Ask an engineer to describe the greatest hurdle in delivering next-generation entertainment services to the home, and most will give the same answer.
The problem does not lie in the operators’ back haul networks. Mostly, it no longer lies in the “last mile” between the service provider’s network and the consumer’s home.
Instead, the problem lies squarely in the home itself. The difficulty is bringing rich media content to where consumers want to watch it.
In a typical European house – with the exception of recent construction – the main phone socket will be in a hallway or on a landing. Some homeowners will have added extensions or second lines in the study or den for a broadband computer connection. But few will have anything other than an antennae socket in the living room.
The situation in North America is a little better, with more homes wired up for cable TV and home working. But bridging the last 10ft or few metres between the broadband network and where consumers want to watch TV or listen to music remains a challenge.
Wiring a 2,000 sq ft property with Ethernet (Category 5) cables costs between $1,000 and $1,500, takes four hours and requires two labourers, according to Ruckus Wireless, a vendor of wireless multimedia networking gear.
Ruckus calculates that wiring costs add between $100 and $300 to the per-subscriber start-up costs for an Internet TV (IPTV) service.
Even alternatives, such as re-using existing TV coaxial cable, or using the domestic power circuits to carry data, face challenges. The equipment remains relatively costly, often needs professional installation, and data speeds can be affected by interference from other devices.
For newly built properties, especially at the higher end of the market, an Ethernet based network remains the preferred solution. A house builder can install the cabling at the same time as electrical circuits, and homes ready-wired for multimedia can attract a premium in the marketplace.
For existing homes, as well as less expensive properties, wireless networking looks to be the strongest contender. But the industry appears split over whether a standard or proprietary approach will deliver the best performance and price.
Existing wireless LANs, based around the 802.11b and g standards, only provide limited support for multi-media applications such as video or live TV. The throughput is low but, more importantly, these networks are not designed for streaming media, which needs constant and consistent throughput.
“Most of the people developing wireless devices don’t understand the problems of video,” says Mark Gray, CEO of Internet TV technology company Kasenna. “They think in terms of error correction as best effort or retransmission. That doesn’t work in video: in video you cannot drop frames.”
The problem is commercial as well as technical: an advertiser that has paid for a slot in a major sporting competition such as the Superbowl or World Cup would be looking for money back if viewers missed part of the transmission.
The emerging 802.11 “e” and “n” standards promise to solve some of these problems by supporting both higher maximum transmission speeds and by introducing quality of service and error correction protocols that are better suited to video streaming, as well as to applications such as voice over IP.
Better antennae technologies are also being touted by manufacturers as a way to solve problems that affect wireless networks in the home. These include interference from devices such as microwave ovens and cordless phones, and the difficulty of transmitting through dense materials, such as brick and steel.
MIMO, a multi-directional antennae technology that forms part of the 802.11n standard, is already being used for home wireless data networking. Significantly, the technology is being adapted for use in IPTV set-top boxes and other consumer electronics hardware.
Wireless technology company Airgo Networks has developed MIMO-based systems that can be installed directly in set-top boxes, and that can use either the 2.4GHz or 5GHz radio spectrum in order to avoid interference. One European service provider is testing the devices with consumers.
The semiconductor arm of Philips also recently announced that it will use 802.11n technology in the 5GHz band from Metalink for its Connected Home chipsets.
These chips are designed to go directly into devices such as TV sets and DVD players; Philips claims that the technology can deliver multiple, high-definition TV streams up to a range of 100ft (30 metres).
But devices incorporating such technologies are some way from widespread availability. Philips, for example, will be making the chips available in the fourth quarter of this year, and there will inevitably be a delay before manufacturers can incorporate the chips into their devices.
For IPTV companies that want to deploy networks now, that leaves two real choices: a proprietary wireless system, or cables.
Proprietary wireless systems can handle the 10-20Mbps of sustained throughput TV needs. Even if the operator has to subsidise the cost, it will still be cheaper than cabling in most cases. But operators may well have to offer a combination of solutions, including 802.11n, powerline and Ethernet in order to service all their subscribers.
“There is no one right answer, and operators will have to use different techniques even within a single geography,” says Ed Graczyk, director of marketing for Microsoft TV.
“But at the same time, IPTV devices are a lot less expensive than dedicated set top boxes for cable or satellite, and it can act as a home gateway.” This, he suggests, will go some way to offsetting the cost of completing the home network.