The Internet disappeared at midnight on Dec. 1 for 850,000 subscribers to AT&T Broadband’s cable modem service. That’s when AT&T’s insolvent broadband partner, ExciteAtHome, officially pulled the plug, cutting off Internet access for AT&T customers as a cost-cutting measure. In the following days, AT&T scrambled to restore Internet service to its customers to try to prevent them from switching in disgust to another Internet service provider.

Never has the messiness and stupidity of the broadband industry been quite so apparent. Not that ExciteAtHome’s demise is a surprise — the company had been teetering on the edge of insolvency for months, and the main question wasn’t whether it would go out of business, but when. However, the incident does raise some questions. Why is it that broadband companies like ExciteAtHome, Rhythms NetConnections, and NorthPoint Communications (two other recent flameouts) can’t seem to stay in business? This occurred despite the investment of tens of millions of dollars, a genuine demand for bandwidth among consumers, and an array of slow-moving, sclerotic competitors (the cable TV and telephone companies) who seem to have only heard about the Internet sometime last year.

The Defogger is here to tell you not to panic. There’s nothing fundamentally wrong with broadband technology, and as sure as I write this, it will ultimately become as common as television (better, as common as electricity). Right now, according to a recent study by Nielsen//NetRatings, the number of consumers surfing the Internet at high-bandwidth speeds is greater than ever. Some 20 percent of the Net’s 106 million active users have a broadband connection at home — a record number. What’s more, Nielsen estimates that more than half of Internet users at work have high-bandwidth connections. For all but the smallest of companies, a T-1 or T-3 connection has become a nonnegotiable cost of doing business, like paying the telephone bill. (And incidentally, that’s also why so many people choose to do their online shopping from work: faster connections.)

No, the real problem is not with broadband technology, but with the inevitably cyclical nature of the business. A lot of optimism about broadband’s future — as well as incessant hype — led to a period of massive infrastructure buildup and investment, which far outstripped demand and created way too much excess capacity. Only 3 percent of the high-bandwidth fiber-optic cable currently laid in the United States is in use. The other 97 percent lies dormant. As a comparison, think about auto manufacturers producing 100 cars and selling only 3. With broadband companies sitting on that much idle inventory, there’s no way they can recoup their costs, and they’re predictably dying off as a result.

A second problem is that the government’s attempts to deregulate the telecom industry created a big incentive for companies to build backbone networks (which carry data over long distances), while creating a disincentive to build last-mile networks that would carry data to consumers’ doorsteps. Deregulation attempted to force telephone companies to share their last-mile connections, a decades-old monopoly in most parts of the country. As a result, most broadband companies figured they could simply piggyback on the telephone companies’ copper-wire networks, or partner with cable companies and use their cable TV lines (as ExciteAtHome did), so they concentrated on building backbones instead.

The problem is that telecoms and cable companies refused to give up their most valuable assets — those last-mile monopolies that they’d spent decades and billions of dollars building — even though they had been ordered to do so by the federal government. Meanwhile, cable companies like AT&T realized that they could build or buy their own backbone networks fairly cheaply, so they didn’t need partners like ExciteAtHome anymore. Result: The telecoms and cable companies are still sitting on the only thing worth having, and everybody else is left holding a bunch of worthless backbone capacity.

The third problem is that cable companies and telephone companies can’t provide good customer service to save their lives. Who can blame them? After decades of near-monopoly status, collecting $20 to $40 a month from millions of customers at a time, it’s easy to get fat and lazy. There’s just no incentive to provide good customer service — which is exactly what’s needed to persuade residential customers to spend their time and money on a balky new technology like DSL or cable modem service.

Is there any quick way out of this morass? Probably not. DSL providers will likely see a temporary lift in business as cable modem subscribers, disgusted with AT&T and ExciteAtHome, switch over to a completely different technology. But DSL still isn’t readily available in rural areas (or in many suburbs), so cable will likely keep its customers there. Companies that provide broadband Internet via satellite haven’t managed to get much of a foothold in the market yet. And with high-bandwidth connections waiting at the office, many people just don’t feel compelled to get broadband at home yet.

During the next few years, though, the number of residential broadband subscribers will keep rising, slowly but steadily. The current glut of capacity will be absorbed, the few broadband companies that survive will start to make money, and after a few more boom-and-bust cycles, broadband data connections will be nearly ubiquitous, and they will indeed form the underpinning of the next Internet revolution. In the process, however, broadband data providers will be relegated to the status of commodity suppliers or utilities — an unglamorous, cutthroat kind of business, but an immensely profitable one for those who can corner the market.

Link: UnexcitedAtHome

Link broken? Try the Wayback Machine.


Are You Broadcasting Secrets Over the Airwaves?

