Revised January 20, 2000
Contact: Ron Olton
As part of long-range network planning, the colleges were asked to project their needs or additional Classroom and Conference Room facilities. The results show that there is a desire to add approximately 80 end points. As might be expected, these projections are contingent upon the availability of funds and the exact nature of the increased functionality.
The long-range strategic plan for videoconferencing is to use native ATM SVCs to carry video traffic. This method makes the best use of ATM resources and is also the most cost effective way to implement video services across an ATM infrastructure. When the PVC based video model was implemented neither the ATM equipment nor the video equipment could provide the necessary SVC support. Advances in technology over the past two years now make native ATM SVCs possible.
This proposal recommends changing the PVC based video model to an SVC based model, while expanding the number of sites. The SVC based model provides a long-term solution in that it implements a methodology that is scaleable throughout Virginia. It also eliminates the need to continuously update the video model, produces the lowest overall cost over time, and provides the most flexible video services architecture possible.
Conceptually, there is no "video network" in the SVC model. It is a collection of end points (colleges, high schools, state agencies, etc.) and video bridges connected to the ATM infrastructure. The end points connect to each other dynamically, just like placing a telephone call. As with a telephone call, the only thing an end point needs to know is the ATM address (telephone number) of the other end point. This architecture allows end points and video bridges to be added as needed, without disrupting existing sites. It supports both point to point and multipoint conferences. Any Net.Work.Virginia customer may upgrade their codec and ATM switch to the same level as described in this proposal. Sites that do will have the same level of video services as the VCCS colleges.
The SVC based model provides for local management of college resources and the sharing of video resources that cannot be cost justified by every college. The cost justification of a video bridge is the most difficult task for a college. Therefore, the SVC model treats video bridges, and their associated scheduling system as shared resources. An ATM address translation function is also needed in the SVC model, and is also a shared resource. The distributed resources are the college Classrooms, Conference Rooms, and scheduling systems. With an SVC based model the colleges have the flexibility to manage and operate their video conferencing environments, using the shared resources only when necessary.
The address translation device noted above provides a way to associate easy to remember names with the ATM addresses of specific locations, thus eliminating the need to remember ATM addresses. This provides a user friendly way of dialing remote sites. The SVC proposal calls for two shared devices, a primary source and a backup device. Colleges may elect to install a local device that would serve as the college primary source, and use the shared devices as backups.
Provision is made for temporary support of the H.320 PVC based video model. This allows colleges and Net.Work.Virginia sites that do not upgrade to the SVC model to retain the PVC level of video services. Connectivity between the two models will exist, but specific conference service levels will depend on the capabilities of the sites in the conference. This temporary support for the PVC based systems will be provided until June 30, 2000.
Some of the benefits that will accrue to the colleges by upgrading to the SVC based model are as follows (See Appendix A for details):
The PVC model could be expanded, rather than implementing the SVC based model. The one time and ongoing costs of upgrading the shared resources in both models are shown below. Total college costs cannot be compared because they depend on what specific actions each college might take under each model.
To expand the PVC based model by 80 sites (see Attachment B.) -
One Time Cost - … $ 2,521,500.00
Ongoing Cost - ……. $ 312,150.00 / year
To use the SVC based model (see Attachment C.) –
One Time Cost - ….. $ 1,054,000.00
Ongoing Cost - ……... $ 105,400.00 /year
In the SVC model, colleges have five upgrade options (See Appendix D for details and cost estimates). The five options are:
1. The first option applies to existing Vtel 227 codecs. It replaces the codecs with codecs that are compatible with the industry standards specified in the proposal, and upgrades the ATM switches to support ATM switching. This will result in the campus being able to take full advantage of the new functions noted above. Vtel has tended an offer that provides a trade in allowance for existing codecs, on the purchase of new models.
2. The second option applies to new Classroom and Conference Room equipment. It also applies if a college elects to purchase a new codec and keep the existing unit. It calls for the purchase of codecs that support the proposed model. This will result in the new facility being able to take full advantage of the new functions noted above. This option assumes that the campus ATM switch has already been upgraded to support ATM switching for an existing Classroom.
3. The third option applies to sites that have a newer Vtel ESA codec that is already standards based. It involves upgrading the codec and campus ATM switch. An ATM interface module is added to the codec, and its software is upgraded. The cost of the software upgrade is covered by the existing maintenance on the codec. If not already done, the campus ATM switch will also have to be upgraded to support ATM switching. This will result in the facility being able to take full advantage of the new functions noted above.
