advertise here



Industry Comment Research   RSS Feed

Webinars Buyers' Guide Podcasts

Related Publications Foward Features




  In partnership with:

May/June 2006 issue

Telecoms – convergence challenge?
Ian Grant

Churchill said, "Give us the tools and we will finish the job." We have the infosecurity tools, but the job of managing infosecurity in an age of converged digital networks is only just beginning.

Earlier this year German scientists working with Fujitsu transmitted a data signal at 2.56 terabits per second, the equivalent of 60 DVDs/s, over a 160km link. This doubled the previous five year old record and is 50 times quicker than the fastest commercial high speed links, now around 40 Gbits/s.A little later, following his reverse take over of cable TV company NTL, Virgin boss Sir Richard Branson outlined his plans for a new communication s service. This will introduce a single bill for access to and use of telephone service, mobile telephony, broadband connection to the internet, and television. In true Branson style, he suggested the Virgin service be called 4Play.

These moves are the more visible foundations of an unparalleled integration of content over a single communications connection. Less obviously, scientists are starting to unleash a new breed of objects, typically tiny sensors, that detect changes in ambient conditions and communicate them (wirelessly) to a controller for analysis and response. The net result is that, to coin a phrase, Big Brother may be watching you, but only so that he knows when to deliver what you want. And what that is, is a $64 billion question. Researchers at Finland's VTT Technical Research Centre explored the infosecurity aspects specifically of digital television. They say,: “From the information security management point of view, the interconnection of the different systems, and understanding the whole environment are very challenging, and have not yet been solved.”

However, Robert Temple, Chief Security Architect at BT argues that there are “no insurmountable technical or engineering security issues in the way”. He points to the scores of very active standards-making bodies that are striving to build the technical foundations for a trustworthy environment. “We've got all the technical standards we need,” he says; it's who is going to take the commercial plunge.

Temple reckons the market will ‘federate’. In addition to traditional network operators, content providers from movie studios and record companies to publishers and search engine firms are starting to jostle for position with search engine and network operators. Transaction-based firms such as banks, credit card companies and retailers could enter the fray. It is not inconceivable to see AT&T tie up with Google, Time Warner and Visa, or BT with Yahoo, Newscorp/BSkyB and Mastercard. Such alliances could make commercial sense. And there is little to stop them except corporate egos, antitrust lawyers in Washington and Brussels, and consumer distrust.

State of play
The International Telecommunications Union estimates that the world spent $1,248 billion on telecommunications services in 2004. That was up 11% on the year before. According to the latest market assessment from the European Commission, last year the total EU market for computers and networking grew about 4% to E614 billion, of which 44% (E273 billion) was for 'e-communications'. Adding content and service revenue streams would raise these figures dramatically.

The key market driver is the rate at which companies and end users take up broadband connections. The EC says there were nearly 53 million broadband lines in Europe last year, a rise of 20 million over the year before. Market penetration now averages 11.5% compared to 7.3% in 2004. Network operators' capital investment, estimated at E45 billion and rising, ensures that penetration rates will speed up.

For network operators, the key metric is average revenue per user (ARPU). Traditional fixed line voice telephony is still the main source, worth some E85.8 billion in 2005. But it is dropping at an average of 1.6% a year. This is due partly to incursions by mobile telephony operators; voice over the internet protocol (VoIP) will accelerate the decline. Skype already claims its proprietary PC-based peer to peer VoIP service has over 100m users. European mobile market penetration rates average 93%, but some countries are now over 100%.

ARPUs for both fixed line and mobile network operators have stalled. In the UK, the telecommunications regulator Ofcom, has just licensed 11 firms to provide mobile (GSM) 'picocells'. From next year the cost of making a mobile call from a picocell equipped office will drop dramatically. Moreover, the European Commission wants operators to slash the cost of cross-border mobile calls, currently about 10 times the cost of national mobile calls.

Existing network operators are therefore anxious to find new revenue sources. For most, this means adding services and down loadable products to their offering. By and large this requires changing the basic networking technology from circuit switched, the technology of the telephone, to packet switched, the technology of the internet. It also means forming alliances or competing with providers of network enabled services and/or own copyrighted content. Until now, these were merely customers.

Most operators are already responding. BT will trial its IP based 21st century network in Wales this year, with commercial switchovers from next year. Last February mobile operators Vodafone, T Mobile, Orange and others said they will introduce HSDPA (High Speed Down link Packet Access) this year or next. This will allow them to pump data to cellphones at between 1.8 and 14 megabits/s. Soon to follow are equivalent transmission speeds in the reverse direction.

These speeds make possible smooth, high resolution, realtime TV to and from your cellphone, or laptop, or P.A. Interactive digital TV is currently the perceived Holy Grail, partly because of the advertising money that goes into TV and the potential to add transactions to ads, but also because their present pricing model charges for bits delivered, and digital TV is bit-intensive. However, Baby Boomers' kids are less addicted to TV than their parents. They may spend more hours with the TV on, but often they are using their cellphones, IPOs or PCS concurrently. As a result, adspend for TV is falling while adspend on other digital formats is rising fast.

Intelligent environment
Just creeping over the horizon are networks of 'picocomputers' which, when networked together, will create an 'intelligent ambient environment'. The first concrete manifestation of this is the proposed replacement of the barcoded label, the RFID tag. The tag uses low power radio in either active or passive mode to transmit information about itself or its environment to nearby receivers.

Some believe these units will become ubiquitous. The most obvious applications involve identity authentication, monitoring ambient conditions, and transactions, often in combination.

The biggest problem is that all these units, and mobile terminals, use the electromagnetic spectrum. Spectrum is a finite resource, subject to heavy regulation and penalties. Most of it is already earmarked for applications such as TV broadcast, emergency transmissions and the like, leaving little space for new applications.

