The idea of getting unlimited streaming movies for something that might be around $10/month is pretty appealing. And with more and more TV & compatible devices (PlayStation 3, Xbox 360, Wii, etc) with Netflix support, I think this kind of service will have genuine mainstream appeal. No longer will folks have to mess with wires to connect their PC to their TV!

I do have a few concerns about these types of services. I can’t be watching a movie in the living room with my wife and have the movie break-up and begin buffering because there is congestion on the network. In order for services like these to satisfy mainstream customer expectations, the quality of experience should be consistently excellent. High-quality large format TVs are very good at highlighting the difference between high quality and low quality video.

The reality is that streaming video is perhaps the most demanding consumer application in terms of broadband network resources. As such, it demands intelligent and efficient network management; network management that is sensitive to and aware of the specific requirements of various applications. A consumer’s sensitivity to a video buffering mid way through an action scene is very high. Their sensitivity towards a bulk file transfer taking 10 seconds longer is very low. Network operators with visibility to and understanding of the broad range of applications that traverse their network are in the best position to satisfy the growing demands for high quality real-time broadband entertainment.

Here in Canada, the games are broadcast live on CBC and are simultaneously available for streaming from the CBC’s website.  Rogers is also making the games available with their Rogers On Demand TV and website services, which also provide live streaming to mobile devices.

CBC’s service uses the Akamai content delivery network to power their streaming, and the results have been stellar – just look at the detail in the screenshot below.

You can see the akamaihd network at work in the bottom left:

The World Cup possesses a combination of factors that are contributing to its significant Internet presence:

• It is the ultimate competition in the world’s most popular sport, and sporting events must be seen live by diehard fans
• It has global popularity, even in countries who didn’t send a team to South Africa
• Games take place over a period of many hours, spanning many time-zones, so a large number of workers engage in streaming activities at the office (where TV viewing is typically not available)

Other events in the last few years have possessed one or two of these characteristics (for instance, President Obama’s inauguration), but none have had all three to the extent that the World Cup has.

Many service providers have tools in place that can measure the bandwidth impact – one such tool is Sandvine’s recently-launched Network Data Analytics product.  One of the features of this product is a Network Summary dashboard that captures how the network is performing, and illustrates the shifting nature of Internet traffic profiles.  A particular tool on this dashboard is a series of “Top 25” lists that highlight popular websites, video providers, and applications.  One of the early adopters of Network Data Analytics is a Canadian service provider, and the impact of the World Cup on this provider’s network was immediate.

Here is a snapshot of the Top 25 Websites list from June 15th, note that akamaihd.net has surged onto the list.  The snapshot shows the traffic from the previous day, so on June 14th, akamaihd.net was the 12th largest website on this network, by total bytes.  We can also see that on June 13th, the site was the 19th most popular, and between June 13th and June 14th, akamaihd.net increased in total bytes by more than 80%.  Clearly, this service is coming out of nowhere to become a significant generator of traffic on the network.

Indeed, by June 18th, akamaihd.net had risen to second-place, behind only YouTube, and was still increasing its byte usage at an astonishing rate of more than 80% day-over-day.

By June 22nd, the growth had slowed somewhat, and YouTube still held onto its lead – however, let’s keep in mind that Canada is not competing in this World Cup, so the fact that CBC’s streaming is enjoying such success is a testament to the infectious nature of the event.

Figure 1 – June 15th, 2010	Figure 2 – June 18th, 2010	Figure 3 – June 22nd, 2010

While network use is good for service providers, it can also be a double-edged sword, particularly in the case of sensitive applications such as real-time streaming.  Content delivery networks certainly help to maintain high levels of subscriber quality of experience by optimizing delivery, but the network is still a limited and shared resource, so providers need to pay close attention to applications that suddenly explode in popularity.  In particular, network operators need to engineer their network such that it can accommodate abnormal surges of Internet traffic (some predictable, some unpredictable), especially when such events overlap with Internet peak usage periods, in order to safeguard the subscriber experience.

