.gif)
Access technology
With more and more ‘bandwidth heavy’ services scheduled to come on stream, the issue of network capacity is now a high priority for service providers. Keith Donahue looks at how the industry can meet the expected demand
The mobile content industry is quickly clearing the barriers to its own success: the first wholesale 3G and GPRS services are launching to overcome high data charges; Apple's DRM policies are being challenged in court; and companies like Nvidia are starting to drive through standards for high-quality 3D gaming. But one hurdle remains to be cleared.
While all the discussion is of the 'sexy' technologies for delivering and experiencing content, the method for actually getting this content to the users has been overlooked. At the access level, all the talk is of HSPA and WiMAX: but operators need to examine how they get content to the network edge.
Most networks were built for voice services, with data as an afterthought. Even 3G networks typically only offer a few megabits to the base station. Yet new access technologies are offering multiple megabit downloads and real-time streaming.
Network evolution
As the capabilities of mobile devices have evolved over the last 15 years, the network has also developed to support new features. Starting with the introduction of digital voice over GSM networks in 1990, the evolution of mobile services can almost be plotted on a chart of bandwidth requirements against time: text services added in the mid '90s; camera phones and photo messaging in 2000; cached video services in 2004; and now live streaming of mobile TV.
In the last five years, the provisioning of capacity for each new cell site has more than doubled. Where once one or two E1 circuits were the norm, now four times E1 is the minimum. And even this is based on educated guesswork about the future (most likely conservative) rather than a real knowledge of what the demands of subscribers and applications will be. Herein lies the operators' major problem. The prospects for mobile broadband content are good, but the timetable for their success remains uncertain. Most service providers come from a voice centric background and lack the required experience in planning for and delivering data over their networks. For example, some carriers are happy provisioning 4xE1 today, while others are planning up to 16xE1. There really is no consensus on the expected demand.
It's a cliché now, but nonetheless true that no one wants to repeat the failure of WAP. With expensive rights deals, and huge potential revenues, the risk of underwhelming subscribers with poor quality content services is just too great. Yet this remains a very real possibility, unless service providers begin to address the potential bandwidth deficit in their networks. Estimates suggest that video services alone could saturate the capacity of many cellular networks by 2007.
The issues that dogged WAP were many, but fundamental to the problem was the lack of bandwidth available in the last mile. Even with 2.5G, the experience was nowhere near that promised by early images of the wireless web surfer. At bit rates of barely 30Kbps, this was hardly surprising. Today, this problem has been largely overcome with 3G and High Speed Downlink Packet Access (HSDPA). But rather than solving the problem, this has moved the bottleneck deeper into the network – specifically into the backhaul segment. As noted above, 3G networks – rolled out in the last few years – have typically been based on the provision of 1-2xE1 per base station. Before taking any other overheads or bandwidth use into account, this gives a theoretical maximum of 4Mbps. Just one unrestricted user with an HSDPA-enabled device could consume this bandwidth five times over.
Traffic control
Of course, this situation is merely theoretical. It will be some time before any mobile service could consume such enormous volumes of data. But the introduction of rich content to the service mix will place enormous strain on any current architecture. Streaming video is a prime example of this.
Video, like voice, uses a constant bit rate, but the amount of bandwidth required by video is exponentially higher. Any degradation in the available bandwidth will seriously impact frame rates, picture quality and hence the user experience. While 8kbps is sufficient for a good quality voice call, the 3GPP recommends that up to 32Kbps would be required for text distribution with low-quality video, where low-quality video is defined as three frames per second (fps). This rises sharply to 128kbps for audio distribution with low-quality video, and all the way to 384kbps for video streaming, where the quality is not defined.
Clearly, if video services become popular, it won't require many users connected to the same cell site, making calls, browsing the web, downloading games and watching streamed content to exhaust the available capacity. When this happens, service quality and the consumer experience will begin to degrade.
Of course this isn't just a mobile telephony issue. Managing a mix of traffic with different requirements has been a conundrum in the fixed line world for many years. The solution – at least in theory – is QoS/CoS/SLA controls that assign priorities to the most important data traffic. These are generally the packets associated with constant bit rate services such as video and voice. However, while these tools can certainly help in a situation with limited capacity, some services and users will inevitably suffer when a finite amount of bandwidth is being shared between users. The long-term solution has to be more bandwidth.
Build a better network
The problem, of course, is that no one knows precisely how much more bandwidth is required. While provisioning 16xE1 to every base station may provide long-term flexibility, it is unlikely to be economical. One unlikely source of guidance is from service providers in developing countries. Here it is the extremely rapid adoption of mobile phones that has caused a headache in network planning. In Nigeria, for example, new cell sites have been saturated within just five hours of going live because of intense demand. The solution for Nigerian carriers has been a phased approach to adding capacity, using point-to-point radio-based platforms with the scope to scale over time. This means that the right amount of bandwidth can be supplied at any one time without the additional cost of over-provisioning.
Whatever the approach, the major issue is ensuring that those marketing the new services are aware of the problem and are communicating with the network planners to ensure a solution. Adding the capital expenditure of a network upgrade into the cost of launching new services could make the difference between profitability and disaster. If not, and the problem is ignored, we could end up with another WAP.
Until this issue has been resolved, the user's experience of downloading mobile content could be so constrained that it may not be one they will repeat – however good the content.
Keith Donahue is Director Business Growth Programs, Harris Microwave Communications
Share this article with others
Post to del.icio.us
Printed from http://www.eurocomms.com/features/111501/Access_technology.html
Comment on this article
Skip to comments
We encourage users to analyse, comment on and even challenge European Communications's articles, including the one above - 'Access technology'
User reviews and comments that include profanity or personal attacks or other inappropriate comments or material will be removed from the site.
Additionally, entries that are unsigned or contain "signatures" by someone other than the actual author will be removed. We will take steps to block users who violate any of our posting standards, terms of use or privacy policies or any other policies governing this site.