VoIP Still Not Ready for Prime Time

Ever since the voice-over-IP hype began several years ago, I’ve been saying it’s nowhere near replacing good old circuit switching. Guess what? After literally tens of billions have been spent on moving it along, it’s still not here. This month’s Publisher’s Letter looks at 10 new reasons VoIP won’t replace circuit switching and where the various industry players stand followed by the 10 hottest VoIP bets that billing executives and their suppliers should be banking on now.

First, the Bad News

So here are my new top 10 reasons for what’s wrong with VoIP, and why it’s just nowhere near ready to replace good old circuit switches.

1. QoS: The VoIP zealots can give conference presentation after presentation, issue press release after press release, and so on, but the fact remains you can’t get PSTN quality with VoIP. Once you packetize voice and send it with other packets, reduce the analog to digital coder rate, suppress silence and all the other efficient tricks, you reduce quality of service (QoS)—period.

2. Reliability: If a company’s LAN goes down you have a problem; when your company’s voice network goes down, you’re dead. For a major carrier, circuit switch disasters happen about once every decade. A major IP Internet-based service failure is a weekly event.

3. Standards: Today there are three VoIP standards (H.323, MGCP and SIP) forming the basis for product development—and, of course, they are essentially incompatible with each other. It’s worse than that. These standards are works in progress: if you’re basing your product on this year’s version of a given standard, it will likely not be compatible with a product based on next year’s version. Bottom line, there are 30 or more VoIP product vendors out there with virtually no meaningful interoperability.

4. Phone prices: The consumer price point for a phone that delivers caller ID and acts as an answering machine is $30 to $40. VoIP phones are 10 times that price. Well, you say, consumers won’t necessarily have to change their phones with VoIP network service. Yes, but how is a consumer going to program all those fancy call-handling features with 12-button touchtone telephones?

5. Power: The Achilles’ heel of the VoIP phone is power. Your phones at home and at your office still work when commercial power goes out. Someone is going to have to spend big bucks for VoIP phones to make the same claim.

6. Complexity: There is a saying, “If your mom can’t use it, it’s probably not very useful,” and it probably applies to VoIP phones. On one hand, it looks like advanced telephony when you walk up to an IP- phone at a trade show and program it via your PalmPilot with an IR interface, or download information about your custom calling features from your personal Web site. On the other hand, if you have to go the IP carrier help desk to figure out how to enter 911 and dial it, you have a consumer problem.

7. Security: The more complex a telecommunications service becomes, the more susceptible to security problems or fraud. Today’s PSTN phones may be dumb, but you have no doubt who is on the other end. You can prove who made calls to whom (remember the Clinton-Lewinsky episode). When you start talking about people writing softswitch applications in their garage, teenagers programming their phones and more, a lot of fraud-meisters and hackers will be licking their chops waiting for VoIP to arrive.

8. Voice over DSL: Most industry publications—except ours, of course—get this technology concept screwed up and imagine VoDSL as voice over IP. First, if you get DSL from the ILECs, your voice is processed by circuit switch. Second, almost all DSL goes through an ATM digital subscriber line access multiplexer (a DSLAM, not an IP multiplexer). Third, voice-over-ATM is much more efficient than VoIP, with no loss of quality. Bottom line, unless you can really add service value to a residential consumer who has DSL from either ILEC or a data CLEC, you aren’t going to get their voice business with local IP service. If you do, you’re not likely to make any money at it.

9. VoIP network management: There are numerous VoIP or softswitch vendors out there, and a few have actually devoted time and resources to working on the network management aspects of an IP network carrying voice. However, very little is being done to standardize network management for multivendor VoIP networks. Europeans (ETSI) are looking at adding VoIP interoperability to TMN; the ITU is looking at adding network management to H.323; and the Internet gang (IETF) are looking at VoIP interoperability from the perspective of enterprise networking. But in my opinion these efforts are not going to lead to an interoperability solution.

10. End-to-end VoIP: The only way VoIP could completely replace circuit switching and achieve good quality is to do it in one big bang. Seeding the PSTN with islands of VoIP networks will not likely lead to a complete conversion to all VoIP. Here’s why. Anytime you complete a VoIP call on an island network (a local VoIP carrier) you have to go through the circuit switch network to another VoIP network island, end-to-end quality goes down the drain. You can’t do a low-bit-rate IP coder to PSTN coder to low-bit-rate coder conversion and have good quality. If both the originating and terminating VoIP have the same coder, at best it will be poor quality. If the IP coders are different at each end, the quality will be unacceptable (namely, zero).

So what does this mean for migrating to all VoIP? You can’t transition from all-PSTN (Figure 1) to all VoIP as in Figure 2, via Figure 3 to 4 to 5 network. It’s either VoIP end-to-end, or you won’t see the PSTN go away.

VoIP Billing Opportunities

So if VoIP isn’t going to replace the PSTN anytime soon, where are the VoIP service opportunities and challenges? Here are my top 10.

1. Prepaid calling cards: Clearly, offering prepaid calling card service via a managed IP network is a proven market and fills a major consumer need. Visitors to the United States who don’t have calling or credit cards, or those who don’t want to pay up to $10 a minute calling home from a hotel phone, will accept a lower quality of VoIP in order to save substantial money and be able access many other VoIP value applications. But the biggest prepaid market is outside North America. In Europe most every call, including local, is metered; and phone bills for equivalent use can be 10 times what you pay in North America. In Africa, where outgoing bandwidth is limited, compressed prepaid VoIP makes sense. So as long as you avoid calling someone who has a VoIP phone (Figure 5 above), you will have a market winner.

