In comparison to reliable plain old telephone service (POTS), IP-based telephony is still a bad idea.
I see that another attempt is being made to drive a stake through the heart of the two-wire POTS, also known as the Public Switched Telephone Network (PTSN). This time it is by Alcatel-Lucent and its PSTN Smart Transform.
Alcatel-Lucent says its PSTN Smart Transform consulting and design services to current POTS providers who want to shift to all voice-over-IP network will reduce migration costs 30–50%, reduce end-user outage time during migration of 20 seconds compared to an average of 20 minutes by competing services, and migrate 99% of their end-users without experiencing issues.
If this plan makes it to my PTSN provider I will probably be one of that 1% who will have "issues," no matter who is doing the conversion because I have just shifted back to a standard two-wire POTS line from my local telephone company. After numerous tries with different companies, I gave up voice-over-IP (VoIP) because it was maddeningly unreliable. And I find nothing in the material on the Alcatel-Lucent website describing Smart Transform that makes me regret that decision. The web page presents a good case to telephone service providers but there is little on what it means to end users.
Limits of IP-based telephony
Based on my experience with Voice over IP (VoIP) systems over the last several years, I don’t have much confidence in the all-IP solution Alcatel-Lucent is proposing. Here’s why:
When you place a "regular" phone call using the Public Switched Telephone Network (PSTN), you use what's called circuit-switched telephony. This system works by setting up a dedicated channel (or circuit) between two points for the duration of the call. These telephony systems are based on copper wires carrying analog voice or digitized versions of it over dedicated circuits from the home to a local central office where it is further compressed and transmitted as a whole to the receiving end.
VoIP, in contrast to PSTN, uses what is called packet-switched telephony. Using this system, the voice information travels to its destination in countless individual network packets across the Internet. This type of communication presents special TCP/IP challenges because the Internet wasn't designed for the kind of real-time and deterministic communication a phone call represents.
While PTSN-based POTS provides only limited features, low bandwidth, and no mobile capabilities, it has something that after my experiences with VoIP I now value more than I used to: dial-tone availability — that is, a live line — 100 percent available, always or as close to that is as humanly possible. That is due in part to a totally battery-backed up network independent of the power grid. Previously in the United States, when AT&T/Bell was a regulated semi-private company, such backup was government mandated. Now, even though they are no longer government regulated, many of the local telecom companies (BabyBells) that were spun off after deregulation still use it on their PTSN systems, especially in rural areas in the Midwest and inner west of the United States.
Aside from the cost of conversion, the only thing that prevents local BabyBells from doing away with PSTN is that users would scream. I live in a rural area in Northern Arizona, where we get varying levels of snow, rain, and thunderstorms during the fall and winter. Invariably, power outages occur, and when my connection to my Internet Service Provider goes down, there goes not only my ordinary Internet access, but my VoIP line as well. Of course I have my cell phone, but only until my battery runs out. I can recharge the cell phone from the cigarette lighter in my truck, but not my computer or my Internet modem.
There is no such end-to-end battery backup for IP, of course, and there is a lot of work yet to be done to overcome the inherent unreliability of voice over IP. PTSN still has something close to the 99.999% (no more than five minutes of a dead line a year) reliability that the original AT&T/Bell system did. None of the VOIP systems I have tried over the years has anything like this. My grade overall for VoIP is not even 90%. Maybe 70 percent at best.
No matter what service I used, there were at least a dozen ways VoIP crapped out on me: dead lines, crossed lines, misdirected calls, varying voice quality dependent on the amount of Internet traffic at the time, and calls lost in mid-conversation. And invariably, when I tried to forward calls automatically to my cell phone, it would end in failure. And I have not been alone: go to Google and enter "VoIP sucks," and you will see thousands of complaints going back years.
I know I am going against the grain in my unqualified support of POTS. In a recent Federal Communications Commission (FCC) Local Telephone Competition report 47% of residential users and 15% of business customers with POTS-based landline service have shifted to voice over IP. Other less reliable estimates I have read place this at 70% and 55%, respectively. However, it’s interesting that in Alcatel-Lucent's promotional material on PSTN Smart Transform, they point out that by 2017 about 66 percent of voice telephone subscribers worldwide will still be using POTS. So maybe I am not alone.
If the traditional POTS and its rock-solid reliability goes away, I am looking at alternatives other than voice over IP. In one of its promotional pieces, Alcatel-Lucent likened POTS to the proverbial tin can and string method of communications. Call me a curmudgeon if you like, but even though tin can and string technology is a purely analog form of transmission with poor audio quality, I bet its reliability would be better than an IP-based telephony system.
For other alternatives I can go back into one of my treasured four volumes of The Boy Mechanic which contain collections of how-to articles published by Popular Mechanics Magazine between 1915 and 1920. There I am sure I can find something like Claude Shannon’s wireless telegraph built to connect to a boyhood friend’s house when he was growing up in the 1920s in Michigan.
/3 
文章评论（0条评论）
登录后参与讨论