Think small

Small networks--strategically located in offices, homes and local retail stores--are one of the biggest opportunities to hit the telecommunications sector. 

What makes these locations strategic is their relationship to the Internet. Small networks are deployed at the Internet's edges, creating the side streets that feed (and are fed by) the information superhighway. They enable the Internet to extend multiple tentacles deep into the places where most people live and work. Small networks might be described as the Internet's eyes, ears, and hands.

RESEARCH REPORT Digital Dwellings: Technologies & Business Opportunities for Small Networks February 2001 Datacomm Research  254 pages $1495 electronic or printed version $1695 for both  Order this report

Most of what is written about small networks sounds futuristic at best and silly at worst. Perhaps some day your refrigerator will talk to your electric can opener; it is a much better bet, though, that your hi-fi, TV and security system will soon talk to the Internet. Small networks present significant near- and medium-term business opportunities for hardware developers, software and content producers, and service providers.

There are four major applications for small networks: 

• business communications

• entertainment

• education and

• security/safety. 

There are far more small businesses than large, and most start in the home. People understand that you have to spend money to make money, and money spent on advanced communications is money well spent. Though it took a little longer for computer networks to trickle down to small businesses, they help make small businesses as efficient and professional as large businesses.

People in developed countries take their entertainment seriously, and the amount they spend on entertainment each year proves it. U.S. consumers, for example, spend about $200 billion annually on entertainment.

Entertainment may be a luxury--but that's not how most people look at it. People in developed countries take their entertainment seriously, and the amount they spend on entertainment each year proves it. U.S. consumers, for example, spend about $200 billion annually on entertainment. For better or worse, many people want to have their music with them at all times--much like the background graphics on a PC screen. Others feel the same about video entertainment.

The Internet has become an essential tool for education. Schools routinely assign homework requiring that students look up answers on the Internet. Households with two or more school-age children will want a second or third PC to ensure all of the kids can go online when needed. Plus, PCs usually become obsolete within two to three years, and households tend to accumulate PCs as they upgrade.

Small networks provide more ways of ensuring security and safety, though people do not spend as much on these things. An alarm system connected to a PC can turn on lights. Sensors for natural gas, carbon monoxide and water can be easily added to the network. Video cameras can be installed at doors and gate entrances. As the U.S. population ages, more people will desire an "always on" connection to healthcare providers.

As the Internet becomes an increasingly essential tool, more people will subscribe to high-speed access services provided by cable operators, telephone companies and wireless carriers. Sharing high-speed Internet access is becoming a killer application for small networks, but it is by no means the only application.

Internet connectivity will spur small networks in ways we can only begin to imagine. The ability to "look out" to the Internet from a small network, or "look in" on the small network from anywhere via the Internet is a powerful argument for installing a small network. Suddenly, the utility of existing products begins to multiply, and new services that would have been impossible just a few years ago start to spring up.

Small networks can "look out" to grab Internet content on demand. Music and videos are obvious examples. There are less obvious examples, too, such as a lawn sprinkler system that grabs the local weather forecast to create its own watering schedule. Security and safety systems can alert remote service providers or an absent homeowner (for example, via an Internet-based alphanumeric paging service) when an alarm is triggered.

Indeed, users and remote service providers can look in on small networks to monitor events, diagnose trouble and remedy problems. Working parents can communicate with their teenagers, check the status of household systems and even look at video stills. Appliance makers can instruct a washer that won't start to run a self-test and upload the results. Some problems may be fixed (or temporarily circumvented) by downloading new software.

The opportunity for small networking gear is significant, but it is dwarfed by the even larger opportunity that surrounds it. Small networks will spur sales of new entertainment systems, household appliances and personal gadgets. Manufacturers will leverage networking capabilities to sell families of products--not just individual products. Consumers may replace certain devices sooner to take advantage of new, network-enabled functions.

Table 1 Forecasted Sales of Home Networks U.S./Canada, 1999-2006
YEAR	1999	2000	2001	2002	2003	2004	2005	2006
AMOUNT	100	500	1,000	1,600	2,300	2,800	3,200	3,500
(Dollar amounts are in $millions)

Competing Technologies

An amazing variety of technologies are converging on the small/home network market. There are different transmission technologies for various media; microcode to ensure plug-and-play compatibility; and software for interfacing and managing devices attached to the network.

Transmission media can be divided into three major categories:

• new wire

• existing wire and

• wireless. 

Today, the leading choices are Ethernet, HomePNA, powerline communications and wireless local area networks. The typical small network will start with an Internet access-sharing backbone. Subnetworks, often using different media, will be gradually added to extend the network's reach. Thus, we may see PCs interconnected via Ethernet or HomePNA, and the network extended via wireless and powerline communications (PLC). There is a good chance that multiple transmission technologies could win, but proponents of each technology need to be realistic about where they fit in.

Ethernet will continue to be a major force in the small office. Ethernet is fast (10 Mb/s); it offers a credible upgrade path (fast Ethernet at 100 Mb/s and gigabit Ethernet at 1 Gb/s); and there is plenty of Ethernet hardware available at competitive prices.

