Wiring the world

Every few years, our homes and neighbourhoods threaten to become “smart,” and quite frankly it gives me the willies. The Internet is the latest in a long line of suitors that promise to bring us intelligent buildings, smart communities and houses that know more than we do.

I have a file folder overflowing with clippings dating back to the early 1980s about how computerized systems were going to reduce the cost of energy, automate facilities management and provide safety, security and state of the art communications. Until now, though, computerization has had little effect on our domestic lives. Our homes have remained relatively calm havens from the high-speed, high-tech world of automation.

That is changing, however, as the internet tidal wave revolutionizes every aspect of our lives. The internet now has the potential to provide a common protocol and sufficient bandwidth to control and connect every major appliance, machine and piece of equipment in our homes, schools and offices. The telephone companies, cable companies and even subdivision developers are betting that now they have you on-line and sending e-mail, you’ll be an easy mark for a whole slew of other services. Those services will be transmitted over a host of competing broadband networks.

Networks are like onions. They are designed in layers that can be peeled away until you’re finally left with just the wires and cables that carry the information. This is called the physical layer. The confusion begins with the fact that each layer has its own set of competing protocols, or the set of rules that networks use to ensure that ungarbled communications take place.

Protocols can be jammed together in an amazing number of configurations. For example, TCP/IP is the name given to the most common protocols of the internet. It consists of the Internet Protocol (IP) which operates at the network layer and concerns itself with sending packets of information from one computer to another. The Transmission Control Protocol (TCP) is used in the transport layer — the layer that takes the incoming stream of bytes, breaks it into manageable chunks and passes it along to the network layer. Protocols such as TCP/IP are open protocols in the sense that their detailed specifications are freely available.

At the moment, the telephone companies with their DSL (Digital Subscriber Lines) products are duking it out in the physical layer with the cable companies with their cable modems. Unfortunately, both of these approaches are based on old delivery systems — twisted pair copper wire in the case of DSL and copper coaxial cable in the case of the cable modem. In the long term, both these stopgap measures will be blown away by fibre optic cable with its ability to carry what seems like unlimited bandwidth.

The turf wars are threatening to become even more intense as traditional boundaries and regulations are swept away and new players enter the field. It is estimated that at least 37 million kilometres of fibre had been laid worldwide by 1999, and this amount is predicted to grow 500 times larger by 2002. Quite apart from the usual players, new Canadian companies such as 360networks are staking a claim to some of those kilometres. The company, which acts as a wholesale provider of bandwidth to telco (telephone company) and internet service providers, estimates that by the end of 2001 they will have installed some 90,000 kilometres of fibre around the world. When fully operational, some of their networks will be capable of speeds up to 1.28 terabits per second — that’s about 44 million times faster than the 28.8 Kbps modem that sits on my desk.

360networks began as the telecommunications division of Ledcor Industries, a construction company originally based in Alberta. According to Michelle Gagn, director of corporate communications, “360networks inherited a great deal of construction expertise from Ledcor that allows them to build networks faster and for less expense than their competitors.” She also highlights an innovative rail plow technology that allows them to lay fibre optic cable along CN railway tracks [R.V. Anderson Associates of Toronto helped design the system.]

Another Canadian company, Futureway, has also staked out new territory. Futureway is a telecommunications company that provides telephone, television, data and internet services to subdivisions in southern Ontario. It runs fibre directly to the home. The company maintains an exclusive right to lay telecommunications cables in those communities. Companies such as Bell and Shaw traditionally charge developers to install their cables to homes in new subdivisions. When Futureway installed its own cables, Bell still wanted to charge the developer the same rates to deliver its services over Futureway’s wires. The CRTC, however, ruled that Bell could not charge tariffs on cables that weren’t their own. This decision opens the door to intense competition in the wiring of new developments.

There’s also a battle brewing in the air as wireless companies begin to offer internet services. Collaborative efforts such as Inukshuk Internet are focused on providing Canada with a state-of-the-art wireless IP access network. In five years they intend to reach some 70 per cent of Canadian households who will then be able to access the internet through handheld devices like Palm Pilots.

The last mile

Above or below ground, the real battleground is the “last mile” — that part of the network that connects directly to our homes and offices. Huge fibre backbones such as CA*net3 (deployed by CANARIE, the Canadian Network for the Advancement of Research, Industry and Education) already snake across Canada. But up until now it has been very expensive to deliver significant bandwidth to our living rooms. And in the world of telecommunications it’s all about bandwidth — the number of bits of information you can shove down the tubes during some unit of time.

