What will happen when the internet can't cope?

In less than a decade, the internet has become an essential part of business.

But many experts are predicting that today's web infrastructure will not be able to handle the load when users are fully exploiting its potential.

Lack of security, speed and reliability contributed to the dot com crash and are still issues for many companies, forcing the industry to look more closely at what the future holds.

'People have talked about a lot of possibilities but couldn't implement them in a cost-justifiable manner,' says Gartner research vice president Alex Linden.

'The infrastructure has matured significantly but it's still not enough, although a lot of it should be there in about seven years.'

A number of initiatives are taking place around the world to boost the web's infrastructure. The four most significant are Internet2, the Next Generation Internet, computational grids and the Semantic Web.

The first two projects focus on making the internet faster, more reliable and more secure, and will result in greater network intelligence, says Forrester Research principal analyst Ellen Daley.

'Applications will have to do less work around authentication and security because they will be built in,' she says.

The goal of grid computing is to make computational resources constantly available as though they were a utility, such as electricity. Daley says this is likely to make services and resources on the network increasingly reliable and scalable.

'It's about faster connectivity and not having to think where data is, and such transparency could change the user experience quite significantly,' she says.

The Semantic Web takes a content-based approach. The aim is for software to take raw, undifferentiated content from across the internet and turn it into meaningful information, to make retrieval and navigation more straightforward and effective.

In simple terms, the Semantic Web would work like this: if a user had food allergies and wanted a recipe for stew, their online search would look at the individual ingredients listed, work out if they were suitable and exclude any offending recipes from the responses, thus saving time and reducing the need for manual intervention.

Of all the projects under way, the Semantic Web has the most potential to change the way organisations operate, says Malcolm Hutty, regulation officer at the London Internet Exchange.

'The concept is huge, and so is the potential long-term impact,' he says. 'Who knows what this could be used for in future? There could be a wave of innovation that has not even been imagined yet.'

Hutty says IT directors should not expect the fruits of such research projects to appear in a big-bang fashion, but to emerge over time. Change is not just around the corner, but we can expect to see it in our lifetime.

'In terms of direct, practical impact, changes are diffuse and tend to trickle down,' says Hutty.

The Extensible Business Reporting Language (XBRL) is one example of Semantic Web concepts in action, says Linden. Still in its early stages, XBRL describes and classifies the contents of financial statements and other business reports, to simplify financial reporting and analysis.

But not all experts are convinced that the Semantic Web is a likely outcome. 'An "internet of things" enabled by web services technology is as close to the Semantic Web as we're going to get,' says Mike Gilpin, vice president and research director of Forrester Research.

'When Tim Berners-Lee conceived the web, web services hadn't come along, and while they're not solving exactly the same problem, they have distracted the industry from the idea of the next big thing.'

The 'internet of things' is the next stage of the web's development, says Gilpin. Devices of all descriptions, including wearable computers, mobile phones and oil pumping equipment, will be connected to private and public networks using the TCP/IP protocol.

'Businesses are starting to realise they should no longer think about what's done over the internet as a connection in terms of this application, user or device, but should instead consider which business process is enabled by applications that are making use of these connections,' he says.

This will give organisations a cost-effective means of loosely coupling internal and external applications, without needing to know where they are, how they were built or with what technology, says Gilpin.

Linden agrees. 'If we want great infrastructure we have to improve the way we reuse software, and that will be driven by web services,' he says.

He also forecasts that mesh networks, based on the convergence of wireless standards such as WiFi and 3G, will play a significant role, allowing device users to switch seamlessly to the appropriate network without losing the signal.

But for all this to occur, two key issues need to be addressed. The first is the development of universal web services and globally accepted mesh network standards, and the second relates to standards for universal billing.

If these challenges can be met, says Linden, the internet will achieve the goal of becoming genuinely ubiquitous.

The future of the internet

Computational grids

This global initiative aims to network computational resources such as servers, PCs and storage systems across the world. Grid software presents the machines as a single unified resource for processing data-intensive applications. The concept has been likened to the electricity grid, because users can access power at any time without worrying where it comes from.

www.gridcomputing.com

Internet2

A not-for-profit research and development body set up in 1996 and led by more than 200 US universities. It also includes 60 commercial organisations and various US government bodies.

Internet2's goal is to create a faster internet infrastructure that is more reliable, available and secure, to support new applications such as digital libraries and virtual laboratories and improve research and communication.

www.internet2.org

Next Generation Internet

This US government project was set up in 1997 with input from federal agencies such as the Department of Energy and Nasa, but has since been disbanded. It was seen as complementary to Internet2, and although its focus was more governmental than educational, it was designed to build on Internet2's work to help it achieve its goals more quickly.

www.ngi.gov

Semantic Web

Developed by Sir Tim Berners-Lee, inventor of the web, HTTP and HTML, this aims to allow a more intelligent means of navigating and searching online content, by introducing standards for describing and showing relationships between pieces or collections of data. The addition of such metadata to data streams will help machines to find, sort, classify and interpret information in a more meaningful way to present it in context.

www.w3.org/2001/sw