Ethernet will continue to be a major force in the small office. Ethernet is fast (10 Mb/s); it offers a credible upgrade path (fast Ethernet at 100 Mb/s and gigabit Ethernet at 1 Gb/s); and there is plenty of Ethernet hardware available at competitive prices. Many PCs sold to businesses come with Ethernet adapters already installed. Plus, small businesses are willing to run new wire, particularly if they have a long-term lease or own the building.

However, Ethernet will not be a big winner in the home. Ethernet requires the use of hubs and either coax or qualified twisted pair cabling. Consumers will be attracted to technologies that use existing wiring (HomePNA and PLC), avoid running new wire (HomePNA, PLC, and wireless), or support mobility (wireless). Many PCs intended for use in the home feature pre-installed HomePNA adapters.

HomePNA is the leading technology for home Internet access sharing and is likely to remain on top for the next three years. It works with ordinary phone wiring, is available in 1 Mb/s and 10 Mb/s versions, does not require hubs, and is inexpensive. It can coexist with telephony and xDSL (high-speed Internet access). And consumers are already accustomed to running new phone wiring to reach new locations within the home.

But HomePNA has two major drawbacks. The condition of home phone lines is not always good enough for high-speed data. While HomePNA can coexist with voice and xDSL, transmission errors may ensue, reducing throughput. HomePNA only runs up to 10 Mb/s, which is acceptable for Internet access sharing but not for video distribution.

Powerline communications sounds attractive, but may promise more than it can deliver. PLC offers the obvious advantage of extending the network to virtually every room, many of which have multiple power outlets. And nothing could be easier for consumers to install; they simply plug into standard power outlets.

But electric power lines were designed to distribute electric power, not provide two-way data communications. Power lines are notoriously noisy, and common household devices such as light dimmers are often the culprits. It is not safe for ordinary consumers to modify or troubleshoot powerlines. PLC promises speeds of 10 Mb/s and faster, but new wire, HomePNA, and wireless are already delivering it. Ethernet and wireless promise speeds powerlines may never achieve--54 Mb/s and higher.

PLC is a good choice for relatively low-speed applications, such as remotely turning lights off and on, or querying household appliances. But powerline communications is probably not the way to go for Internet access sharing and digital TV. New sources of noise may appear on home electric wiring, creating intermittent interference that may be difficult to pinpoint.

PLC is also not a good choice for use with PCs, because the computer, monitor and various peripherals often take up every available AC outlet--even when power strips are used to multiply the number of devices that can be plugged in.

However, powerline communications does have a major advantage in serving sensors, remote switches and household appliances. Consumers will not want to run new cables to washing machines, refrigerators and light switches. Low-speed powerline communications is inexpensive and consumers need not do anything special to add household appliances to existing PLC subnetworks. Household appliance manufacturers, meanwhile, could benefit immensely from remote access to their products, performing remote diagnostics prior to service calls.

The long-suffering wireless LAN industry is finally enjoying success. Wireless LANs are poised to be a major player in small office and home networks. Wireless is the best choice for hard-to-reach locations, or to avoid the cost and disruption of running new wire. And surprisingly, wireless offers a compelling solution for migrating to higher speeds. Wireless LANs using orthogonal frequency division multiplexing (OFDM) in the 5 GHz band (based on the IEEE 802.11a and HiperLAN/2 standards) promise near-term speeds up to 54 Mb/s and 100-200 Mb/s further down the road.

Today's wireless LANs offer speeds up to 11 Mb/s--in products based on the IEEE 802.11b standard using direct sequence spread spectrum (DSSS). HomeRF, which uses frequency hopping spread spectrum (FHSS) is also competing for the home market, offering over 1 Mb/s today and promising 10 Mb/s tomorrow.

There has been a long seesaw battle between these two versions of spread spectrum radio technology. In the early days of unlicensed spread spectrum, DSS grabbed the lead, offering superior price/performance. Then FHSS surged ahead by demonstrating superior noise and interference immunity, which are important attributes in many vertical industry applications.

Noting the proliferation of PCs in the home, the leading independent wireless LAN manufacturer Proxim decided to develop a solution specifically for the home. Proxim put together a coalition of companies to develop and promote HomeRF. While HomeRF has proved an acceptable interim solution, it was based on the incorrect assumption that home networking would be less demanding than office networking. With its emphasis on entertainment and multimedia, the home is arguably the most demanding local area networking environment.

Thus, Proxim's HomeRF created an opportunity for DSSS proponents. Led by Harris Semiconductor (now Intersil), DSSS proponents recognized the need for higher speeds and the fact that their technology could get their first. Thus, IEEE 802.11b wireless LANs now run 11 Mb/s--more than five times faster than the original HomeRF standard--and are far more popular.

In response, the HomeRF Working Group is developing a 10 Mb/s specification, but it will be too little, too late. Three major technical developments are in the works that will make wireless LANs both more robust and more affordable. A planned enhancement to the IEEE 802.11b standard will deliver 22 Mb/s performance. The use of OFDM, combined with CMOS semiconductor technology developed for the 5 GHz band, promises speeds of 54 Mb/s and faster.

