& cplSiteName &

Mobile Wi-Fi Offload

Heavy Lifting Analyst Notes
Heavy Lifting Analyst Notes
Heavy Lifting Analyst Notes
4/7/2011
50%
50%

In the course of researching my upcoming Heavy Reading report on Wi-Fi Offload for Mobile Operators, I came across an interesting academic paper that estimates 65 percent of smartphone traffic is already (or can be) offloaded to Wi-Fi under normal usage conditions.

The paper, "Mobile Data Offloading: How Much Can WiFi Deliver?," attempts to quantify how much traffic is, or can be, offloaded from cellular networks to Wi-Fi by analyzing users' actual physical movements throughout the day and mapping that to a data consumption model.

To achieve this, researchers developed custom handset software that tracked 100 iPhone users in South Korea to determine how often, and for how long, people are within range of Wi-Fi connectivity to the Internet, and what the data rates were at any given point. This data was then combined with a traffic model to simulate how much mobile data could be offloaded to Wi-Fi over a one-month period.

The study demonstrates with a degree of academic rigor what we know intuitively: that many people spend a good deal of their time within range of an accessible Wi-Fi connection. The simulation assumed 7GB of data consumption per month, as predicted by the Cisco Visual Network Index for 2014, adapted to approximate a real-world consumption pattern (i.e., a mix of small and large files).

The key takeaways from this project -- at least, as I see them -- include the following:

  • Users are in Wi-Fi coverage 63 percent of the time during the day (70 percent over 24 hours); each stay in a Wi-Fi zone lasts about two hours.

  • 65 percent of traffic can be offloaded to Wi-Fi under typical usage conditions using on-the-spot offload (i.e., using Wi-Fi on demand when available)

  • Greater offload performance can be achieved if the user is prepared to accept delayed offload (e.g., sync your videos or photos when you get home)

  • This means that out of the 7GB usage per month, 4.5GB would travel by Wi-Fi and 2.5GB by cellular. Therefore a 2GB-3GB per month cellular data plan is probably enough for most users.

  • Offload performance depends, obviously, on the pattern of Wi-Fi coverage as well as user mobility. South Korea probably has denser Wi-Fi than some other markets, but still many of us have Wi-Fi at home or at work.

  • Wi-Fi provides connections of 2 Mbit/s on average, which is skewed to 2.76 Mbit/s at night, (probably home Wi-Fi) and 1.26 Mbit/s during the day (probably shared Wi-Fi).

  • Because Wi-Fi is faster than 3G, offload also generates a 55 percent improvement in battery life. However, the simulation only considered the energy used during transmission and did not take into account the battery drain from scanning for, and attaching to, Wi-Fi access points.


The assumptions and methodology of a study will inevitably have an impact on the output of the simulation, but the data does appear to show, pretty conclusively, that Wi-Fi is already taking much of the burden of smartphone traffic growth.

A logical question from this finding is then: How useful would it be for operators to more actively manage Wi-Fi offload? Would that generate an even greater performance gain? Or are operators getting most of the offload benefit without having to do anything?

— Gabriel Brown, Senior Analyst, Heavy Reading

(9)  | 
Comment  | 
Print  | 
Newest First  |  Oldest First  |  Threaded View        ADD A COMMENT
paolo.franzoi
50%
50%
paolo.franzoi,
User Rank: Light Sabre
12/5/2012 | 5:08:18 PM
re: Mobile Wi-Fi Offload


 


I am really curious to AT&Ts view of the new version of the UMA service (I think its called WiFi calling now and is really just the old service rebranded).  I really liked the service and was just about to re-up with T-mobile to get on new phones that supported this.  I expect AT&T to drop it, so now I am thinking maybe I should jump to Verizon.


seven


 

rate-limit
50%
50%
rate-limit,
User Rank: Light Beer
12/5/2012 | 5:08:13 PM
re: Mobile Wi-Fi Offload


As far as I know this kind of service is not that popular in Japan, while every operators are very keen to do. There could be couple of reasons, but the most serious issue is, such *potentially* offloaded traffic is likely generated at home(>70%), and celluar operator has to navigate thier subscriber to wifi seamlessly. This can be technically done via;


- user device will be automatically provisioned to wifi when user is at home


In this model, celluar operator provides wifi ap to their subscriber home and remotely provisions ssid on wifi ap, allowing only familiy member can join such ssid and so on...


I feel how to remotely provision is key to accelerate offload service penetration.


 

Gabriel Brown
50%
50%
Gabriel Brown,
User Rank: Light Sabre
12/5/2012 | 5:08:10 PM
re: Mobile Wi-Fi Offload


@rate-limit – I agree, remote provisioning of devices to automatically use WiFi would accelerate adoption. This is part of the EAP-SIM and AAA integration effort underway across the industry.


In Japan I've heard Softbank is giving away FON-style WiFi access points with new smartphone subscriptions and encouraging users to plug them in at home.


