Network-2020 : Discussing Role of LTE-U/LAA,Indoor Small Cells and IOT with Mr. Oscar Bexell

Hi Oscar thanks for giving us time and agreeing to share your expert views with us. You have been associated with the telecom and technology industry for more than a decade now, you have seen the evolution and development very closely. Recently we have seen dynamics of the industry changing very rapidly. Specially for countries as India 3G was not even fully deployed and the 4G/LTE evolution came. Also community and vendors are already taking about 5G and Network 2020.

Q 1. Techplayon: How do you see industry and network changing w.r.t 2020?

A 1. Oscar: Mobile operators have always handled indoor traffic by building outdoor networks, and when running in to capacity problems they handle that by buying more spectrum, change the subscriber cap, adding more macro cells or by throttling users. Traffic is sky rocketing and spectrum gets jammed, so the only way forward is to deploy radio networks indoors. I’ve been part of real-case studies showing fantastic results on UE battery consumption, peak rates, average rates, spectral efficiency, uplink interference and mobility when coordinated macro/indoor systems are deployed on the same frequency as the outdoor macro in a congested area.

Outside-in coverage as a business model is as doomed as oil in the car industry. The future is in-building systems, very likely small cell based. The reason why small cells haven’t taken off in other markets than maybe China this far is the multi-operator requirement which today means you need four-five parallel small cell systems per building to get all operators represented. That is of course too costly and messy. So what I expect will happen between now and 2020 is that market forces solve this issue and get neutral host or MVNO small cells closer to large scale takeoff.

That’s the only sustainable way forward. In the US this has already started to happen with CBRS. The Google Fi model where a virtual operator switch between other operator networks in the background is another possible approach. National roaming between local operators is a third. Somone will crack the nut, and when it happens I expect to see system integrators rolling out the small cell systems, not carriers who often lack both trust from enterprises, relevant experience to do the job and ability/will to cooperate with key stake holders.

Q 2. Techplayon: 5G seems to be true Heterogeneous network being deployed with licensed and un-licensed networks working in harmony, do you see Wi-Fi integral part of the same or you see LTE-U/LAA and Mutlifire being strong threats to the same?

A 2. Oscar: One of the main benefits of LTE over Wi-Fi is that two or more overlapping cells can perform simultaneously, allowing for state of the art mobility and QoS handling which you really want on a phone or, in the future, on a pair of glasses. Since MuLTEFire requires carrier sense and listen before talk, it removes this ability so it’s DOA to me. Similar with LWA which will fail because mobile operators never get in on enterprise Wi-Fi networks and because there is no business case in connecting Wi-Fi traffic back to a mobile core.

Wi-Fi will like today keep taking the bulk of “fixed” data even if LTE-U/LAA is rolled out along with LTE small cells for macro offload purposes. This is not a threat to Wi-Fi as such, but to any Wi-Fi business model with the word “mobile” or “offload” in it. In a longer perspective, 5G or something similar is well suited to take over all wireless traffic in enterprises. Exactly how that plays out will be more clear in a few years time.

Q 3. Techplayon: NB-IOT vs Unlicensed LWPAN such as SigFox/ LoRA, who wins in long term? Sigfox and LoRA having initial advantage of getting early to market, many companies (Comcast, Machine Q) which are not Telcos see this as their opportunity to deploy networks. Do you see once NB-IOT comes full flow would Non Telcos find it hard to fight with Telcos?

A 3. Oscar: Many IoT projects start off in small scale and with a low budget where noone wants to spend money on access. Along with already available chipsets this has driven adoptation for LoRa(WAN) and Sigfox and non-MNO operators working in that field. The main challenge for these technologies is of course that they operate in license free ISM bands that are extremely narrow and prone to interference from for example LTE800. When traffic goes up, quality will naturally go down just as in other ISM bands. NB-IoT provides QoS, often better coverage and other benefits such as better downlink capabilities for upgrades, so that’s my long-term bet for large, commercially related IoT services in dense areas like city centers. There are a few other interesting contenders also that could gain momentum for low-band operators, for example Weightless.

Q 4. Techplayon: What is the best use case of IOT you have come across?

A 4. Oscar: Very dependent on vertical. Most manufacturing/industrial companies have been dealing with IoT for ages and have great solutions. Since their solutions operate “in-house” they don’t have to think too much about standardization and security. Instead they can build Intranet of Things. On a more public scale, my favourite use cases are actually those where you manage to avoid sensors completely. Like a parking company that tracks the position of all cars in a city and also knows exactly when they leave the parking spaces. Much better and cheaper solution than tagging every parking lot with a sensor. You just fetch the data via an API and build services on top.

Q 5. Techplayon: So, what your current role and organisation is all about?

A 5. Oscar: I’m working as IoT Solution Architect at the largest IT consultancy company in Scandinavia. IoT is a lot more about understanding business operations and humans than technology. Of course you need to know your numbers and how things work, but that information is useless if you don’t spend most of the time listening to people. I think that’s the biggest difference compared to traditional “IT” where you can knock on the door, sell a server and an access point and then walk away with a service contract. To be successful within IoT, you really need to understand your customers and their pain points. That often involves conversations with a lot more people than just an IT manager. I find myself in meetings with CEOs, HR managers, receptionists and a lot of other random people who are not working on the IT side and it’s always great discussions. I have also participated in workshops where we let various people from the same company discuss topics that help them bridge the gaps within their organizations. Extremely valuable. I believe that’s how all kinds of tech business should be driven. Listen to people, listen to customers, analyze the business, find problems and THEN apply technology to improve.l