Pages

Pages

Tuesday, June 16, 2015

WiFi Offload.... or Onload?


One of the things I get annoyed by regularly is the mis-characterisation of a lot of WiFi use on smartphones / tablets as "offload".

In my view, true "offload" is a term only applicable to data which would otherwise have transited a cellular network, and which has been deliberately pushed to WiFi at a public hotspot, with the intervention of a service provider.

It is very different to "private WiFi" use, where the user has unilaterally decided to connected to a network (for example at home or office) of their own volition, or because a venue or other sponsor has made WiFi access available.

Care also needs to be taken about elasticity - user behaviour may change when on WiFi, even with "proper offload" - if the connection is cheaper or (more often) free/unmetered. At that point, a notional volume of traffic X that might be used on mobile might become 2X or 10X when on WiFi. This could be because of a shift in user perception ("Hmm, yes I will watch Game of Thrones streamed on my phone"), or it could be because an app-developer has created a different experience when connected to WiFi (eg auto-playing video, or enabling downloads of updates).  

[Sidenote: connectivity is not "all the same" - both users and developers make very different decisions when a device is on WiFi vs. 3G/4G. People & apps/OSs are aware of, and care about, the type of network which they're connected to. The notion that it doesn't matter is a core fallacy behind the notion of so-called "seamless" HetNets and integrated infrastructure] 

In other words, only a tiny fraction of smartphone WiFi use can be called "offload" with any reasonable definition. I'd estimate that it's well under 5% globally, and probably under 10% of phones' WiFi usage even on those networks which have extensive operator-driven offload implemented. For tablets, the numbers will be lower still, as the majority are non-cellular and therefore can never "offload", while even mobile-enabled ones are mostly non-activated or primarily used in static locations with private WiFi.

But there is another trend emerging in parallel to "real offload" that will make the numbers even more confusing.

In some cases, people or applications might deliberately switch to cellular from WiFi, for example if the WiFi network is congested, coverage is poor, or there are localised authentication problems. In other words, we will see "offload" from both cellular-to-WiFi AND WiFi-to-cellular. It may be that one direction of this gets referred to as "onload". It may also be that the WiFi-to-cellular onload is larger in volume. This would mostly driven by users' deliberate switching, but perhaps also by WiFi-primary policy clients on devices, for example from services provided by cable operators.

Takeouts from this:
  • Be skeptical of most alleged "WiFi offload" figures - they're usually nonsense
  • Most smartphone WiFi usage is private - traffic that would never have used cellular anwhere
  • Be aware that WiFi/cellular onload happens, as well as cellular/WiFi offload
  • Claims that "nobody cares which network they are on" are either ignorant or duplicitous
  • View all discussions of cellular/WiFi combinations through the lens of WiFi-primary users as well as cellular-primary viewpoints

Friday, June 12, 2015

Qualcomm's MuLTEfire is what LTE-U should have been, instead of LAA

Yesterday, Qualcomm rather quietly announced a project called MuLTEfire on its blog.

It describes it as "a new, LTE-based technology that solely operates in unlicensed spectrum, and doesn’t require an 'anchor' in licensed spectrum"

This is a very different proposition to the other type of LTE-U, called LAA (licence-assisted access), which requires a provider to "anchor" the service in a separate (licensed) band. That has proven very controversial in recent months, with fears that its coexistence with WiFi in the 5GHz band could prove damaging, with extra interference. There are claims and counterclaims there, with both technical and "moral" viewpoints.

But I've been critical of LAA for another reason - I think it is potentially anti-competitive, as it is only usable by operators that have (paid) spectrum for other LTE networks. It could be seen as a way of extending an oligopoly position into an adjacent marketplace, as inevitably its use in a band reduces theoretical capacity available to others, even if it behaves "politely".

My view of unlicenced-band cellular has been that it should be available to all to implement, in the same way that WiFi is. At least in concept, MuLTEfire is what I'd envisioned when I first thought about unlicenced 4G.  