Wireless technology has developed ahead of most companies’ ability to keep their networks safe.

In principle, wireless networks sound like a great idea: no cabling to pull through the walls and ceilings, no nasty wires tangling themselves up behind your desk, no need for laptop or handheld users to remain rooted to a single location. They’re a great idea, that is, until some facially pierced teenager in the coffee shop across the street from your office taps into your wireless network, downloads a few files, defaces your intranet homepage, and plants a virus or two just for fun.

The reality of wireless networks is that, in most cases, their convenience far exceeds the level of real security they offer. Network administrators with years of experience safeguarding traditional wired setups often fail to plug obvious security gaps in wireless networks simply because they lack the requisite tools and knowledge. (It’s not their fault: Either the technology is too new, or wireless technology vendors haven’t built security safeguards into their products so that they’re secure right out of the box.)

In many cities you can drive down the street with a laptop equipped with a $100 wireless network card and pick up the signals beamed out by various offices as you drive past. In a recent experiment, techie website ExtremeTech set up a wireless laptop on the roof of a building in Manhattan and discovered it could access 61 wireless networks, of which 48 were completely unsecured. In Silicon Valley, the site found 100 networks — many accessible from within a car zooming down Highway 101 — of which 66 lacked any kind of security.

This sort of hacking is made easy by the way most wireless local area networks are set up, using the 802.11b transmission standard (also known as “WiFi”). With WiFi networks, a techie links wireless ports off the wired system using a bridge device called a wireless access point. That access point communicates on regular old radio waves to laptops and handheld computers.

The problem is that WiFi security measures are weak — and in some cases, nonexistent. Wireless access points are often installed behind the company’s firewall, so if you can get to that point, you’re essentially in the system. Elias Ladopoulos, the founder and chief strategy officer of security consultancy Digital Frameworks (and a former hacker), has a vivid analogy: “You think of medieval times, when China was building up its wall to stop all these raiding tribes from attacking their cities. Well, then the tribes discovered you could fly balloons over those walls and attack the cities. That’s what’s happening on our networks. We’ve spent all these years building firewalls, and what happens is someone plugs a $250 wireless device into your network and bypasses all that security.”

WiFi does have a built-in security system — known as “Wired Equivalent Privacy” (WEP) — but network adminstrators have to take steps to install it, and even if they do, it’s easy to get around. Readily available software with names such as AirSnort and WEPCrack can get a hacker through WEP’s defenses in a few hours or less.

There are more effective solutions to address this problem, though, and they aren’t all that complicated. First, your techies need to make sure the wireless component remains separate from the rest of your network. That way, if hackers do get in, they won’t gain automatic access to every piece of digital information in the company’s files. Second, if you’re concerned about the security of wireless data transmissions, use virtual private networking (VPN) technology to create a secure connection over the wireless link. VPN systems work by encrypting the information sent over a network so that it can’t be deciphered, even if someone manages to intercept it (see “Your Own Private Internet”, Defogger, September 2000). Digital Frameworks uses VPN technology to protect wireless networks, and VPN vendor Check Point also offers versions of its products aimed at securing WiFi systems.

Finally, make sure that your IT staff is up to speed on wireless security technology. ExtremeTech is a good place for hard-core, nuts-and-bolts security information. Check out Business 2.0’s Web Guide for a less technologically intense overview of the subject. And send your IT folks off for some training on wireless security if you can. The bottom line is that if you’re going to let employees into your network through wireless devices, you should take steps to make sure everyone else is kept out.

Link: Are You Broadcasting Secrets Over the Airwaves?

Link broken? Try the Wayback Machine.

Are You Broadcasting Secrets Over the Airwaves?

Table Is Set for Web Telephony

Switching to a voice-over-IP phone system can cut your interoffice long-distance bills to zero and reduce administrative headaches. So what are you waiting for?

For years now, the promised convergence of voice networks and the Internet has failed to materialize. It’s a bit like watching Bullwinkle Moose try to impress his buddy Rocket J. Squirrel: “Hey, Rocky, watch me pull a rabbit out of my hat!,” the providers of Internet telephony technologies promise, year after year. “Aww, that trick never works,” a jaded public sighs. “This time for sure,” the technologists gamely insist.

To no avail: After years of promises, the Internet still hasn’t replaced the telephone, or driven the cost of consumer long-distance calling down to zero — even though companies like Net2Phone have offered free telephone calls via Internet-connected PCs since 1996. Isn’t it time we put this whole idea to rest?