While some providers will reap the rewards of the World Cup’s popularity thanks to consumption-based billing and high subscriber satisfaction, others are struggling with overwhelming data levels and problems with filtering technologies.  Those service providers who have invested in traffic measurement and network policy control tools are in a perfect position to smoothly manage and benefit from this type of network usage scenario – for others, the 2010 FIFA World Cup South Africa might be just the event to sell them on the value of such tools.

So let’s talk about data or more specifically unlimited data and the growing number of service providers who are rolling out plans that better reflect user behavior. Its probably worth noting here that there were never really any completely unlimited plans, they all had some fair-use limit. The sea-change is in the overage charge model. O2 announced last week that they will end unlimited data plans for their smart phone customers later this month as did AT&T early in June (Source: http://www.engadget.com/2010/06/02/atandt-makes-sweeping-changes-to-data-plans-iphone-tethering-comi/). The million dollar question is “will users suffer”. And happily for Better Broadband the answer looks to be no.  “Based on current usage patterns, 97% of O2 smart phone customers would not need to buy additional data allowances. And according to the network the lowest limit of 500MB is 2.5 times the amount used by average customers on a monthly basis.” (source: http://www.thisismoney.co.uk/bargains-and-rip-offs/broadband-and-phones/article.html?in_article_id=506082&in_page_id=182&ito=1565)

If service providers can continue to offer fair packages that cover users needs (and then some) consumers can, and no doubt will, continue to embrace the variety of new and exciting things they can do with the mobile Internet.

How much data are you using? This really puts into digestible terms what you can do with a 500MB plan.

*Source: http://www.thisismoney.co.uk/bargains-and-rip-offs/broadband-and-phones/article.html?in_article_id=506082&in_page_id=182&ito=1565

Alongside the very Hollywood-esque show is CableNet. This always well-attended “product demo-focused” pavilion of the show, organized by CableLabs, provides insight into timely key themes such as interactive TV, video on demand, and broadband over cable. It gives operators the chance to see a broad range of product demonstrations within a very focused area.

Sandvine took the opportunity to demonstrate its latest product in our Network Business Intelligence Portfolio.  Network Data Analytics is focused on providing an executive level dashboard view of how subscribers are using their broadband service. It aims to answer questions like “How do my DOCSIS 3 subscribers use their service compared to the DOCSIS 2 subscribers?” or “Where will I get the best return on my CAPEX budget?”. It was great to see the many cable operators who  dropped by to explore. They were quite intrigued by the flexibility of the product and how it could potentially help them to enhance their businesses and their customers quality of experience. After all, providing better broadband is the end goal.

In addition to a wide range of operators, we were pleased to host FCC Chairman, Julius Genachowski, when he dropped by to see some of the trends that Sandvine is currently reporting on.

Were you at the show? Please share your better broadband observations and experiences.

The most popular island app was web browsing at 43% followed by email at 32%. Twitter and Facebook came in at 5% each and a bunch of you picked “other” which primarily broke-out into the websites for your favorite airline

I guess the results shouldn’t surprise – the vast majority consider the basic applications as the must- haves. We can all wait until we get home to resume gaming and watching movies.

So what does this mean for mobile operators? We believe that as more and more applications are offered on mobile devices, operators are looking to predict usage patterns so they can offer personalized service tiers that increase customer satisfaction, and to build out their networks to match user demand. AT&T recently stated that “5 percent of users account for approximately half of the data traffic on average” so clearly flat-rate billing does not align with subscriber usage (see Fierce Wireless: Mobile data traffic patterns look similar to fixed broadband patterns).

And for subscribers? Hopefully it means more and more applications along with appropriate service tiers that economically keep our lives connected to the Internet, everywhere.

Some of the high-level findings of Sandvine’s report include interesting observations like social networking accounting for up to 9% of total bytes on any given mobile network and YouTube accounting for 10 to 15 percent of total bytes on any given mobile network. To download the full report visit http://www.sandvine.com/news/global_broadband_trends.asp.

