Monday 11 June 2012

The 'Virtual' Femtocell and a competition for OTT Apps

Over the last few months we have been thinking of so many ideas around small cells and this is something that we thought. It looks very simple and straightforward and having talked to a few small cells experts, off the record, none of them seem to be able to see anything wrong with this concept. With the 'Small Cells World Summit' just round the corner I am sure this could be something worth a discussion.

I am explaining the concept using an HSPA+ setup but there is no reason why this would not work in an LTE Setup. This is a typical connection for HSPA+ Femtocell setup with the gateway acting as a concentrator for all Iuh connections and having a single Iu connection towards the core. I have not shown CS/PS connection separately for simplicity. 
We propose a 'Virtual' or 'Invisible' Femtocell concept where we think that the Femtocell is redundant but the concept can be used to avoid the coverage and capacity problems faced by the operators and at the same time avoid the 'Signalling storm', atleast on the access network side. Now most smartphones have WiFi stack inbuilt. For this concept to work, WiFi in the phone is a must. Instead of having a Femtocell in between, a modified stack could be embedded in the phone itself. The output of the phone over WiFi are the Iuh messages that can terminate at the gateway and no difference would be needed from the core network side. This is illustrated in the picture below.
The phones would also need to have an enhanced UI to be able to allow a user to select only this option when roaming. You don't want a situation where the user thinks that he is camped on the 'Virtual' femtocell and making/receiving calls while he is not and run up a huge bill.

Advantages of this approach:

  • The Femtocells are no longer really needed and the end customer does not require to buy a separate equipment, which is different for different operators.
  • The phones can be working whenever a reliable WiFi connection is available, even if they are abroad without incurring costly roaming charges.
  • Some operators that do not have a lot of spectrum available avoid using Femtocells as they can cause interference and black holes in the coverage. 
  • There is no worry of a femtocell being used abroad illegally thereby causing interference with spectrum in another country.
  • Some security issues can be totally avoided and it would be worth for the operators that the keys being used cannot be seen by others.
  • A lot of people use OTT apps like Skype, Viber, Whatsapp when abroad, being camped on WiFi to avoid costly roaming charges. This approach would mean that the normal Voice and Messaging becomes similar to OTT and can help operator avoid losing out to the OTT apps. 


Disadvantages of this approach:

  • WiFi spectrum is already congested and does not always give reliable coverage.
  • Security issues would have to be looked in detail to make sure this would be secure enough. Since this concept is similar to creating a VPN between the phone and the gateway, I wouldnt think there would be any issues though.
  • Roaming revenues are a big cash cow for the operators, most of them would be unwilling to lose this if the phones are using this approach.

I think this concept is more suitable for the Residential Femtocells rather than the other Small Cells (enterprise, metro, pico, etc.) and there will always be a need for them. The main reason being that on a large scale, WiFi is extremely unreliable, prone to interference and not future proofed. A new device may cause interference that may take forever to resolve. Operating a small cell in the licensed spectrum would always make sense and the reliability would be much higher.

If you think this makes sense please click the 'Useful' checkbox so that I know.

As a company we are always looking to engage with other companies to discuss similar ideas. If you are a company dealing with Small Cells and are open to discussing similar ideas, please let us know.

8 comments:

Rupert said...

Haven't you just re-invented UMA (or GAN) ?

The issues with that were embedding appropriate stacks into the handset, and operator concerns about quality over unmanaged WiFi

Zahid Ghadialy said...

Hi Rupert, agreed that its not very different from UMA/GAN concept and WiFi issues will be present.

Kit said...

Yes, it is very similar to GAN. Handover issues will need a little bit of work in the HNB-GW if you want to hand-in (but in a similar way to the deployment of GAN / Smart-WiFi, this can be done nearly at the application layer if you don't want to hand-in and limit "stack distortion" and handset vendors willingness to maintain a code base.) Since WiFi is more efficient and widespread now, & increasingly considered by operators, I think the barrier is a lot lower than it was.

Ankit said...

What about paging? What will UE send in LAC/RAC?
What about relocation?
What about location aware services?

I think this will require changes in Wifi access point and thus making it very similar to FAP except the radio interface.

Zahid Ghadialy said...

@Ankit Not sure why the Paging, LAC/RAC and Relocation would be any different from an existing HNB solution. Location aware services would definitely not work as for a Femto solution but operators can always take advantage of location based measurements to find a user's location information.

Anonymous said...

How would the eRelay standardization going to affect Femtocell concept? now they talking about inband outband relay and how to realize them. once this eRelay are in place, it will eventually making femtocell obsolete? or wifi more prominent?

Anonymous said...

Which companies do you think can work on this concept and put it in handsets. Will we see them in iphone first? Maybe iphone 6?

Zahid Ghadialy said...

eRelay: As far as i understand, eRelay would be used for Backhaul of Small cells so Small cells should not be affected in any way.

Companies: I would expect someone like Samsung or Huawei to incorporate this concept in future but there should an operator willing to support this, otherwise this wont work.

Operators can also get clever and agree on a universal OTT service that can avoid the need for Femtocells, etc. all together but we can safely assume that wont happen for long time to come.