Both telephone company DSL and cable company modems claim to deliver in the neighbourhood of 10 Mbps (or 10 million bits of information per second), but the reality is that you’ll only attain speeds that range from 60 to 150 Kpbs (or thousands of bits per second). Broadband technologies using fibre promise speeds in the range of 100 Mpbs and are creeping towards Gigabit (or billions of bits per second). Wireless services, while they hold great promise for connecting rural and remote areas, are limited by their bandwidth, which is currently in the range of four Mbps, although it will grow as the technology improves.

What puts fibre into a class of its own is the fact that today’s most advanced networks are transmitting light instead of electrical signals. Optical networks allow signals to be carried on different colours of light. Each colour can carry a gigabit or more of information per second, and as the technologies become more efficient a single fibre will be able to carry thousands of colours.

What can you do with bandwidth like that? Well, many people in the telecommunications business see live, full-screen, full-motion video as the Holy Grail of networking. This kind of video currently requires 40 Mbps. Do the math for yourself: 25 video streams of that bandwidth will chew up an entire gigabit of bandwidth. We’re going to need a lot of fibre to wire up this country and copper wire just can’t cut it.

Some engineering firms are already heavily involved with wiring up whole communities. The IBI Group has its own telecommunications division spread between its Toronto, Vancouver, Denver and Seattle offices. According to Peter Zurawel, a director with the firm, it is “deploying city-wide networks for telecommunications companies, which combine voice, video and data.” They have projects on the go in Puget Sound, Portland, Salt Lake City and Washington D.C. For these networks Hybrid Fiber Coax (HFC) is the physical delivery mechanism. As the name implies, this is a combination of fibre and coax cables. Optical fibre is run from a central office to a neighbourhood where it is distributed to some 600 homes through standard coaxial cable with standa
rd coax bandwidth.

Zurawel points to infrastructure issues that are often overlooked in the deployment of these networks. The Puget Sound network, which will serve some two million homes, will also include 55 buildings for the service and maintenance of the network. About a fifth of these will be inhabited for training and administration. A critical concern for IBI is being able to reconfigure these facilities quickly.

Dessau-Soprin, consulting engineers with offices throughout Qubec, have also installed a number of innovative networks. The Centre Hospitalier Universitaire de Qubec (CHUQ) connects three health care facilities in an integrated network that provides voice, video-conferencing, tele-consulting and imaging capabilities. In this instance yet another protocol was used: ATM or Asynchronous Transfer Mode. ATM is a protocol that transmits all information in packets of a fixed size called cells. Because of its unique form, ATM is considered to have its own set of layers that sit on top of the physical one. Michel Famery, ing., the vice president for telecommunications at Dessau-Soprin, explains that “For CHUQ, ATM was chosen because of its robustness and efficiency.” This is because the cells in an ATM stream are guaranteed to arrive in the order they were sent, which is particularly important with video applications.

A few years ago ATM was considered the wave of the future but it is losing ground to simpler protocols. When you’re wiring up whole neighbourhoods or provinces for telecommunications you’re talking about WANs, or Wide Area Networks. The latest wrinkle is that simple, cost-effective LANs (Local Area Networks), the technology used to connect devices in the home or office, can now be deployed on a WAN scale.

Ethernet, for example, is a LAN protocol that was originally designed by Xerox PARC, Intel and DEC as a means of letting computers talk to one another and to printers. Recently, however, it has become possible to use the inexpensive technologies of Ethernet to span large distances. The Netera Alliance, a not-for-profit corporation working with advanced networking and computing in Alberta, has just deployed a 370 kilometre Gigabit Ethernet Network linking Calgary and Edmonton. The idea behind Gigabit Ethernet is that if you throw enough bandwidth at a telecommunications problem then it goes away. You don’t need the guaranteed service of ATM because there’s more than enough bandwidth for all the videos being sent.

Laundry day the new way

It’s the applications not the protocols, however, that will make or break the acceptance of broadband networks. A lot of the new scenarios revolve around your refrigerator suddenly realizing that you’re almost out of milk and sending an emergency all-points-bulletin to the nearest grocery store to have that precious commodity delivered in record time.