Thus, the IEEE 802.11a standard is poised to be the big winner. OFDM transmits data simultaneously over multiple radio carriers and, therefore, shares most of the advantages of spread spectrum. It also opens the door to a highly scalable radio technology, with smaller, lower speed and less expensive devices using fewer carriers.

It is even conceivable that within a few years IEEE 802.11a wireless LANs will offer higher throughputs than most wired LANs. The use of CMOS should permit wireless LANs to ride the price curve down in similar fashion (as described by Moore's law) to microprocessors. Thus, we may see people choosing wireless not to avoid wire, but simply because wireless works better and cost less.

There is also a need for a complementary wireless solution for linking PDAs, portable computers, smart phones and cordless phones to each other, desktop devices and nearby access points. The Infrared Data Association (IrDA) addressed this need, and nearly succeeded: infrared transceivers became fairly ubiquitous, but users never warmed up to the technology.

That's why Ericsson's short-range, low-cost radio solution, Bluetooth, was greeted with great enthusiasm when first unveiled. Everyone agrees there is a need for such a solution. Many believe infrared failed because users don't like having to aim devices at each other--a problem Bluetooth eliminates. However, Bluetooth is experiencing its own birthing pains, and it's not clear the technology will succeed. There is an opportunity for another technology, such as a scaled down OFDM solution--to jump in.

Bluetooth has taken longer than planned. The first products are nowhere near the $5 cost promised, selling instead for $100-$200. Plus, many of the first Bluetooth products take up precious add-on card slots. What's needed is a technology so inexpensive that manufacturers integrate it with all of their handheld device designs. Ideally, it would also be compatible with wireless LANs, allowing the devices to work in multiple environments.

Atheros' "5-UP" proposal is intriguing, providing both a path to higher speeds and a very scalable system. Atheros has developed CMOS-based, 5 GHz radios that meet and exceed IEEE 802.11a specifications--running up to 70 Mb/s) in the Unlicensed National Information Infrastructure band (U-NII). The firm reports it is quoting $35 per chipset (with integrated power amplifier) for quantity 100,000. If the technology works as promised, it could replace both IEEE 802.11b and Bluetooth.

Business Opportunities and Illusions

While the market for small/home network equipment will grow rapidly, its impact will extend well beyond telecommunications equipment makers. To put it succinctly, a few billion dollars worth of network adapters, access points, gateways and outlets will profoundly alter markets for home entertainment systems, household appliances, security systems, home automation gizmos and PCs. Thus, the total magnitude of the opportunity will probably be measured in the tens of billions of dollars.

Manufacturers of TVs, VCRs, DVDs, video games and sound systems that leverage home networking will gain market share. As synergy between home entertainment systems, PCs, and the Internet grows, people will start replacing components that are not network-capable with components that are. Expect these devices to work together and expect manufacturers to focus on networking capabilities as a means of differentiating their products and creating a new motivation for brand loyalty.

Gateways will succeed in offices, where people are willing to spend up front for future return. Standalone residential gateways, however, will not do well. They are too expensive and expect consumers to invest in a concept.

Gateways will succeed in offices, where people are willing to spend up front for future return. Standalone residential gateways, however, will not do well. They are too expensive and expect consumers to invest in a concept. Instead, expect residential gateway functionality to be gradually integrated with cable modems, DSL modems and set-top boxes.

Structured wiring systems will also not do well in homes. Structured wiring systems are only practical for new construction. But new construction represents a small fraction of homes. Thus, the market for structured wiring systems is a subset of a subset. Like the residential gateway, the structured wiring system is an application enabler, but few consumers buy application enablers.

There will be opportunities for service providers--but not necessarily the ones envisioned by local phone companies. Remote services will spring up to monitor, control, diagnose, and repair various devices and systems. For example, consumers will tap communications applications service providers (CASPs) for more advanced telephone features, such as conference calling. Outside services might negotiate lower electricity rates by scheduling use of power-hungry appliances and systems. Tutors will offer their services over the Web, making use of interactive multimedia.

Small networks will spark the growth of Internet-based streaming media, distributing media to every nook and cranny of the home and office. Internet radio is available, but not yet widely used, because it is bottled up in the PC. New devices have appeared that liberate Internet radio from the PC, and consumers appreciate the ability to choose from a wider selection of programming sources. Plus, Internet radio is digital radio, though it suffers from occasional network congestion.

As the U.S. population ages, interest in home-based medical technology will grow. Small networks will enable mobility throughout the home and yard for patients wearing heart monitors or panic buttons. Patients will also want faster and more direct ways to communicate with healthcare providers and pharmacies.
Ira Brodsky is President and Paul Pauesick is Director of Research for Datacomm Research, Chesterfield, Missouri. To contact the authors, e-mail them at info@datacommresearch.com. 

This article is based on the report, "Digital Dwellings: Technologies & Business Opportunities for Small Networks" published by Datacomm Research Company in Q1, 2001. 

Visit Datacomm Research online at www.datacommresearch.com.