Thursday, April 28, 2011

The best way to deliver video to mobile?

An article form Gigaom's Ryan Lawler recently caught my eye as a potential game changer in the mobile video delivery market and got me thinking.

As a user of mobile video, I get often frustrated by latency, stop and go, long loading times, etc...
I can't imagine that content owners can be too happy about video experience being mangled for their customers.

If we look at the trend for content control and compression, the early way to reduce content size, at the time of dial up and mobile narrow band was compression and encryption.
The main problem was that it required a client. Clients were difficult and expensive to update, fix, maintain over the air, so the business model died.

Fast forward 10 years later, seemingly everyone walks around with a smartphone or wants one. One of the major catalysts for the smartphone adoptions has been the apps explosion. Apps are nothing more than standardized clients that can be downloaded, upgraded over the air...

My YouTube app is a client with an embedded player that allows me to navigate and select content in a predefined manner. It is similar to my browser experience but at the same time a little different. In a browser, Google, Microsoft, Apple control my user experience. In the YouTube app, YouTube controls my user experience.

In the case of mobile CDNs, there is an additional potential benefit to reduce cost of delivery by hosting, caching, delivering content as close to the user as possible.

If you add to this the trend towards using P2P technology for video delivery,YouTube, Dailymotion, Akamai, Limelight... could decide to encrypt and tunnel the traffic, to try and keep control of the user experience.

I will examine the potential implications of this scenario in a future post and look at possible alternatives.

Wednesday, April 27, 2011

LTE: it's a Little Too Early part 2

My analysis:
On mass market penetration, we are still far. Out of the 140 network commitments, 17 are launched so far, in many cases at the city level. I could not find many figures in term of LTE subscriber numbers except for Verizon, and I think in itself it is a sign. Additionally, out of the nearly 100 LTE devices out there, only 6 are smartphones. 7 are tablets, the rest are notebooks, modems, dongles, routers, etc...
The technology will certainly take off but we are far from mass market adoption.

On ease of use, I can't really comment. I haven't had yet my hands on an LTE device or accessed an LTE network. The peak speeds advertised make my head swoon and certainly raise questions about unplugging fixed line access. I am hoping LTE in itself won't negatively impact the current trend of intelligent devices that allow me a better connection to my network and apps.

Interoperability, once again in my mind is going to be the big issue here. One of the key advantages of LTE for an operator is to be able to reuse spectrum and bands allocated to older technologies (think GPRS, EDGE, 1X...). This result in about 20 paired bands frequency division duplexing that can be used by a given operator to roll out the service. Additionally, LTE can be implemented  in time division duplexing (TDD), adding another 12 bands.
Operators have bid or have been allocated frequencies to operate LTE in their country. In many countries, like in the US, operators have different bands, which will force device manufacturers to have multi-band devices. Multiply that at the scale of the 600+ network operators worldwide and you understand that it will be a long time before we get a world phone on LTE.

Additionally, you have to think that each band requires a specif radio module. It comes with an impact on antenna and battery life. Economically and form factor-wise, it is difficult to have a smartphone today that will accommodate more than 5 bands.

This situation will lead to a variety of short term hindrance:
  • Operators asking manufacturers for devices supporting as many as 12 bands to cater for international roaming. This will increase price of devices, decrease battery life and slow down adoption of technology for business users and travelers.
  • The only cost-effective devices in the short term are going to be stationary ones or devices tied to one operator. they will accelerate the transition from fixed broadband to mobile broadband.

In conclusion, I see LTE time frame for mass market to be ~7 years. I have no doubt it will be successful but the current growth in data traffic might choke current networks before LTE has a chance to flourish if the above issues are not addressed.

Tuesday, April 26, 2011

LTE: it's a Little Too Early part 1

As we are starting to see the first announcements about LTE deployments and their promises of speed comparable only to those experienced on the Autobahn, I wanted to dig a little bit into the maturity of the technology and its mass market time frame for applicability.
My conclusion? It's still a Little Too Early for LTE .

According to GSA(the Global mobile Suppliers Association), the current number of commitments from network operators to implement LTE is 140, as of March 24th2011. GSA is a vendors-led association, so the forecast are somewhat optimistic, but this is a respectable number of operators who have bravely started to invest in LTE.

When I traditionally look at the penetration capability of a new technology in the wireless ecosystem, I usually rely on a few indicators to gauge where the technology is in its adoption cycle.

  • Mass market penetration: I would look for at least 30% of a given population having implemented the technology and having devices in their hands capable of using it. That could mean 30% of network operators in the case of LTE, but more importantly, I would look at 30% of the subscribers of a given operator having an LTE subscription as a sign of maturity.
  • Ease of use, ease of adoption: In this case, I look at what the barriers to entry are for subscribers or operators to acquire and use the technology. The cost of license auctions, the relative cost of increasing HSPA+ network density and investments. What will Femtocell or data offload do to the demand? 
  • Interoperability: This is a key criterion, that is overlooked time and again in wireless technology introduction. This is about interoperability between the devices themselves, the devices and the network, backwards compatibility, the networks between themselves, roaming, interconnectivity, etc... This has been a consistent issue that has plagued the adoption and success of many wireless technologies over the last decades (WAP, MMS, PoC, IMS...).

In my next post, I  will provide my opinion on the challenges of LTE and the time frame for its mass market adoption.

Wednesday, April 20, 2011

Welcome to {Core Analysis}

Hello and welcome to {Core Analysis}.

This blog will examine the myths and realities of mobile broadband. It will feature comments and opinions on vendors, analysts and operators announcements surrounding the transformation of mobile networks, as video and all IP communication takes a more and more important role.

As the name points out, {Core Analysis} strives to look beneath the surface and to offer an informed view on the direction market, technologies and vendors are taking.

Here are a few subjects you can expect me to address :

  • Why you shouldn't hold your breath for LTE just yet
  • Video optimization myth and realities
  • Tariffing and charging strategies in a video world
  • Policy management in a video world
  • Mobile CDN and video delivery
  • Over the top strategies for video content providers
  • Caching strategies for video

As for me, I consult and advise operators, vendors and investors on technology and vendors qualification.
You can find out more about what I do at

Don't hesitate to comment or drop me a note!