2. Managed IP: Every major IXC, particularly the international carriers, must have a managed IP offering (intranet, extranet or virtual network services). Frame relay and ATM services can’t deliver the switched connectivity and bandwidth on demand of managed IP. Since the big guys have to offer managed IP anyway, corporate network customers ask for VoIP even if they have not committed to using such a service—and if you don’t offer it, some startup will. VoIP via managed IP is a reality. So enterprise or private network VoIP is a reality, even if it’s bundled as “free” to paying data networking customers.

3. Browser on the lips: Voice access Internet Web sites will take the Internet to the next level. With VoIP, voice is just another type of data and eventually can be interfaced via XML to Web sites. You can’t do this with circuit switching alone; you need VoIP.

4. Voice ASPs: When you speak of application services provider (ASP) applications, where you couple the above browser on the lips with outsourced IT or ASP products, you add a new dimension to the ASP opportunity.

5. Call handling vs. PSTN replacement: While the $500 IP phone connected to a LAN is not going to make the “black phone” obsolete, there clearly are people who love their PalmPilots and would routinely reprogram their IP phone or download customized instructions (or data) into their IP phone. In this case and for this market, it’s not about VoIP as the end, but as a means to be more innovative with call-handling features. Yes, there is a market for things like having your phone ring at the office and at the same time ring at home. The PSTN can’t do this. VoIP can. But this service is not for all, and you don’t have to replace the PSTN to support it.

6. WAP and integrated messaging: VoIP introduces many new innovations coupled with other developments, particularly in wireless. The Wireless Application Protocol (WAP) allows a wireless phone with a microbrowser to communicate with a server. VoIP, or voice coded in IP, allows for voice mail server-to-server communication. Since this is not in real time, quality of service is not an issue (voice goes over TCP, not real-time UDP as in VoIP networking). So what are the innovations? Many! You can see on a screen what voice mail is in your box without listening to the message, leave and send voice mail without having to enter addresses, and send voice messages in seconds thanks to compression, instead of in minutes uncompressed.

7. Wireless softswitch: The world needs a local (class 5) or long-distance (class 4) circuit switch replacement like it needs a hole in the head. Most of the advantages are lost with VoIP if you get into VoIP-to-PSTN-to-VoIP networking (Figure 5). However, a VoIP softswitch and managed IP network for a wireless carrier and the industry makes sense. Note, a softswitch centralizes the provisioning, service routing, call control and signaling software, leaving behind a simple, low-cost VoIP or media gateway. As discussed in last month’s Publisher’s Letter, a wireless softswitch makes sense because the wireless carrier is going to be serving a mix of 1G, 2G, 3G, data-only wireless Internet access devices and more, and wouldn’t want to deploy a switch for each new terminal. The wireless softswitch could have all information converted to IP at the base station via a cheap media gateway. IP packets, under the control of the softswitch, are routed off to the PSTN if it’s voice, the Internet if it’s data, or a CODEC converter where it’s off to another wireless network, to ensure no degradation in service quality (see Figure 5). By the way, if you are looking for the road to an all-VoIP network (Figure 2), this is it—except you are talking about wireless VoIP coexisting with the PSTN, and in some countries phasing out the PSTN. Note that this scenario will work with wireless but not local VoIP network replacement, because at least today you can distinguish between wireless and wireline phones by the dialed telephone number.

8. Cable softswitch: Cable companies plan to make big bucks delivering a vast array of new services by combining entertainment video, Internet access and, of course, VoIP. The synergy of video and VoIP, as well as Internet access and VoIP, will enable new services. Just as in wireless, a cheap media gateway at the home and the softswitch at the cable head end makes sense. It’s highly likely that VoIP will be a success for the second, third, and fourth phone lines to the home, or the so-called “teen-line.” As for the first line or “lifeline” that gives dial tone for 911 when the power is out, or PSTN-quality voice for work at home, the jury is still out due to the poor economics from UPS or network power, no use of cheap coders, no silence suppression, and no statistical multiplexing with Internet access.

9. Digital signature: One of the hottest new services, particularly in the wireless world, is instant messaging. But just like any data message, you run up against the problem of managing encryption keys or something that the industry calls the public key infrastructure (PKI). Today’s wireless phones have encryption keys, but they lack that second requirement of authentication (yes, you are who you say you are) and non-repudiation (yes, you did electronically sign this check). The ultimate digital signature for a wireless phone could very well be your voice carried over IP as a prelude to an instant messaging session or making a purchase.

10. Free voice: The ultimate opportunity for VoIP could very well be free VoIP. This has been bubbling around since the VoIP hype started, and there are companies providing it via PC to phone, or advertisement-supported. It certainly solves the QoS problem—you only get what you pay for, which is zero. But since you have to make money somewhere with free VoIP (ad revenue, commissions from transactions, and more), the above VoIP opportunities do create IP billing challenges.

If you are looking for billing solutions in general or IP billing in particular, don’t forget to mark June 26–29 on your calendar. Plan to attend TeleStrategies Billing 2000 in Washington, D.C., or check out our Web site (www.telestrategies.com), if you feel you need a seminar in telecommunication technologies, IP or billing.