The smartphone is pre-configured to use this WiFi, and under the FON model, each user can access any other of the Softbank/FON WiFi APs around the country. Eventually Softbank will build a WiFi network with substantial footprint. 


 

Gabriel Brown
50%
50%
Gabriel Brown,
User Rank: Light Sabre
12/5/2012 | 5:08:10 PM
re: Mobile Wi-Fi Offload


 

Seven, I do not know AT&T’s view of the T-Mobile UMA services. Historically AT&T has not backed UMA; it would take quite a change of heart to embrace it now and I do not think that’s likely.

But my understanding is that T-Mobile’s WiFi Callling service is going great guns from a call volume perspective. The new Android devices have really helped. Since T-Mobile is also said to have more than adequate “UMA capacity” it would be strange to close down the service, even under AT&T ownership.

More prosaically, isn’t it going to be ages before anything much happens in the service portfolio of either operator as a result of this proposed acquisition?


 

Francis McInerney
50%
50%
Francis McInerney,
User Rank: Light Beer
12/5/2012 | 5:08:00 PM
re: Mobile Wi-Fi Offload


Wi-Fi bypass, not offoading, now accounts for all the growth in wireless.

Gabriel Brown
50%
50%
Gabriel Brown,
User Rank: Light Sabre
12/5/2012 | 5:07:56 PM
re: Mobile Wi-Fi Offload


Thanks rate-limit,


I'm not convinced always-on WiFi makes much difference to battery life when in idle mode. 


Where there is an impact is when the device is active and you're actually using the network. You could argue that WiFi is more power efficient than 3G, although from a practical point of view I don't notice much difference, all else being equal.


I do agree that managed WiFi offload is potentially complex and possibly more effort than its worth -- that's part of the point this column makes. Half of traffic is already sent over WiFi without operators having to do anything.


BUT, I also think that WiFi in mobile operator environments is not neccesarily about "offload" per se. It is as much about extending coverage and services to hard-to-reach indoor locations. It is about a better customer experience.


EAP-SIM is being deployed very much in that vein: to make it easier to connect to hotspots, which are typically indoors at hotels, convention centres, etc.

rate-limit
50%
50%
rate-limit,
User Rank: Light Beer
12/5/2012 | 5:07:56 PM
re: Mobile Wi-Fi Offload


Gabriel,


 


I think real challenge they are faced with is, mobile user does not really like to "always on" wi-fi, in order for battery life. Therefore, ideally when UE moves to celluar provider's wi-fi hot spot, network preference for UE will be automatically provisioned to UE, with EAP-SIM authentication. However, I think this requires wi-fi hot spot to routing area mapping but clearly this is operationally too complex or expensive....hence my impression is, wi-fi offloading cannot become majority of offloading access due to such restriction...from this aspect i would guess femto is much better.

Gabriel Brown
50%
50%
Gabriel Brown,
User Rank: Light Sabre
12/5/2012 | 5:07:55 PM
re: Mobile Wi-Fi Offload


Hi Francis -- would you care to elaborate?


Second-guessing what you might mean, there's certainly an impact from the fact that half of smartphone usage is now via WiFi. That has implications for operators' value-added services strategies.


For example, anything related to the use of subscriber data management is impacted, since half the experience is invisible to the carrier.

armor.chen
50%
50%
armor.chen,
User Rank: Light Beer
12/5/2012 | 5:03:53 PM
re: Mobile Wi-Fi Offload


Wifi can switch to a sleep mode if communication is not used.


When a mobile device use wifi, it may consume more battery power but may obtain more bandwidth. It is a tradeoff. So, it needs an intelligent method to help user or communication system to select a best network according to the user preference, network operator's policy, device's condition and network condition.

More Blogs from Heavy Lifting Analyst Notes
IPV6, security, DNS evolution and automation were all on the agenda at the UK Network Operators' Forum gathering in London.
Developing an open architecture for domain-specific accelerators could empower service providers to accelerate edge computing and related applications.
Auctions of 5G-suitable spectrum are now generating billions. Meanwhile, vendors are waiting on a different bounty — the further billions operators will need to spend to finally realize their 5G ambitions.
US cable providers are gearing up to play a crucial role by supporting transport of 5G communications, but their aspirations are clouded by the complexity of 5G itself.
Many operators are expected to initially deploy 5G RANs in non-standalone mode on an LTE network, migrating later to standalone mode. But that migration is less than straightforward, and there are multiple options to consider.
Featured Video
Flash Poll
Upcoming Live Events
March 12-14, 2019, Denver, Colorado
April 2, 2019, New York, New York
April 8, 2019, Las Vegas, Nevada
May 6, 2019, Denver, Colorado
May 6-8, 2019, Denver, Colorado
May 21, 2019, Nice, France
September 17-19, 2019, Dallas, Texas
October 1, 2019, New Orleans, Louisiana
October 10, 2019, New York, New York
November 5, 2019, London, England
December 3, 2019, New York, New York
December 5-3, 2019, Viena, Austria
All Upcoming Live Events