4. The fourth option applies to existing cell multiplexers. This option retains the existing codec, but replaces the cell multiplexer with a device capable of providing the ATM switching capability. This will result in adding switched capability, reducing the cost of shared resources and providing the ability to conference directly with other sites in two site conferences. But it will leave the college in the position of having only basic video and audio support when conferencing with sites that have upgraded their codecs. This option retains the current proprietary data sharing capabilities when conferencing with other sites that have not upgraded their codecs. This option also requires an upgrade to the campus ATM switch.
5. The fifth option is to do nothing. As noted above, the PVC based model will be maintained until June 30, 2000 for those sites that do not or can not upgrade to the SVC model. These colleges will have only basic video and audio support when conferencing with sites that have upgraded to the SVC model. The PVC model retains the proprietary data sharing capabilities when conferencing with other sites that have not upgraded their codecs.
The nature of the SVC design allows a migration from the PVC video model to the SVC model on a room by room basis. The first step is to install the new video bridges and scheduling system. The rooms may then be converted, individually or in groups. New rooms may be added as needed.
As might be expected, conferencing capabilities during the migration period will depend on the capabilities of the rooms involved in the conference. The SVC implementation schedule will be coordinated with the colleges to ensure that classes scheduled during the migration period are not interrupted.
The implementation events or activities are as follows:
Appendix A.
The adoption of additional International Telecommunication Union Standards will ensure that the colleges will be able to communicate effectively among themselves and enhance the quality of courseware through the increased use of file and document sharing, white boarding, etc. It will also improve the chances that a college will be able to communicate with a non-VCCS partner, since the trend in video conferencing is to use a video standards based, rather than a non-standards based approach. Colleges will have the flexibility to select any Classroom or Conference Room equipment that complies with the standards adopted in the SVC based model.
The weakest link in the PVC based model is the scheduling system, and its connection to the video bridges. The SVC based video bridges have a new scheduling system that connects to the video bridges via the VCCS Intranet, not telephone lines. This makes the connection more reliable.
Studies in January 1998 show that approximately 75% of the scheduled conferences contained only two sites. The SVC based model introduces the concept of separate college scheduling systems to manage these conferences. This provides the flexibility to manage these conferences at the local level, between the two participating sites. Bypassing the video bridges allows a dramatic increase in the number of sites that can be supported system wide. The current estimate is that 4 video bridgeses can support the needs of the approximately 120 sites in the current projections.
The proposed video bridges are capable of building conferences between sites running at different speeds, and sites using different audio and video protocols. This expands the range of sites that colleges can conference with. The new bridges will also have an on-board calendar to store the schedule of future conferences. This improves reliability by allowing the video bridge to start conferences even though the scheduling system may be inoperative. In addition, the new bridges have the ability to provide split screen support for the instructor. This, along with picture-in-picture support, reduces the need for dual monitoring at the sites.
Appendix B.
To expand the PVC based model, without changing the basic technology will require:
One time costs -
1 new hub video bridge @ $550,000.00 ….………...$ 550,000.00
4 upgraded video bridges @ $87,500.00 ……….….... 350,000.00
6 new video bridges @ $173,000.00 ………….…... 1,038,000.00
11 Scheduling systems @ $25,000.00 ………………. 275,000.00 *
1 new ATM DS-3 line installation @ $1,000.00 …….… 1,000.00
8 new CP-300 multiplexers @ $35,000.00 ………….. 280,000.00
11 new ATM-25 switch modules @ $2,500.00 …….… 27,500.00
Total One Time Costs …………………………... $ 2,521,500.00
Ongoing costs -
1 new ATM DS-3 line @ $5,000.00 / month ………….. 60,000.00
Equipment maintenance @ 10% of cost / year …….… 252,150.00
Total Ongoing Costs ……………………………….. $ 312,150.00
* The assumption is that one scheduling system will provide adequate performance. Since the scheduling system has not been tested, a worst case scenario is used in this estimate(one scheduling system per video bridge).
Appendix C.
As noted above, the VCCS conferencing workload consists of approximately 75% point to point and 25% multipoint conferences. This estimate is based on supporting only the multipoint workload. The estimate is that 4 video bridges will support the workload in an SVC based model that contains about 120 sites. Two ATM address translation devices (V-Gates), and a new scheduling system are also be needed.
