About net neutrality

THE DEBATE over net neutrality hit the headlines again as World Wide Web founder Sir Tim Berners-Lee proclaimed that the death of the web will be upon us unless net neutrality is preserved. 

That dire prediction followed the clamour that greeted Google and US telecoms operator Verizon published deal that they claim will preserve the ideology while catering to the needs of information providers and Internet service providers (ISPs). As other telecoms operators and content distributors start eyeing up similar deals, the importance of understanding what net neutrality means is growing in significance. 

There are several competing visions of acceptable and appropriate types of net neutrality that only increase the confusion and misunderstanding surrounding this important concept. The most common are:

• No data traffic discrimination whatsoever - all protocols and data types are treated equally.
• Data traffic discrimination with quality of service (QoS) differentiation between different protocols and data types but no costs attached - data traffic is treated based upon rules, with no price to access high priority QoS levels.
• Data traffic discrimination with QoS differentiation between different protocols and data types, with priced performance tiers - data traffic is treated differently depending on QoS levels and pricing.

The 'no discrimination' policy is truly absolute net neutrality. This puts forward the notion that data from all Internet services such as websites, email, Voice over IP (VoIP), streaming and peer-to-peer (P2P) services are treated the same and given identical priority to use network resources, first-come first-served. However, while this might seem like the ideal, a panacea, the reality is somewhat different.

As network bandwidth is limited, QoS approaches have been discussed and deployed for as long as there have been telecommunications networks. In its most basic form, QoS is a set of rules which prioritises network resources to particular services that are deemed important. Data is analysed in real time and, depending on which QoS rule it matches, is either transferred, delayed or discarded.

Large organisations are likely to impose some form of QoS and firewalling on their network traffic, limiting employee Internet connections, for example. Given that the rules are set in accordance with company policy or budget, it is a reflection of what the organisation deems important for its employees to get work done.

For instance, it is understandable that VoIP data is given higher priority over loading web pages. A few tens of milliseconds of delay when visiting a social networking site is no big deal, but such delays while on a call are perceptible and, if severe, unacceptable.

In this case, the rules on prioritising Internet access are in the hands of the person or company paying for the connection. Proponents of net neutrality fear that, if such rules were implemented by an ISP, it would choose rules that suited itself rather than its customers.

Aspects of a limited Internet already exist. Many ISPs handle P2P traffic at a lower priority, and impose data transfer limits during particular hours of the day. In extreme cases services are blocked completely, such as T-Mobile's decision not to allow its subscribers access to Skype.

Consequently, content providers such as Google, Yahoo and Microsoft may be forced to do deals with ISPs to ensure that they receive priority over rivals. Not only could this lead to an arms race between content providers, but the major concern is that the ISPs will turn round and ask consumers for money to access content from those providers at a particular quality. This is what is commonly termed as creating a 'tiered Internet'.

In the UK, a debate sponsored by US telecoms giant AT&T on the subject was held in Westminster in March 2006, but key companies such as Google declined to take part, calling it "biased". In June 2010 Ofcom published a discussion paper in the hope of gathering industry opinion on net neutrality. At the time, Ed Richards, chief executive of Ofcom, said that transparency was key to customer satisfaction. Ofcom is expected to release an official framework on net neutrality, although no date has been set.

Regulatory bodies in Europe and the US face the problem of balancing the need to be seen to foster free enterprise, while ensuring that citizens have unrestricted access to the Internet.

Given that many telecoms operators are no longer nationalised, governments have limited reach, and could face litigation if they are seen to be taking action which could hurt the profits of publically traded companies.

Comments made in February 2010 by César Alierta, president of Spanish telecoms operator Telefónica, which owns O2 and Be Internet in the UK, only served to raise further concerns that operators are looking to monetise access to key content providers.

"Internet search engines use our net without paying anything at all, which is good for them but bad for [Telefónica]. It's obvious that this situation must change. Our strategy is to change this," said Alierta.