(I'm not 100% certain, but I think I may have personally invented the concept of unlicenced-band LTE myself, as per this blog post from July 2008 . I also suggested SIM-free LTE a couple of months later) 

Fully-open unlicenced LTE has some rather interesting possibilities.  By decoupling LTE from the constraints of licenced spectrum - and, ideally, without a SIM card or with some sort of soft- or programmable SIM - then we could see a set of revolutionary new business models. For example, it would become possible for venues to offer "free 4G" to visitors, or for all sorts of novel "anti-roaming" propositions to be provided. We could also see true "private cellular" networks - which have already been proven in concept by the use of light-licensed GSM guard-bands and pico/femtocells in the UK and Netherlands.

Obviously, any company considering its deployment could equally-well use WiFi in the same places. But LTE-U in MuLTEfire might allow easier roaming, especially in devices which don't have SIM-based WiFi capabilities enabled.There are also all sorts of interesting options for hybrid MVNOs/MNOs, neutral-hosts for indoor coverage, and a bunch of other concepts I've got at the back of my mind.

In particular, given this is cellular technology, it is actually much more aligned with the notion of "seamless" connection than the WiFi is. I'm a deep skeptic of integration of WiFi with cellular, as it introduces too many compromises in terms of user choice and policy/preference conflicts.

Qualcomm's timing here is very interesting - the FCC has been asking for submissions about LTE-U / LAA, with the initial comments also due yesterday. And there's a big spectrum management event in Brussels next week - I'm presenting on Tuesday afternoon and will be mentioning LTE-U on a panel which also includes a Qualcomm speaker.

Now clearly, a lot depends on the details (eg IPR costs, whether the coexistence works as billed) and whether the project gets traction. But for a mobile-industry giant such as Qualcomm to even suggest a SIM-free variants of cellular is a major step forward, and one that I've been advocating for years.

Thursday, June 04, 2015

What's YOUR view of contextual communications?

In recent months, I've been drilling into the new "hot topic" of contextual comms. Martin Geddes & I are so enthused by the topic that we're running a workshop on June 15th in London (details here), and we're already considering follow-ups, maybe in the US later in the year.

We're combining both the "here & now" of context with a view on where we might be heading in the medium-to-longer term. Martin wrote a very forward-looking and provocative piece on the possible future recently (here).

I'm really interested in what "contextual communications" means to everyone else. There's no fixed definition at the moment, and I suspect that we're going to get an "Olympic Rings" multi-way Venn diagram. Some views of context will overlap, while others will be miles apart. For instance, I've seen or heard all of these described as Contextual Comms:


  • Sending web-form info to an contact-centre agent during "click to call"
  • Embedding video/telepresence into a robot
  • Using mic & speakers on a phone to map out a room acoustically & tweak the echo/noise processing
  • Use a media-server to analyse a caller's tone (eg angry vs. happy) or facial expressions, and adjust the experience or script for a salesperson
  • Using a device orientation sensor to work out if a phone is flat on a table, or help to the ear, and adust the UI accordingly
  • Using machine-learning and analytics to assess the best time to call someone
  • Mechanisms for indicating the purpose of a call
  • Embedding a call into a timeline or activity-stream interface for UC and collaboration, so it can be recorded, captured & seen alongside text commentary or speech analytics
I'm sure there are dozens more as well. I'm looking forward to distilling some sort of map or ontology, so we can collectively understand this new landscape a bit more clearly. Is it one thing with lots of variants? Or 5 separate trends with a little overlap?

Do YOU have a good example or definition of Contextual Comms? I'd love to hear from you, either via a comment here, or by doing an interview briefing.

And if you'd like to talk about it publicly, we're offering all the workshop attendees an chance to present or demo their view - basically an "open mic" section of the day to showcase their unique take on context.

If you'd like more detail about the event, or to get in touch separately about context, please comment,  see this page to book a spacea, or email information AT disruptive-analysis DOT com.

Monday, June 01, 2015

WebRTC & Contextual Comms may help so-called "OTT" apps avoid regulation



Last week, the Belgian authorities tried to claim that Skype is a telecoms "operator" and should comply with a similar set of laws to fixed and mobile telcos, especially around data-retention and lawful interception.

The Indian regulator TRAI has been undertaking a consultation about whether so-called "OTTs" should be somehow "regulated". (TRAI's consultation document is one of the most woefully-written and factually incorrect pieces of official literature I've ever read).

