I retired from personal blogging in July 2008.
But you can find me over at http://blog.xero.com.
Making progress behind the scenes on Wellington 2.0.
As a point of clarification (because I’ve been thinking VOIP for a few days) VOIP is not primarily about saving money (though it helps), its about connecting to markets.
For example www.packet8.net allows any one in the world to present themselves with a US phone number. That means that when you dial the US you can lessen the ‘where are you from?’ question and present a local number for sales and support queries.
Why am I angry enough to push this Wellington 2.0 concept? Because we could not use the Packet8 service reliably - so we had to keep dropping back to a standard phone service and present as being from NZ with +64 n nnn nnnn. It just makes it harder.
The international part of this strategy is to provide a clean Quality of Service (QoS) connections that we can run low-res data communications services, first consumer VOIP like Skype, Vonage and Packet 8, then desktop video and eventually hi-res video conferencing - like Access Grids.
Because VOIP is so inexpensive compared to traditional carrier based call plans - it is seen as a threat to their revenue. No one seems to be able to confirm or deny but it appears that VOIP data traffic is actively ‘managed’ down to a low QoS such that it becomes unusable.
But it’s not about the money. If the carriers offered VOIP services where we could have a US, UK and Australian phone numbers and CD quality voice audio (which is very possible) we would probably pay a reasonable amount for the service.
The current situation makes the carriers appear anti-business and spoilers. In this time of carrier’s claiming to ‘get that consumers hate us’ why not spin it around with some positive management and allow us to connect to world. Use the technology they have at their disposal to make us more competitive and connect us to the world - rather than try to constrain us with old and tired business models.
blog
CD quality voice is hard because the telephone network doesn’t support it. You won’t get it if the call has to go through the telephone network at any point.
QoS … that sounds really odd. You’re not trying to run an office’s internet over DSL, are you? There are plenty of ISPs (even Wellington-based ones) who will happily sell you committed bandwidth, which should mean you don’t need to compete with others to make your VoIP work.
??? have you tried to do a VOIP call overseas?
An ISP cannot commit to QoS from here to the US or anywhere because they do not control the entire link. They pass to the international connected networks which I assume is where the problem lies.
The premise that VOIP is about connecting to markets is accepted. (Example: I got a Wellington DDI for WIC as it’s useful us to be able to offer a local number to business contacts there.)
Connecting to overseas markets by the same means is trickier, for the reasons that you point out.
The solution that I would propose would be to commit a reasonable sized pipe (say 10Mbps) direct in to a hosting centre connecting to a VOIP gateway. (My specific solution would be an Asterisk server onto ISDN PRIx4). That’s not undoable with current service provision.
Location of this gateway? US, Los Angeles or Hong Kong. (We looked at feasibility of putting gateway into one of our client’s offices in HK). Globally, no difference between the two locations as US DDIs could be added to HK exchanges.
Cost?
International bandwidth $800/Mbps monthly, so $8K for the pipe, another $1K monthly for national to pick up all the NZ IXes. 1 VOIP call is around 80Kbps, so that pipe does 120 or so simulataneous calls (in and outbound). Trunking helps that number.
That gives you fantastic QoS end to end, digital voice quality all the way and low-cost international since HK rates are the cheapest anywhere. Get NZTE to commit to building exchanges in LA, HK, Singapore and UK and you cover the globe for around $50K capital and $100K operating. Not huge numbers by any means.
There are people able to build this very quickly. The two spurs to either get it done are demand and funding, pick any one. Let’s do it.
Great info. That’s exactly the scenario!
Why buy the pipe?
Would it not be cheaper/more reliable to locate gateways (Asterisk nodes) in the markets you care about that take incoming local calls and switch them back out as voice calls using the local PSTN to a +64 number?
Your cost then would be related to the per minute fees which (I don’t know the math) could more easily be renegotiated as the market dictates.
Lower fixed costs anyway–just the local PRI’s and electricity.
In any case, you could set that up as phase one and add IP voice routing later.
This is really a side case from the Wellington 2.0 topic which has a lot of merit unrelated to voice calling.
Rod’s point was regarding QoS which can’t be assured routing VOIP over “public” Internet; you can get closer to high-quality if you pay for an assured pipe. Locating the gateways in various locations means buying local access at each of those which bumps up the price (hosting plus PRIs) but the architecture is pretty much the same regardless.
I know this is off-topic from the tech comments and I totally agree with your points. But to those who need a solution. Get the Packet8 number anyway and put it on your card. Dial out from NZ using 0197 to mask your number. If you have to leave a number just leave your US number, but if you do reach someone they won’t know from Caller ID that you are ringing from NZ or Timbuktu and you have the call quality levels of normal lines. It won’t solve your cost problem as well as VOIP but you can play the “we are in your market” game with absolute confidence.
I might have to semi disagree on this. It is about connecting with markets but its all about the money in my opinion lol. I mean VOIP gives a lot more revenue for business owners and more high speed internet providers are jumping on the bandwagon. http://ds3providers.com