Given Alierta's comments, Google's widely reported deal with Verizon earlier this week should come as little surprise. Net neutrality threatens firms such as Google just as much as consumers, but many feel that the precedent will open the floodgates for back-room deals given Google's staunch support of the net neutrality principle.

The firms claim that the agreement between them does not renege on net neutrality, but merely addresses the realities of today's market. Both agree that all data should be treated equally on wired services. However, the compromise did leave the door ajar for wireless communications to have limitations applied.

Google and Verizon had hoped to appease net neutrality advocates by reiterating support for the FCC. Many groups pointed straight to the gaping fissure that is wireless internet, and slammed it as a corporate pact.

Google said in a statement defending itself that "the wireless market is more competitive than the wireline market, given that consumers typically have more than just two providers to choose from".The firm believes that competition will result in consumers getting the best deal but, given the few telco-independent ISPs that exist, and the percentage of internet data they carry, the agony of choice may not be the solution. Such deals could very well lead to the consolidation of the ISP industry as smaller outfits are unable to compete with large multinational operators.

It should be noted that the deal, or "proposal", laid out by the two firms is merely a voluntary code of conduct to which Verizon has said it will adhere. This 'gentleman's agreement' in effect puts into writing what many had already assumed is taking place. However, the need to reiterate these points in public has led to questions about what hasn't been made public.

Few should be surprised that AT&T has since adopted Verizon's stance that wireless communications is somehow different to wireline. Suspicions are raised by the fact that neither Verizon or AT&T actually offer crediable arguments to back up their claims instead opting to give up on the technical challenge and instigating financial barriers to customers in order to lower investment in infrastructure.

AT&T took the additional step of calling net neutrality oppressive, a move which is sure to be a signal of intent in how it decides to tackle this issue. If all telecoms operators take such an attitude then it is highly likely that net neutrality supporters will be left with nothing.

Assuming that net neutrality fails and leads to the creation of a tiered Internet, the result for users is not necessarily just higher prices. The two most probable outcomes are an end to ISPs offering 'generic' Internet access, and the proliferation of paywalls.

ISPs will look to monetise groups of QoS rules to provide niche products that focus on optimising access to particular services such as VoIP, video streaming and gaming. Given the EU's aim of creating a baseline service, it is likely that ISPs going down this route will still offer a catch-all product which transfers data at lower priority than its QoS-enabled products.

This is not necessarily a bad thing for companies. If ISPs offer service level agreements (SLAs), businesses can rely on Internet access doing what they need without worrying about secondary Internet use affecting key business operations.

However, for many smaller firms, which do not require SLAs and get by on the current 'best effort' Internet, such packages will offer little more than an extra expense.

As content providers try to recover the cost of doing deals with telecoms operators, they will have to look beyond traditional means of revenue generation such as advertising and content syndication.

The notion of a paywall has recently re-entered the public's consciousness owing to The Times charging for access to its website. At present, people have many choices for online news, but choice is likely to suffer if the widely predicted battle of the bank balance plays out.

Ultimately, the argument for and against net neutrality boils down to cost. Data consumption is growing and traditional methods used to service demand have simply become unprofitable. ISPs need to consider using smarter networking technologies to overcome barriers, rather than try to limit consumption.

Even if laws are passed to preserve net neutrality, questions over how it will be policed remain. Whether such activities will end up eating into citizen's privacy and the cost of monitoring cause Internet access charges to increase is not known. 

Net neutrality may be a fashionable buzzword at the moment, but the consequences of its failing are very real. Perhaps ironically, as net neutrality advocates see such a policy as something that will cement the future of the Internet, such a policy would bring about no change at all for consumers.

As for Berners-Lee's prediction for the Web, perhaps the legendary scientist can be allowed some creative license with his words, however it is certain that should net neutrality fail, there will be significant changes to the Web and other services that run on the Internet. µ