One time costs -
4 SVC based video bridges @ $229,000.00 ……... $ 916,000.00
2 Address translation devices @ $14,000.00 ……….. 28,000.00
4 Scheduling systems @ $25,000.00 ………………. 100,000.00 *
4 new ATM-25 switch modules @ $2,500.00 ……… 10,000.00
Total One Time Costs ………………………… $ 1,054,000.00
Ongoing costs -
Equipment maintenance @ 10% of cost /year ……… 105,400.00
Total Ongoing Costs …………………………….... $ 105,400.00
* The assumption is that one scheduling system will provide adequate performance. Since the scheduling system has not been tested, a worst case scenario is used in this estimate (one scheduling system per video bridge).
Appendix D.
Option 1:
One college upgrade path is to replace the original Vtel 227 codecs with codecs that are video standards based, and that have an ATM-25 interface to connect them to the campus ATM switch. This provides the college with the maximum flexibility. Vtel has agreed to take the existing codecs as trade ins on new Vtel codecs, to help the colleges minimize the cost of upgrading.
A new ATM-25 interface module will be needed in the campus ATM switch to connect the new codec to the ATM network. This new module will replace the ATM DS-1 module in the switch. Those campuses that host a video bridge will not need this module. They can use available ports in the module supporting the video bridge.
The cost of this portion of the upgrade is -
- 1 new codec (less trade in allowance) - …….… $ 19,388.00 *
- 1 ATM-25 PAM @ $2,500.00 - ………..………... 2,500.00**
Total cost - …………………………...…………$ 21,888.00
* Includes codec, ATM interface and maintenance (1 year).
** May not be required.
Option 2:
To add video facilities a new, standards based codec with an ATM-25 interface is required. This provides the college with the maximum flexibility.
An ATM-25 interface module will be needed in the campus ATM switch to connect the new codec to the ATM network. This estimate assumes that the college has already installed one to support the existing Classroom upgrade.
The cost of this portion of the upgrade is -
- install new campus wiring for the new room - ?
- 1 new codec (assumes a Vtel LC5000) - …………. $ 38,145.00 *
Total cost - ………………………………………….$ 38.145.00 +
* Includes full room equipment (codec, camera, monitors, etc.). This cost may be lower, depending on which equipment the college selects.
Option 3:
Those sites with Vtel ESA based codecs (which are already video standards based) may upgrade them with an ATM-25 interface to connect them to the campus ATM switch. The upgrade consists of installing a VC-NIC card in the codec and upgrading its software. The software upgrade cost is covered under the existing maintenance on the codec.
A new ATM-25 interface module will be needed in the campus ATM switch to connect the new codec to the ATM network. This new module will replace the ATM DS-1 module in the switch. Those campuses that host a video bridge will not need this module. They can use extra ports in the module supporting the video bridge.
The cost of this portion of the upgrade is –
- 1 VC-NIC card @ $9,000.00 - …………………… $9,000.00
- 1 ATM-25 PAM @ $2,500.00 - ………………….… 2,500.00*
Total cost - …………………………………….…..$ 11,500.00
* May not be required.
Option 4:
Those sites with Vtel 227 codecs (which are not video standards based) may continue to use them. The local cell multiplexer is replaced by a local ATM switch (V-Room), which provides the ATM switching capability. Colleges selecting this option will not have data sharing capabilities with sites that have standards based codecs.
A new ATM-25 interface module will be needed in the campus ATM switch to connect the new codec to the ATM network. This new module will replace the ATM DS-1 module in the switch. Those campuses that host a video bridge will not need this module. They can use extra ports in the module supporting the video bridge.
The cost of this portion of the upgrade is –
- 1 V-Room switch @ $14,400.00 - ……………. $ 14,400.00
- 1 ATM-25 PAM @ $2,500.00 - …………………. 2,500.00*
Total cost - ……………………………………...$ 16,900.00
* May not be required.
Option 5:
Those sites with Vtel 227 codecs (which are not video standards based) may continue to use them. In this option, the local cell multiplexer is retained. Colleges selecting this option will not have data sharing capabilities with sites that have new (T.120 Standard compatible) codecs, nor will they have the ability to connect directly to other sites for point to point conferences, without using of a video bridge.
The ATM DS-1 PAM in the campus ATM switch is also retained.
There is no cost associated with this option.
Return to Projects