Louis, you bring up some interesting and very good points:
>>Nothing I've read so far has been a compelling argument for investing in any of these companies. If we're going to find the next "Cisco", you have to find a company that manufactures a critical component generic to all technologies that arise out of this new market trend. I've haven't heard anyone argue what that might be.<<
Perhaps "different" critical components can satisfy the same generic requirements, hence permitting diversity and choices. But I get your point.
You may have touched on something else here that I was waiting for someone to bring up. You probably weren't alluding to this, but I'll offer it up anyway, at the risk of boring some:
What's missing, IMO, is a common glue that will enable interoperability, not only among different VoIP vendors' wares, but with the existing PSTN model, as well.
The interoperability has to go beyond that which will permit same-box-to-same-box, and local-calling-area- to-local-calling-area calls. The latter refers to the end-office local calls which most of the vendors mentioned here are able to complete today.
Granted, calls may now be completed beyond the end office, but they are not leveraging the lower costs of internet transport, since they must traverse the PSTN at regular calling rates, and must be billed (reconciled) on a case by case basis, depending on the destinations.
I'm suggesting that there must be ties to the signalling system #7 (SS7) network, and to the Advanced Information Networking (AIN) databases that have been in use since the early Eighties, since the inception of modern 800 calling, and the use of service control points to oversee enhanced connections. It's conceivable to me that these "ties" may only be a transitional requirement, as the paradigm shifts, but they will nonetheless be a requirement for at least the next three to six years, at least, IMO.
The SS7 network is perhaps the best kept secret in industry, mostly because it is only accessed by carriers and very large users and enhanced service providers. And the IN/AIN databases, accessed by the same parties as the SS7 Network, are similar to DNS and directory servers in the Internet space, but significantly different in structure. They are responsible for line and subscriber look ups, policy-based routing decisions, billing, to name a few of their functions. They are also required (at least links to them are) for calling card account billings and settlements. To the best of my knowledge, only a few VoIP vendors have even remotely addressed these capabilities so far, and since I haven't been overly impressed with their progress to date, I'll defer mentioning who they are. I am quite certain, though, that this is an issue that is on the back burners at many a VoIP startup and ITSP alike. In fact,it is the substance for which ITXC (Tom Evslin's company) was formed.
While discussing these issues, it would be interesting to place them alongside the pending draft of the ITU/IMTC VoIP standards, and how the standards may potentially be used to satisfy the shortcomings that currently exist, and there are many. This issue does not receive, IMO, the attention that it should here, or elsewhere.
Another issue that is creeping up on us, but has received little attention so far due to the high levels of interest and excitement in the Internet protocol space, is that of ATM. Voice Telephony over ATM (VTOA) in the near future will incorporate the abilities to co-exist with VoIP, or may flourish on its own merits on dedicated backbones, optionally. In advance of the standards being ratified by te Forum, there are some non-standard adaptations of this in some of Nortel's Core and Edge products, as well as by others, and they are highly efficient for those carriers and SPs who use them. It's a little early to predict what the impact of this technology will be, once the standards are in place and boxes are shipping, but it does loom to be highly competitive, potentially, in terms of both bandwidth utilization and routing via both native ATM and multiprotocol (IP) techniques still in the RFC stages of the IETF.
Do you or anyone else here have any thoughts along these lines concerning ATM, or regarding the issues that I brought up above, i.e., SS7/AIN-like interoperability features in VoIP products?
Regards, Frank |