Not so fast. Voice-over-Internet Protocol (VoIP) technologies, which transmit conversations over IP-based networks, are on the verge of transforming corporate telephone systems. Now that companies have adequate bandwidth to handle VoIP, the technology’s voice quality rivals that of the Baby Bells. For companies with multiple locations, routing calls between them via the Internet can save a bundle on long-distance charges. What’s more, turning voice conversations into digital data makes it easier to integrate a telephone system with enterprise software applications. This puts customer relationship management (CRM) and sophisticated call center systems within closer reach, and makes possible new features — such as unified messaging systems that deliver e-mail, faxes, and voice-mail all on a single webpage.

These advantages, among others, have pushed sales of corporate VoIP software and equipment to $829 million this year, according to Phoenix-based consulting firm Synergy Research Group. In the coming year, that figure is projected to swell to more than $1 billion, with vendors such as Alcatel (ALA), Cisco (CSCO), Lucent (LU), Nortel (NT), and 3Com (COMS) leading the way. VoIP converts include Dow Chemical (DOW) and Merrill Lynch (MER), as well as many smaller companies.

For Buca (BUCA), operator of the Buca di Beppo chain of kitschy Italian restaurants, VoIP is part of a calculated expansion strategy. The company has 5,500 employees in 68 restaurants, and has been adding about 15 new outlets each year. Prior to installing an IP-based telephone system, each Buca di Beppo had 12 ordinary phone lines (at an average monthly cost of $50 per line) and called long-distance for dial-up access to database and e-mail systems in Buca’s Minneapolis headquarters. Long-distance charges alone were “a huge, huge number,” says John Motschenbacher, Buca VP for finance and purchasing.

In October 2000 the company began installing a new Nortel-based VoIP system. Now each restaurant has an IP-based PBX system and a T-1 line (used for voice and data) that connects to Buca headquarters via a private data network. Buca’s five-person IT staff oversees the entire network, using simple Web-based control panels. Motschenbacher estimates that each restaurant saves $300 to $400 on its monthly telephone costs.

But cost savings are only part of the story — the VoIP system has also boosted the company’s revenues by allowing restaurants to take reservations at any time of day. With the Nortel system, calls made to a restaurant before it opens at 3 p.m. are automatically routed over the T-1 to an eight-person staff at Buca headquarters. The call center books 400 to 800 reservations per day, generating an estimated $1.5 million in additional annual revenues. Not bad for a system that costs about $10,000 per restaurant.

To understand why Internet telephony is taking off now — after years of false starts — it helps to take a look at the technology under the hood. Consumer-oriented Internet telephony services bypass long-distance charges by sending voice data over the public Net, which is essentially free (you pay only for the local connection to your ISP). But this makes for crummy call quality — if one packet of data takes a fraction of a second longer to arrive, your conversation gets interrupted with pauses and glitches. Using the public Internet to place telephone calls is like hauling muddy water from the river, bucket by bucket, just to save a few cents, when clear water is already being piped directly into your home. “You get what you pay for,” says Anil Uberoi, VP at network monitoring provider XACCT.

Calls between your company’s various offices, however, go over the company’s local area or private IP network. And there it’s possible to control the amount of bandwidth available to voice data — giving it a higher priority on the network than, say, e-mail messages. Once you’ve solved the call-quality problem, VoIP offers significant advantages over the old telephone system — as Buca found — and costs the same as a traditional PBX, about $1,500 to $2,000 per phone line, according to research company Meta Group, based in Stamford, Conn.

However, a few obstacles remain between you and VoIP utopia, the biggest of which is the lack of interoperability between equipment from different vendors. Disappointingly, VoIP vendors have chosen to implement standards inconsistently, meaning that equipment bought from one company won’t always work with that of another. To steer clear of IT disaster, you’ll want to buy all of your gear from one source — and check the standard it’s based on. One potential bright spot in this mess is Session Initiation Protocol (SIP), an emerging standard for routing of calls over IP networks. Most VoIP vendors have already added support for SIP or plan to do so soon, which should smooth interoperability in the coming year.

Until this issue is resolved, I wouldn’t advise ripping out your existing phone system and replacing it wholesale, unless you have a manageable project with a clear payoff, as Buca did. Instead start with pilot projects in branch offices, and then extend VoIP throughout the company as the standards catch up with the market. (Even at VoIP vendors like Cisco and Nortel, the phone systems aren’t entirely IP-based yet — despite years of migration.)

Buca’s move to a big VoIP network was a gamble — especially for a company in the technologically conservative restaurant industry. But the payoff came so quickly that now the whole thing seems like a no-brainer to Motschenbacher. “I’m really surprised other companies haven’t grabbed onto VoIP faster,” he says. If Buca’s experience is any indication, it won’t be long before they do.

Link: Table Is Set for Web Telephony

Link broken? Try the Wayback Machine.

Table Is Set for Web Telephony