The only way to get the optimal efficiency out of spectrum assets is to use as much of them as possible for LTE services, according to Paul Ceely, head of network strategy at EE, the first operator in the UK market to launch LTE.

The firm gained permission from Ofcom to refarm its 1800MHz spectrum for LTE services in August 2012, allowing it to launch ahead of its rivals, which only secured LTE spectrum in the country’s 4G auction in February this year. In the same spectrum auction EE also bought 2x40MHz of extra spectrum to add to its existing lot of 1800MHz and 2600MHz frequencies.

By the end of May, Ceely said during a presentation at the LTE World Summit in Amsterdam, EE had launched LTE in 80 cities around the UK, providing 55 per cent population coverage, with an aim of increasing that to 98 per cent by the end of 2014.

As a result of the auction, EE was able to double the amount of 1800MHz spectrum bandwidth that it dedicates to LTE services — from 10MHz to 20MHz – also doubling the average speed 4G customers experience to more than 20Mbps. The company also cut its prices, well aware that its first-mover advantage is shortly to expire—in a bid to attract as many users onto LTE as possible.

Ceely said there were only three clear ways to build more capacity into a network: through greater spectral efficiency; by acquiring more spectrum; and by adding more cells.

For EE, the quickest way to tackle the first part of the solution and increase spectrum efficiency was to refarm some 2G spectrum to be used for 3G. The auction took care of the need for more spectrum, while the merging and migration of the legacy T-Mobile and OrangeUK networks will result in a better deployment of cells and infrastructure.

But Ceely acknowledged that the network is becoming very complex, with a great deal of different spectrum in use and said the company is about to trial carrier aggregation technology with a view to deploying it next year.

But he also spotted a gap in the market for tools that can help make sure the end user is directed to spectrum fit for their specific purpose.

“We need to develop analytics that can hook into the network, to make sure the user is using their chosen application on the right spectrum,” he said. “This way we make sure for example, that 800MHz is only used when absolutely necessary – like for situations where it’s the only option in terms of coverage – so the user is kept on more appropriate spectrum like 1800MHz or 2600MHz if that suits their current use case.”