What's really causing the capacity crunch?

The data capacity crunch is shaping up to be one of the biggest themes in 2010, highlighted by the popularity of smartphones in the U.S. and Europe, in particular the iPhone. But is this capacity crunch that is so often mentioned really a matter of customers consuming too much data on operator networks?

Throughout much of 2009, AT&T became the poster boy for data capacity problems as the popularity of the iPhone and other smartphones pushed data usage through the roof. Company executives admitted that AT&T's network was not operating up to par in densely populated cities such as New York and San Francisco, but the company says it has now alleviated those problems. AT&T has indicated that about 3 percent of its smartphone users are generating approximately 40 percent of the operator's data traffic, and the company is looking at ways to encourage these customers to modify their usage.

Other HSPA/WCDMA operators came forward during the Mobile World Congress trade show in Barcelona, Spain, last month to indicate that data congestion is becoming a problem. Vodafone CEO Vittorio Colao expressed his concerns that the rapidly surging growth of smartphones could soon lead to the outstripping of network operators' capacity and their ability to ensure the smooth flow of data across their networks. Echoing the words of AT&T, Colao claimed there is urgent need to implement new data tariffs to allow operators to invest more to boost the capability of network infrastructure, while still generating a profit.

On the other hand

Conversely, Santiago Tenorio, group head of radio development at Vodafone, stated that there is not a data capacity crunch issue, claiming that all it takes "is a robust capacity [planning] process and sufficient investment, both of which we do group-wide."

The contradiction may lie in what is causing the bulk of the capacity problems.  According to Michael Thelander, CEO of Signals Research Group, the majority of the network congestion is related to signaling traffic coming from smartphones--which are ever increasing on 3G networks--making constant queries of the network as they move among cell sites to push email, access social networking tools and conduct other repetitive actions. Thelander recently spelled out the problems in detail in his Signals Ahead research report.

While data traffic is also growing, by many accounts, signaling traffic is outpacing actual mobile data traffic by 30 percent to 50 percent, if not higher, Thelander said. For instance, a Yahoo IM user may send a message but then wait a couple of seconds between messages. To preserve battery life, the smartphone moves into idle mode. When the user pushes another message seconds later, the device has to set up a signaling path again.

"Smartphones are causing a problem, but it isn't data usage," Thelander said. "The base station controller is spending a lot of its resources trying to process the signaling so it can't do other things like allocate additional resources for data. You'll see dropped calls and data service degradation."

Moreover, even when the signaling resource is released by the smartphone, the network can't react fast enough to allow for the next station to use resources until several seconds and sometimes minutes, noted Mark Pecen, vice president of advanced technology with Research in Motion.

Fast dormancy

One of the ways smartphone vendors try to combat this problem and save battery life on their devices is by incorporating a fast dormancy feature, which allows the mobile device to quickly make a query to the RNC to release the connection so that it can return to the idle state faster. That means the device is relaying the fact that the phone is going dormant rather than network, saving signaling channels, Pecen said.

But the feature seems to be more art than science. Thelander said Apple upset several operators last year when it implemented firmware 3.0 on the iPhone with a fast dormancy feature that prematurely requested a network release only to follow on with a request to connect back to the network or by a request to re-establish a connection with the network. Apple soon after released firmware 3.1, which removed the feature but spurred complaints among iPhone users when it came to battery life, Thelander said.

Fast dormancy "can prematurely request a network release, but it depends on the method you use to detect when to go dormant," Pecen said. "It's a clever thing, but it's not 100 percent correct. It's a guess. But we're about 98 to 99 percent correct ... You are going to incur 30 or more signaling messages to set that transfer back up, but most of the time you are gaining something."

While fingers typically point at BlackBerry devices as being signaling hogs, RIM argues that the efficiency it has created in terms of scalability, compression techniques and smart puncturing with push email and other features such as browsing means its devices don't have to wake up as often to query the network.

"We take these things seriously and do examine ourselves," Pecen said. Of course, even the most signaling-efficient smartphone can clog a network if there are enough devices connecting to the network.

This is why operators will continue to look at a variety of solutions. One solution is implementing signal saving network features that are already 3GPP standardized but have yet to be implemented, thus reducing signaling traffic while preserving battery life, Thelander said. This is a solution Nokia Siemens Networks began touting at the Mobile World Congress last month.

"We as an industry have to get together and continue to tailor the operational part of the radio network to be better suited to data traffic," Pecen said. "It's only been eight years [since smartphones were introduced]. It's a matter of continually going back to 3GPP and working with others like the Wireless World Research Forum for future technologies."

Alternatives

In the meantime, Thelander lists a number of solutions operators can look to ease the congestion, which also work for real data traffic surges. These include offloading data traffic onto femtocells or WiFi hotspots and adding more network resources. He also said enhancements that will come with HSPA+ that will enable the networks to reduce their transmission time from idle to active state and should help ease the congestion.

Another tool advocated by Thelander is intelligent network monitoring, a notion furthered by Alcatel-Lucent through the 9900 Wireless Network Guardian. The product offers real-time traffic monitoring at the packet level across both the radio access and core networks from one platform.  
 
"Traditional traffic policy solutions that measure data traffic in the network core aren't effective because traffic in the core isn't truly representative of traffic models in individual cells," said Marcio Nespatti, corporate program manager with Alcatel Lucent. "In reality, only a fraction of network cells today even need traffic balancing solutions."

As such, the vendor-agnostic product, which works on 2.5G, 3G and 4G networks, is designed to give operators greater insight into what devices, which users and what applications and when and where congestion problems occur, enabling operators to intelligently add additional resources to certain areas when necessary, Nespatti said.

Operators can also monitor and control which devices and applications are generating excessive amounts of signaling or hogging more than their fair share of bandwidth. With this information in hand, operators can have increased awareness of the overall cost of delivering specific applications and services and can even map out their smartphone portfolios accordingly, Nespatti said.

The Wireless Network Guardian will be part of AT&T's Long Term Evolution network, according to AT&T's LTE contract with Alcatel Lucent.