Slashdot Mirror
VoIP, WiFi and the Future of Traditional Telecom
PetiePooo writes "Those of us in the telecom industry have been watching it wither and die in the past few years. Here's why. The Register has an article about the future of mobile communications using VoIP on WiFi. From the article: "... voice over IP would gradually come to be a prime driver of mobile Internet." VoIP has been considered by many for a while now to be the future of traditional telephony. Combining VoIP and WiFi makes a compelling argument for the convergence of voice and data services over a single platform. Here's a previous slashdot discussion on industry's efforts to make this happen."
Previously telecom systems were prohibitively expensive to set up unless you were government or had big backers. Today the playing field is leveled, because both the big corps and the private individual to both establish communications over the last mile.
It will be interesting to see what tiny telcos which are miraculously on the same standard and able to communicate seamlessly will be able to do.
As someone who has, on regular occasion, the responsibility of supporting Wireless access technologies for Companies, I can state categorically that the current standards are NOT up to scratch as yet.
What do I mean? Well, for a start I have lost track of the number of times individual machines on Wireless simply 'drop out' of communication, leading to perception on the part of our customers that this isn't a reliable , responsible technology.
We have seen, in implementing Wireless, a whole host of different issues - in ideal circumstances Wireless access works well, is fast enough to be used for most internal office purposes and so on.
The problem with Wireless in any form is that it is not as tollerant of non-ideal conditions. Adverse weather conditions (especially during the summer, when static build up knocks out entire Wireless networks on a regular basis), passing vehicles, other communication devices (especially mobile phones, which regardless of advancements in tech will continue to operate alongside any upgraded solution for some considerable time) and simple things like the type of clothing work by the person using the computer, have been known to knock a machine out of a WAN.
Solutions of phone technology over existing Cat5e UTP cable networking, such as that provided by Nortel Networks work well, with integration into existing office apps, but Wireless for Data is still, in the field, an unreliable technology. Wireless for VoIP still runs the issue of packet lossage (which on any Wireless solution i have ever seen runs at upwards of 25%), which is far more serious than equivalent signal loss for conventional mobile telecom solutions.
An infinite number of monkeys will eventually come up with the complete works of
I have been involved in the telecom industry for over 8 years, mostly in the call center arena, merging voice and data, and the thing I have yet to see is a good implimentation of VoIP to the desktop.
An earlier poster made the comment that "a number that follows you anywhere." This would not be a function of the pipe that delivers the call to you. WiFi as it stands now is not a good protocol for VoIP. In general IPv4 is not a good protocol for VoIP, and there Internet is VERY MUCH not a protocol for VoIP. It all has to do with the bandwidth that voice takes, and the unusually high quality that us humans need to have to feel the service is good.
If you want a good VoIP solution, you have to run a seperate pipe to the desktop, on seperate routers to ensure decent bandwidth. You have to use propriety IPv4 QOS and you have to sratch you head a bit when it doesn't work right. Also, you Data folks tend not to understand Voice applications and you have a hard time getting pratical support from your WAN/LAN administrators.
We have heard a lot about carriers switching over VoIP. Well, what they are mostly doing, which is pratical these days, is using it for intra-Central Office traffic, which is fine and dandy when the only thing going over your Pipes is Voice. You can guaruntee the quality, know what the bandwidth usage is, etc... but this isn't much different than ATM except that it has a cool name. A lot of us forget that almost every networking technology (ATM, T1, Fiber, etc) was orginally a voice pipe before it was used on the data side.
GSM, CDMA, etc are GOOD wireless protocols that show what adaptive bitrate protocols can do, WiFi would be abosolutly horrible in its current incarnation. It is a fully cooperative very limited bandwidth protocol. Great for our data bursts, but very bad for the sustained traffic of voice. It has a VERY large overhead, plus you had the overhead of IP and you are at a pratical 3-4Mbs which then has to content with the guy down the hall dragging porn files off a remote server or someone playing Warcraft III with 20 other players. Now even 802.11g/a would be a tough bandwidth to deal with. I don't know the specifications in detail, but I doubt they have any standard QOS features.
Anyways, that is my 2-3 cents...
D.O.U.O.S.V.A.V.V.M.
Back in the old days when Internet was embryonic, most of the traffic was telephony. Carriers were operating networks designed for voice and carved into it some channels for the little data applications that were required then.
With the growth of the Internet and entreprises IP networks, this model broke. Carriers had to implement and operate telephony and data networks in parallel.
But the growth of IP traffic did not stop there. Nowadays, most of the traffic is data. Voice requires little bandwidth in comparison and its share of the overall requirements will be less and less every year as the data continues to grow fast. Carriers that can find a way to carve bandwidth for telephony on their data networks can just shut down their their telephony backbone infrastructures with huge cost savings. Then they will be able to pass some of these savings as price reductions to the customer and undercut any competition that doesn't make the same move.
This is basically transparent to the general public because this is purely a backbone thing. Your home phone line will still use the ame old technology. VoIP to the phone technology exist but it is technically less mature and does not have such a clear cut business case.
The future of WiFi with carriers is a different story. You need to understand the state of the mobile Internet to understand why.
Several carriers have invested billions in spectrum licenses to operate 2.5G and 3G mobile networks without having a clear profitable business plan. They did so because the goverments were holding auctions for the spectrum and if the carriers let these go without bidding they would not be able to get into the wireless market at all. These companies are now stuck with heavy debts and are desperate for a killer application to attract customers.
One candidate is the mobile Internet, that is the ability to connect to the Internet anywhere, anytime just like cellular phones let you call a friend anywhere anytime. But there still is a problem of how to make a workable business model out of this idea. One issue is lack of speed. GSM delivers n actual troughput comparable to a 28.8k modem. CDMA-1X is just slightly faster than a 56k modem. In these days of ADSL speed, this performance just does not cut for many people. Carriers are now thinking of installing WiFi hot spots and bundling their 2.5G and WiFi in one single mobile Internet offering. You get WiFi speed in places where people hang out most of the time and 2.5G speed everywhere else.
I just happen to have tested the idea of mobile Internet yesterday. I had a wine tasting party with friends at a restaurant. We wanted to know the grape variety that was used for the wines. I took my PDA and connected to the net over my GSM phone to get the answer. That was cool. But there are many issues to resolve before this kind of things fly in the general public.
The most important issue is pricing. In Canada Fido offers unlimited usage for 50$ CDN. This is right although a bit pricey. Some other carrier offer 500 kbytes for 5$ and each additional kbytes at 5 cents. Considering I consumed 500 Kbytes in just one wine tasting party, these expensive usage pricing models will just stiffle adoption.
Another issue is the mobile hardware. I tested a tablet PC with a built-in WiFi card and CDMA-1X add-on card. It worked cool for work, but the device is too bulky to be used in all situations like a phone could be. If you have a cell phone, it is a shame that the add-on 1X card also require a separate phone number with all the associated fees.
A more mobile solution would be to have a Bluetooth enabled phone and PDA. This is what I used in the wine tasting party. It was cool to be able to surf at the table without wires in the way and without having to make room on the table for a laptop. But the small screen size doesn't work well with many sites.
A