Mobile networks: The race is on to keep the data flowing

Mobile networks: The race is on to keep the data flowing
By Stephen Pritchard

Published: June 16 2010 01:02 | Last updated: June 16 2010 01:02

Smartphones – such as Apple’s iPhone or Research In Motion’s BlackBerry – have given mobile operators a much needed boost during the recession. Sales of these high-end, usually higher margin handsets have continued to grow.

Research by US-based analysts comScore, found that smartphone sales in Europe’s five largest markets grew 32 per cent in 2009, and by a staggering 70 per cent in the UK.

Internal data compiled by Nokia Siemens Networks, the cellular equipment vendor, found that 11 per cent of consumers said they accessed the internet using a mobile handset in 2008; by last year, the figure was 33 per cent.

As a result, mobile networks are starting to show the strain from far greater numbers of more sophisticated users and handsets. Smartphones, with their “always on” connections and data-intensive applications, are rapidly using up scarce radio spectrum, with one chief technology officer calling it “the iPhone problem”.

The problem is not, of course, restricted to smartphones: 3G dongles, 3G-equipped laptops and even devices such as satellite navigation sets or Amazon’s Kindle, which downloads data in the background, are causing capacity shortages.

According to research carried out by analysts at Exane BNP Paribas, an investment bank, for WiFi operator The Cloud, mobile networks are likely to run out of data capacity within the next three years – and in some areas, as soon as a year.

The typical mobile network subscriber will use two to three times as much data in 2012 as they do today, the research predicts. But the aggregate increase in data traffic – combining both subscriber growth and those subscribers each using more data – could see network data traffic increasing 10-fold year on year.

Some operators are certainly seeing phenomenal growth. At the CTIA Wireless event in Las Vegas in March, AT&T reported a 500 per cent rise in data traffic. AT&T is the exclusive US carrier for the iPhone, as well as for the Palm Pre.

“There are three main problems,” explains Ed Marsden, partner in the tele­communications practice at Deloitte. “The first is the volume of devices using the network. There are now 600m mobile broadband connections, and smartphones have overtaken PCs.

“The second factor is how consumers are using these devices. With ‘all you can eat’ data plans, we’ve gone from under-utilisation to congestion. One large US operator has identified a 5,000 per cent increase in data growth in the past three years.

“The third is our level of understanding of, and the optimisation of, networks. Some smartphones strain the network: they generate eight times the signal load of a mobile broadband [dongle] connection. It’s about understanding how the network is being used and where the pinch-points are.”

There is little doubt that the mobile data industry is becoming a victim of its own success. But as yet, it seems unable to agree on a solution, nor on the likely threat of mobile network congestion to business users of data services.

At the network level, some operators are pinning their hopes on a move to a fourth-generation technology known as Long Term Evolution, or LTE, which promises more efficient use of the radio spectrum, especially for data traffic. The greater spectrum range should allow more users to connect at maximum bandwidth at the same time.

Even vendors of fourth-generation equipment, however, acknowledge that LTE might be too little and too late to solve the network capacity problem.

Operators have been conserving cash during the downturn, so the first large-scale LTE deployments, such as that of US carrier Verizon – will only start in earnest this year.

Moreover, few devices currently support LTE, and for an operator to make full use of the spectrum, it will need to move significant numbers of users to the new radio technology.

As an interim measure, operators are starting to look at alternative radio technologies, including WiFi, WiMax, and even multicast transmission technologies, such as Qualcomm’s MediaFlo, to handle some of the data traffic.

On a more localised level, operators are looking at femtocells to add capacity. These small cellular base stations plug into a business or home user’s broadband line, and route traffic via that connection to the operator’s network. Larger, picocells could be used in public areas.

Samir Khazaha, a senior director at Qualcomm, the cellular technology vendor, says: “Networks are going to rely on more picocells and femtocells, and a proliferation of antennae, so wherever users are, there is one nearby.”

Although almost all smartphones are equipped with multi-radio chips, which can operate over both WiFi and 3G connections, not all handsets manage the transition to WiFi with ease. As a result, operators would need to develop software that switches phones to the nearest available connection without the need for user intervention.

Unfortunately, outside some very busy areas, such as airports and railway stations, WiFi rarely provides seamless coverage.

To fill in these gaps, operators would need to persuade private businesses or even individual broadband subscribers to open up their access points to wireless traffic. As yet, there is no proven business model for this.

Sylvain Fabre, a research director at Gartner, the industry analyst, says: “WiFi is not ubiquitous, so it won’t allow operators to offload everything they need to offload.”

Nor does better local radio coverage solve all the operators’ problems; they will also need to upgrade their core networks.

“Operators will have to build a core network that supports any type of traffic and have the technique to monitor traffic in real time,” explains Thierry Maupilé, head of strategy and business development at Starent, a Cisco company.

Such networks, he suggests, will have to prioritise traffic depending on who the user is and what they are trying to do.

A “gold” tier business user might have traffic routed over a connection with a guaranteed service level; a standard-rate consumer might receive a “best efforts” service, but pay rather less.

This could, ultimately, offer the best solution to operators needing to manage capacity and constrain capital expenditure: to restrict demand.

It might be hard, or even impossible, to wean consumers from flat-rate data pricing, but that need not mean they have to receive the same service as higher-value, business users.

Copyright The Financial Times Limited 2010. Print a single copy of this article for personal use. Contact us if you wish to print more to distribute to others.