg. + std msg fee. Details @ vtext.com, TexttoLandline”. The number was a mobile number, the message was not delivered, and I was charged for the service. So I got all the people who were not on Whatsapp onto Whatsapp (except one who is in the 4% of Singaporean 2G users) so we could reliably text.
As a closing note, with the roll-out of VoLTE we need to make sure it works well, I’ve grown increasingly dissatisfied with the quality of critical telecom standards. Any failure will adversely impact customer perceptions, and promote their willingness to use alternatives, which could set in motion a collapse in revenues as people seek simply a data only plan, and regulators react to customer outrage by force such plans at reasonable prices. That for me is the most critical issue we have to address coming out of LTE Asia – get VoLTE working.