Another idea is that you’ll be able to watch a baseball game on screen and select the camera angle you want to watch from. In real life, the applications are a little more prosaic. Echelon’s Lonworks system is a control network for connecting and monitoring everyday devices. For example, Lonworks products will be integrated into washers and dryers from Merloni, an Italian appliance manufacturer, and by working through the existing electrical power wiring, these devices will be able to balance energy consumption and ultimately reduce costs. This is what Steve Nguyen, director of marketing for Echelon, calls, “intelligent power.” Nguyen suggests that all new construction should be wired with Category 5 (or Cat5) cable that can carry 100 Mbps in order to facilitate communications between devices.

In a similar fashion, Nortel has begun developing what they call the Open IP Environment, which will allow consumer appliances to communicate through the internet. The real promise of these home or office networks is that they will be combined with wireless internet technologies. From your laptop you’ll be able to send e-mail or print documents without wires no matter where you are. In this case the smaller bandwidth of wireless systems is less of a problem since you’re already inside the house and focusing on a single application.

Naturally these wireless developments are leading to a whole new set of protocols. Bluetooth, for example, is a standard being developed by a consortium of over 1,600 companies including Ericsson, IBM, Intel and 3Com/Palm. It uses radio waves to allow devices to communicate within a range of about 10 metres. Bluetooth-enabled devices will carry miniature radio transceivers and should start coming on the market later this year.

By now you should be completely overwhelmed and hopelessly confused by the profusion of different standards and protocols that are crowding on line. Is there any way to wire our offices and subdivisions so that they maintain some form of long-term viability? There is no clear answer, but Bill St. Arnaud director of network projects at CANARIE, makes the pitch for an innovative “third residential network” in a paper entitled “Gigabit internet to every Canadian School by 2005.” Rather than integrate cable and telephone services into one network, he advocates building another network to the home that is dedicated to the internet only. He believes in this divergence, rather than convergence, in order to avoid the regulatory and technical constraints of telephone and cable services.

St. Arnaud’s scheme is predicated on the use of neighbourhood competitive access interconnection points (NCAIP). One of these would be located in each neighbourhood and would be the point where each house could connect to the service provider of their choice. St. Arnaud suggests that governments should require companies who provide publicly-funded internet access to Canadian schools and libraries to simultaneously deploy a similar infrastructure to our residential areas.

Make no mistake, the manner in which Canada deploys its high-speed networks will have an impact on its productivity for years to come. While Canada is an important player in the world of the internet and some 75 per cent of the world’s internet traffic flows over Canadian equipment, we are facing stiff competition from other countries. Finland and Sweden have both embarked on government-assisted programs to wire their countries. Stockholm, for example, has invested $100 million in a shared fibre optic infrastructure. And on the other side of the world, Japan has a considerable lead in the deployment of wireless networks.

Nonetheless, at the municipal level, Canadian cities are being energized in this area. Ottawa has just announced an agreement with Vidotron to create a community optical fibre network for the city. Toronto’s Telecommunications Steering Committee has called for a city-owned fibre network. Even smaller cities such as Brossard, Qubec have issued requests for proposals to develop their own civic networks

But if the deployment issues are sorted out, there is still the thorny dilemma of applications. MP3, the software for downloading music, came out of left field to be the killer application that demanded bandwidth and created a whole new class of users. Unfortunately, since most of the music was illegally provided for free it didn’t provide a robust economic model and many organizations, including universities and internet service providers, were forced to curtail the use of MP3 because it was overwhelming their networks. Then, too, most video-on-demand trials in the U.S. and Canada were abject failures. Will people really flock to the internet to watch television? Or is it more likely that we’ll have whole houses filled with appliances flashing midnight like our VCRs?

There’s been a great deal of talk about the “last mile,” but what about the “first inch” in which the designers of these new networks and services carefully consider how they can be used easily, simply and elegantly? One article from the Globe and Mail which I have kept since 1985 describes how residents of an Ontario Housing Corporation apartment complained that they were freezing because of a new computerized heating system, and were informed that t
hey were, “just not accustomed to the system.” Many of these new applications will fail, not because of the technologies or the bandwidth involved, but because they are too difficult to use and their designers are too arrogant to consider the real needs of their customers. While the era of the internet may herald a golden age of communications, no amount of bandwidth can compensate for poor design.

Douglas Macleod is the director of projects for the NETERA Alliance of Alberta and is based in Calgary. He thanks Bill St. Arnaud for assistance with this article.