The useable spectrum is finite, so researchers are exploring other opportunities. The US regulator, the Federal Communications Commission (FCC) found that some 70% of all allocated spectrum may be unused at any one time. This has prompted researchers to consider 'cognitive radio'. The idea is to make transmitters that listen to the traffic, figure out which frequencies are quiet, and then switch to them. The snag is that frequency changes must be synchronized if the receiver is to keep the signal. The military has had frequency hopping radio since the 1970s for secure, unnameable battlefield communications, but so far it has not caught on for commercial applications.

New network; same worries

Chief security officers who plan to use BT's new IP-based network, the so-called 21st Century Network, need to work on their policies now
because things are going to get complicated.

BT aims to test the new all-digital network in Wales in November this year. If all goes well, it will start cutting over customers, 50,000 at a time, in 2007. The aim is to migrate everyone within two or three years.

Technically, the new network won't care what content it carries, as long as it is digital. For CSOs it means that all the hazards of the internet might now affect other forms of communication, such as telephone calls and transaction data.

BT's chief security architect, Robert Temple, says the present internet and other services such as fax and data run on networks overlaid on the existing circuit-switched network. This is optimised for telephone calls.

The new network means that each service will run in its own 'logical domain', says Temple. “Each will have its own security in its own domain.”

Temple warns that end-to-end security is a matter of 'partnership' with the customer. “There is no substitute for good hygiene at the customer level,” he says.

“We may do more as the network matures, but customers will be wise to use firewalls, end-point authentication, the latest patches and operating system updates, and the like.”

This suggests that, in the short term, BT and other IP network operators are more concerned to stop people from accessing services to which they have not subscribed. But bundled offerings such as NTL-Virgin's 4Play are likely to become standard very quickly.

“BT faces a competitive market,” says Temple. “The commercial model must make sense. At present there's no logic in doing a lot of anti-virus etc. in the network.”

But with carriers like BT hungry for new sources of revenue, watch this space.

Who owns the customer?
However, BT's Temple notes the real question is who owns the customer? Terminal devices are increasingly customized to accommodate the user's personal needs and desires. Product choice is presently how users gain the desired personalization. But cost and logic suggest that in future personalization is more likely to be a function of the SIM card.

Until now, cellphone makers have worked with mobile network operators to roll out new features. However, the switch to IP based traffic is likely to weaken those links. In some countries users can already make financial transactions using their cellphones, with the cost of the item debited to their cellphone account. As a result, some banks are issuing bank branded cellphones.

It is a small step technically to use the information on the SIM card to authenticate the user to the network, to the vendor, and to the bank. This makes the SIM in effect an ID card. Taking this further will be tough because all the players want account control, Temple says.

Although the cellphone improves convenience all round, widespread take up is likely to be governed by the trust that account holders place in everyone (and everything) in the value 'net'.

The VTT researchers say "The most important factor is the customer's trust in the service and its provider. The enterprise's reputation, in addition to costs, is important from the end user's perspective when selecting the service provider."

Referring to digital television, they say: “For the time being, the application environment has been restricted and strictly under the control of the digital television network operators and broadcasting channels because the application comes within the programme signal. This is going to change due to the emergence of MHP version 1.1, (which enables) applications to be loaded via the return channel.”

Shifting to IP and terminals with data storage opens them to the hazards of the internet. As Temple notes, as the value of transactions and information on the networks rises, so it is likely to attract better financed, better organized and more motivated bad guys. Potential new threats include organized crime gangs, terror groups and hostile states.

The DoS-resistant working group is part of Cambridge-MIT Institute's Communications Research Network, which is researching ways to prevent attacks. At its inaugural meeting in January 2005, delegates from 50 interested concerns, including the military and intelligence communities, heard that botnet-based denial of service attacks might drain up to 3Gbit/s bandwidth from the networks. “Alarmingly low amounts of traffic suffice for causing damage,” they heard. This could be as little as a few hundred bots for static web pages, dynamic pages and SSL, or even a few dozen if fired rapidly; SYN scans and bandwidth are vulnerable to anything from a dozen to a few hundred attackers. Botnet attacks by up to 80,000 nodes were reported.

Current defensive measures include securing the core; responding to the incident through preparation, detection, classification, trace back, containment and post-mortem analysis. (For a fuller treatment see http://communicationsresearch.net/dos-resistant/meeting-1/cii-dos-summary.pdf.)

The consensus at the meeting was that future defensive architectures are very hard to predict and build. This is because no-one is responsible for it. Furthermore, fixed security standards could inhibit innovation of profitable new products and services.

The group planned to set up a registry of attack profiles and attack types to track trends. It also hoped to classify attackers' motives. It planned to draw up a registry of defence techniques, a 'standard' picture of the network at any point in time, and to coordinate responses to attacks.

As they noted, end users cannot do much against DoS attacks, so the need is to deal with such attacks as far upstream as possible. This requires coordination at network level and security regarding specific responses to slow what Temple calls the arms race against the hackers.

Even so, infosecurity means different things to different people. As the Finns note, the emphasis on threats varies in severity and solutions in different parts of the value net. Content producers worry most about unauthorized use and distribution of programs or other content. Network operators worry that erroneous content will affect terminal devices and/or users' data and applications. End users' concerns are primarily invasion of privacy and identity theft.

Up to now, companies like Google and eBay have shown staggering success, despite the infosecurity hazards. In future, the risk-reward ratios for users are likely to change for the worse. Unless everyone in the value net, including the justice system, cooperates, end users are unlikely to trust the system as much as they should. Without that trust, despite global networks, the world will become a very small place. •

About the author
Ian Grant is a freelance writer on business issues.

Back to features index



 

 

Search this Site:
Google Custom Search



Click here...