<< I might mention here that the difference between the 800 phone line reception and the webcast is the difference between ATM and IP protocol transmission. >>
fumble, the same can be done with a frame relay sla. Our company uses a frame relay T1 betweem sites; Israel, US, England; and uses spare bandwidth on the T1 for voice circuits, with no discernible difference in voice quality than circuit switched. Much cheaper than ATM at this point, which providers are not provisioning (much) down to the T1 level, and certainly not down to the 64K or 56K level. I think the starting point for managed ATM services from most carriers is around 10Mb/s.
However, moving up from there they are provisioning in much better granularity than going right from T1 to T3. For T1 and below, frame relay will continue to dominate for the forseeable future.
Gary
That could very well be true for straight deals through carriers today. The carriers are and will be careful not to cannibalize their existing gravy train - to the extent possible.
Frame Relay is an older technology, but with some of the attributes of ATM. ATM offers true QoS and easier management (from the carrier side - even delegating sub pipe management to sophisticated customers), so there is a trend to slip ATM underneath FR and expose it as customers request. ATM/FR relative pricing depends on demand, supply, and competition which varies widely over the world.
There is a hunger in corporate circles for cheaper T1 lines. This has been noted on this thread with Quest as an example. They are offering T1 at 1/3 of competitive rate and are choking on orders and provisioning (They should have gone for 2/3 of going rate and then worked down from there, but as a new guy on the block, they needed to put the stake in the sand somewhere).
As competition mounts around the world (and huge P/E IPOs are financing them), the carriers will begin to mind their costs and ATM will take an increasing share of the traffic, because of the provisioning and traffic management advantages.
Delivery of 1.5Mb ADSL to end-user businesses is even more competitive ($90/mo in Chicago) than Quest's T1 offering. Couple the ADSL to an ATM 'cloud' supporting QoS at a reasonable price (more than current Internet pricing, but below Quest) and you will have a nice win-win situation for any carrier or carrier competitor who has the vision to get into this business. Perhaps the Newbridge 50G+ will fuel this arms race.
On the issue of provisioning ATM at slower bandwidths on the edge of the network, take a look at the paper mentioned previously Message 11683009
-----
Service Provisioning in an ATM-over-ADSL Access Network
Reuven Cohen, Technion and ECI Telecom
Abstract
Asymmetric digital subscriber loop technology is a new platform for delivering broadband services to homes and small businesses, thus bringing the information highway to the mass market. Successful deployment of an ADSL-based access network is not only a matter of delivering faster rates, but also offering solutions that are cost-effective and easy to use and manage. As deregulation and competition spread globally, an ADSL-based access network should also enable end users to choose their service providers dynamically, and to switch from one provider to another easily and rapidly, even on a real-time basis. In light of these requirements, this article addresses the issue of service provisioning in an ATM-over-ADSL access network. It concentrates on the provisioning of services using the PPP protocol. PPP is the common protocol for service provisioning in circuit-switched telephone networks. However, it is also considered a good choice for the delivery of broadband services since it has built-in mechanisms for IP address assignment, layer-2 security, and AAA (authentication, authorization and accounting). This article presents the problem of initiating a PPP connection at an Ethernet-based host. It presents several solutions for this problem and discusses the advantages and disadvantages of each solution.
...
...
Conclusions
This article has addressed the framework of service provisioning in an ATM-over-ADSL access network. It has described the architecture of such a network and discussed the "PPP over Ethernet" issue. It has presented several solutions for addressing this issue, namely PPPOE, L2TP, BMAP, and proxy-PPP. The article has shown that proxy-PPP has many advantages over the other approaches, the most important of which is that no modification is needed in the users' hosts, and that no new protocol needs to be defined.
The article has also shown that the proxy-PPP solution works extremely well in conjunction with a Web-based service selection scheme. In such a scheme, the ATU-R also serves as a lightweight Web server. A user who needs connectivity to an NSP gets from the ATU-R an HTML service menu through an HTTP session. The selection of the user, along with the associated parameters, is then forwarded to the ATU-R using another exchange of HTTP messages. The ATU-R then has all the information needed in order to set up a PPP connection to the target NSP. Such a Web-based service selection scheme may help to promote new services and to attract new NSPs to offer their services. Moreover, no configuration is needed at the user host; users can get online information regarding the available NSPs at any time, and service selection based on QoS considerations can be supported. |
