Singapore broadband speeds to get real-world tests through SamKnows

After years of complaining about slow connections, Singapore’s broadband users can now join a study that uses their actual day-to-day experience to offer a clear snapshot of just how fast their Internet service is.

One of the long-standing problems here is that much of the content that users access is based overseas, and links to these sites seldom reach the advertised speeds, say, 100Mbps, that service providers promise. Now, instead of second-guessing or asking for advice from strangers in a forum, users can get a better sense of the actual speeds from real users’ feedback.

Some 900 volunteers are now sought to participate in a study conducted by well-known research firm SamKnows, which has carried out similar research in Europe and the United States. Singapore’s Infocomm Development Authority (IDA) is backing this survey in Singapore.

While the IDA has been publishing monthly throughput results based on its own tests, the new tests will have users collecting the information from their homes. This promises a more realistic look at how fast things are in the real world.

After they sign up, volunteers will be provided with a network device that helps test the connection speeds to several data centres in Singapore, the US, Britain and China, according to the IDA.

All that users have to do is plug in the “white box” between the modem and existing home router. The device does not collect any browsing history, according to SamKnows, which has more details on its website.

After each month’s tests, the result will be posted on IDA’s website, so buyers of broadband services can easily check out how well each service provider is doing before signing up with them. The study, which is expected to run for 12 months, is known to be very comprehensive and strict when it was run overseas.

The measuring device, modified from a TP-Link router, is easy to connect and is largely temper-proof. In fact, SamKnows is known to be strict in selecting volunteers and telecom executives asking to volunteer for the test have been rejected before.

How will this sturdy change things? For one, users will be able to see clearly how fast each telco is, based on the actual experience of fellow users. This cannot come sooner for many users who are about to sign up for fibre broadband services, as the nation’s next-gen network hooks up 95 per cent of the place by mid-2012.

Secondly, smaller players such as ViewQwest and MyRepublic, which have prided themselves on faster connections to game servers, for example, may be able to compete based on a better online experience instead of just price.

What else can be improved?

While the tests are pretty comprehensive in testing for bandwidth and latency, it is crucial that the Internet servers that they are checking against are randomised and changed regularly so that service providers cannot “game” the system. Users will not want them to just cater to these test sites but ignore or throttle the access to the rest of the popular sites that they go to.

Perhaps, in future, the test devices at home can also provide a general overview of just how fast users here are connecting to their favourite online destinations, rather than the usual few test sites that IDA is now mandating.

It’s understandable that users will not want to reveal too much about where they are surfing to, but could the device capture aggregated information that can be shared without compromising privacy?

Say, can it measure the overall download speeds and latency experienced for local sites and overseas sites in general? Can it measure the experience based on the type of traffic, say, peer-to-peer traffic and standard Web surfing (HTTP)?

That would definitely give more meaning to the numbers that service providers now tout on their brochures. It also means they have to give a good experience to the type of  customers they want to sign up.

If they want to throttle peer-to-peer traffic, say, they would know that these users will hop on to another telco which appears not to throttle their downloads. If they promise smooth YouTube streams, then they will do well to make sure the latency is low for that service.

In other words, the experience reported each month by real users will shape the way they shape their Internet traffic and how they cater to their most important customers. If that is implemented, users will finally get the transparency they deserve before signing up for a two-year contract.