We’ve already talked about subscriber expectations regarding web page load times on broadband connections.  However, consider a subscriber’s expectation of a web page load time on a dial-up connection, or on free Wifi, or on a mobile handset.  Most subscribers would expect the load time to vary across these technologies.  Now, what if instead of talking about a web page, we examine streaming video, or voice-over-IP, or downloading a movie file…would the expectations be consistent?  And here we find the biggest challenge in measuring Quality of Experience – expectations change based on technology, application, and subscriber.

We can loosely identify some characteristics that differentiate a good experience from a bad one:

• Web Browsing: pages load quickly, without errors
• Voice and Video Conferencing: voice/video is smooth, conversational, and audible/visible
• Gaming: movements are smooth, network is responsive, you don’t get ‘fragged’ or ’pwned’
• On-demand Entertainment: video or audio starts quickly, there is no re-buffering, no compression artifacts, video and audio are synchronized
• Bulk Downloads: completes in a reasonable time

Of course, words like “quickly”, “smooth”, “responsive”, and “reasonable” betray a certain amount of subjectivity and have the potential to greatly from subscriber to subscriber.

To muddy the picture even further, subscribers vary greatly in the applications they use.  A subscriber who relies on the Internet for web browsing and e-mail could have a terrific experience on a network that is hated by a hardcore gamer with a roommate who loves BitTorrent.  Finally, the consumer hardware will impact perceived quality, too.  Imagine having a 50 Mbps connection that is squeezed through an 802.11b router before heading to a computer that was state-of-the-art in 2001.

Minimally, different applications require different measures of QoE.  Web Browsing QoE can be measured in terms of the average time to load a page, VoIP QoE can be assessed by combining quantitative measures such as R-factor and Mean Opinion Scores (MOS), and gaming QoE can be determined through some measure of network latency, packet loss, and jitter.

Once a service provider is confident that application QoE can be accurately determined, then they can apply network and subscriber models to determine the average QoE across the entire subscriber base, or even for individual subscribers.

Tara and I discussed the concept of “better broadband” and what that means for subscribers, operators and the network overall.

To view the article visit:

http://www.xchangemag.com/articles/billing-better-broadband.html

Two of those factors were latency from the client to the content servers and latency associated with DNS.  Various techniques have been implemented to reduce both these latency figures in practice.

Today, most things viewed online are served off a content delivery network (CDN), such as Akamai and Limelight Networks. These companies position content servers as close as possible to the ‘subscriber edge’ of the network (as close to your house as is possible), so that subscribers get a low-latency, high-bandwidth experience when consuming content served by the CDN’s customers.  This is generally a win-win-win-win scenario, with consumers getting better quality of experience, Internet providers experiencing lower transit (and happier customers), content owners delivering increased reliability, and the CDN provider getting paid.

Most CDN providers use tricks with DNS to optimize the experience for the end subscriber.  One such trick is geo-targeting, in which the same hosts (for example, Facebook) will resolve to different IP addresses depending upon the origin of the look-up.  In other words, a subscriber in a particular city will get a different IP address for Facebook than a subscriber in another, even if the cities are relatively close together.

Now, what happens if you override your DNS settings so that instead of relying on your Internet provider’s DNS servers, you actually direct all your queries to a global service (for example, OpenDNS or Google Public DNS)? You might buy into the message of “better caching equals better DNS performance”, even though you know that the latency to these public servers is much greater than the latency to your ISP’s server.  But did you know that you’ve just broken the CDN model?

You can easily test the difference. Resolving Facebook via a public DNS server returns a particular IP address (for example, 72.246.43.8), which a ping test tells us is 40ms away.  Now, using my Internet provider’s DNS server returns 67.69.247.135, which is only 7ms away.

While the extra 33ms might not sound like a lot, when it’s added to each round-trip and considered in light of all the serial and parallel TCP connections required to load a typical web page nowadays, then we’ve suddenly added a full second to the page load time (and we’ve already seen that the difference between a 2 second load time and a 3 second load time is the difference between a happy Internet user and a user who closes the web page in disgust).

Join in for the final part in this series next week, “Measuring the Quality of the Internet”