Slightly different but in the same general domain, the UK Government is looking at ways to limit the use of encryption and anonymity with Internet communications services, again to collect metadata. (Sidenote: I think some of the proposals are rather technically ignorant and will get sidelines - and it's worth noting the UK remains one of a dwindling set of places without either official ID cards, nor a requirement to register SIM cards).

A major principle that keeps cropping up is that regulators, telcos and governments assert that services such as Skype and WhatsApp are somehow "equivalent" to traditional phone calls or SMS, and therefore should attract similar regulation.

This also intersects with another set of regulatory pushes (notably by Telefonica's increasingly shrill & incoherent policy team) to try to force interoperability onto 3rd-party communications apps. Sometimes calling it "platform neutrality" this seems to be a transparent attempt to reduce competition from new voice/video/messaging apps by dis-allowing "walled gardens". Given that the only likely medium for mass interop is the PSTN (and E.164 numbers), this is a blatantly defensive move to ensure the old phone network remains at the centre in future. It's unworkable, but unfortunately the current EU Commissioners seem more keen than their predecessors to try to implement stupid/unworkable ideas from the telco lobbyists.

Yet this is all very rearward-looking. The most successful future communications apps are not going to be yet more "free standalone messaging" services that look like SMS or WhatsApp, nor "cheap generic VoIP calling" ones that emulate Skype. 

Those ships have sailed already. It's another reason why most telcos' "IP communicator" apps will fail, especially if based on lower-than-lowest common denominators like RCS.

Instead, any new winners are going to be unique in some way - features like disappearing messages (SnapChat), blending realtime 2-way voice with asynchronous (eg Talko), embedded voice/video as a secondary feature in other social or business apps (probably with WebRTC), or with a strong contextual-comms element (using the user's physical status or intended purpose).

The interesting thing here is that not only would these be differentiated but it would also seem impossible, or at least much harder, to claim (note: I'm not a lawyer) that these are "equivalent to the phone service".

I also think existing services need to assert their "non-equivalence" much more vehemently - and point to the lack of innovation in telephony and SMS over decades. 

Regulators should not be accepting telcos' arguments that they need to cross-subsidise network investments with profits from over-priced, near-obsolete services.

In the Skype case, I'd say that one option Microsoft has in Belgium is to ditch the interconnection to the PSTN, and possibly move to a video-only model. Both would indicate that it is not a "phone service" but something entirely different. Given that there is no successful telco video-calling service (nor, with RCS & ViLTE as proposals, will there ever be) it would be much harder for the authorities to claim equivalence.

A more interesting defence of Skype's uniqueness could come from analysis of the proportion of calls preceded by a messaging session. In my view, the user experience of Skype is very different to the PSTN, as it is not based around unexpected, interruptive calls, but is instead an "escalation" method of rendezvous and arrangement. You use presence, chat with IM, and then say "OK for a call?". That is different to traditional comms experiences.
 
In fact, I'd argue that designing a new service to be too unique & differentiated to meaningfully interoperate with the PSTN or SMS means that:

a) It stands a chance of success, against 100s of "me too" apps and installed bases of 500m+ for entrenched competitors.
b) It will be harder to capture with pernicious regulations and telco lobbying, as it's clearly something new, and not just a cheaper substitute for protected legacy services.

Having given this a lot of thought, I've reached the following conclusions:
  • New communications apps SHOULD NOT interop with phone calls (like SkypeOut or iMessage) if at all possible. If they do, they risk being classified as "similar" to regulated services.
  • Avoid using E.164 phone numbers as identifiers as possible, for similar reasons
  • Ensure that user behaviour and features are very clearly distinct from traditional "calls" or SMS, to the degree that "interoperability" is meaningless
  • Concentrate on communications-as-a-feature rather than as a standalone service, unless it is a completely unique and differentiated format. WebRTC is the likely key enabler. (Click here for my research report)
  • Create "clear blue water" between legacy phone-calls / messages by using contextual communications capabilities that cannot be replicated in traditional telco service. Focus on how, and why a specific instance is occurring, and use external data to help reach the desired outcome.
  • Where there is a specific business need for interop, avoid using 3GPP/telco standards wherever possible (SMS, SS7, IMS, RCS) and use the web or proprietary mechanisms instead.

As a reminder, I'm running a workshop on Contextual Communications on June 15th in London, along with Martin Geddes. Sign up here.