February 5, 2010

IM, Twitter, Facebook and the Signaling Impact on 3G Networks and Capacity

As the industry prepares for the annual Mobile World Congress in Barcelona February 15-18, we are seeing an uptick in press announcements and articles with regard to the issues of network capacity crunch how to solve these challenges. As some industry insider are (accurately) predicting, “data offload” will be a hot topic at this event, CTIA, and other events later this year.

Though the issues of network capacity are related to many facets of a network (backhaul, core, number of active “heavy bandwidth” 3G devices, software upgrades, etc) – the impact of low-bandwidth applications have gone relatively unnoticed until now. We may think that a small number of subscribers are taking up all the bandwidth in small metro area with the latest and greatest 3G devices, but as one long-time industry expert points out, application “signaling” traffic may be taking up more capacity than earlier anticipated.

Have you really considered the impact of signaling (and thus capacity impact) of active/passive subscribers’ use of Twitter®, Facebook®, and Instant messaging (IM over Yahoo, MSN, Skype) on network resources? Just think about that for a second. How many applications do you have on your Blackberry, iPhone or Android-powered device? How many of them pull down scores (Sports), weather, peer-to-peer IM and other “thin app” data? If you add it all up, the impact on capacity far outweighs the impact of some heavy data download/upload by 3 to 5% of subscribers.

Connection attempts may take up as much as 60% on the RNC processing power whereas 40% is dedicated for actual throughput. This data suggest that backhaul alone may not solve the capacity crunch.

In his latest report (“The Trouble with Twitters”*), Michael Thelander, founder and ceo of Signals Research Group, states: “Without question we have demonstrated that smartphones generate a disproportionate amount of signaling traffic, even when the smartphones are not being used. The question then becomes what can be done to combat this problem – a problem that will only intensify with the increasing popularity of smartphones and social networking websites and services.”

He goes on to say that there are many ways to solve the problem (You have to buy or subscribe to SRG to see his proposed solutions). As it relates to indoor wireless system, he reflects that… “Interestingly, picocells wouldn’t do much to address the issue of smartphone signaling, since, unlike femtocells, they connect directly to the RNC via the Iub interface. Picocells would, however, help offload data traffic from the network.” Though we did not partake in the research study, Signals Research Group notes that SpiderCloud Wireless and our E-RAN approach may play a role in the solution.

“The impact of smartphone data and signaling traffic on macro network capacity has strong ties to enterprise usage or usage from other large indoor locations where people congregate. In this case, a “mini-RAN” approach may be a more appropriate solution since additional intelligence would be needed for cell handovers and to minimize the interference from a large number of access points. Further, to the extent that voice calls or IMs are remaining within the enterprise, a “mini-RAN” solution, such as what is being promoted by SpiderCloud Wireless, would be able to keep the traffic within the “mini-RAN” and completely removed from the operator’s core network. The same can also be said for social networking services.”

As noted in our previous blog from 2 weeks ago, “3G does not have a coverage problem (….it’s a capacity issue!) ” – we conclude that “The net impact of the success of smartphones means that network operators must find ways to handle data offload with and without licensed spectrum and leverage Ethernet LAN where it’s feasible.”

See you in Barcelona.

Ronny A. Haraldsvik
Vice President of Marketing

*Report mention and quotes reprinted with permission from Signals Research Group.