Voice services on campus, and at some off-campus sites, are provided by an Ericsson MD110 PBX. There are approximately 12,300 campus lines and 1,000 trunk lines. The telephone system has been growing by about 250 lines per year for the last few years. On a typical day there are 41,000 outgoing trunk calls and 64,000 incoming trunk calls.

The central switching matrix, the Group Switch, is located at Node M in the Central Utilities Building. Telephones are cabled back to a total of 66 Line Interface Modules (LIMs) which are located in the node rooms on and off campus. The remote node rooms are connected to Node M by fiber or occasionally by leased T1 lines. The Ericsson PBX was installed in 1986, and has been upgraded many times since then. Only a few of the original components remain. The critical Group Switch is one of these original components, and while there have been no problems with it to date, we do not know if there will be future issues with this aging equipment.

We have been watching developments in the Voice Over IP (VOIP) field, and have tested Ericsson’s VOIP system. A major advantage of integrating Ericsson’s own VOIP system with our present network is being able to move numbers from one system to the other. Also, most of the features are accessed in the same way, and the look and feel of the phones is very similar. The trial installation of VOIP at the Mesa Housing maintenance office was successful, and at the customers’ request, we made that installation permanent.

We provide voice mail services through a Pulsepoint 3110 voice mail system, which presently has 6,600 mailboxes and 128 ports. Voice mail usage has increased by about 200 mailboxes per year, but we have recently reduced our voice mail rates and so growth may accelerate. The Pulsepoint 3110 is being discontinued, and although we can still renew the maintenance, we will need to replace the voice mail system within the next few years.

We have recently purchased some other enhancements to the voice network, including an Automated Call Distribution (ACD) system and a Reverse 911 system. We are also doing some initial testing of Ericsson's mobility service, which allows a cell phone to be used as an extension of the PBX with five-digit dialing.

UCSD continues to grow, and some major new buildings will be occupied over the next few years. The growth in telephone lines, however, has not been matching the growth of the data and wireless networks, and we do not anticipate any capacity problems or resource issues on the voice side.

UCSD is also moving more staff to off-campus buildings. Whenever possible we install a LIM in these buildings so that the customers there can get the same level of service as they would on campus. It is in this type of location where Voice Over IP may become a more cost effective alternative to a LIM.

An increasing number of people are using their cell phones as their primary or only voice instrument. They either forward all of their Ericsson calls to their cell phone or, especially in the academic community, they do not have an Ericsson number.

A few years ago there was a sense that Voice Over IP to the desktop was the newest and best way to provide voice services. This "coolness factor" has now worn off, and VOIP is seen as a maturing technology which in some circumstances is more cost effective than circuit switching. As the years progress VOIP will be used in more and more installations, and it will overtake circuit switching as the most popular voice transmission system. VOIP is still improving and already offers most of the features which we need, with the possible exception of multiple line appearances. The main benefit of using VOIP to the desktop will be to save money in equipment, maintenance and support costs. It lends itself well to buildings which have good closets with extensive UPS systems, and does not work well in older buildings where space and environmental conditions are more of a challenge. It needs to be deployed in a well thought out manner. There have been some very high-profile failures of VOIP installations that were not well planned.

For VOIP to work effectively the data network must be modern, well engineered, highly reliable, and centrally controlled with strong policies. Over the last few years UCSD’s network has been making great strides toward becoming this type of network, although there are still some areas where system administrators and other departmental users reconfigure their switches or do other things that can damage a local part of the network. VOIP will not work well in these areas.

Unlike VOIP, wireless technology offers something new. Mobility allows people to work in a different way, and will have a much bigger impact on the future of voice services. Ericsson is a leader in wireless services in many parts of the world, and we will be in a good position to integrate effective wireless capabilities into our network.

Some PDAs will be built with 802.11 voice applications, allowing them to be used as a telephone when using the wireless data network. This may be of interest in some technically oriented areas such as Engineering, especially if cell phone capabilities can also be added to the same device, which would automatically connect to the cellular network if there were no 802.11 signals in the area. Such dual technology telephones are several years away.

Flat-rate calling is becoming more common for residential and small business users. It is not yet available for institutions of our size, but we anticipate that within two years we will be able to buy flat-rate services from our long-distance providers.

VOIP to the desktop is not to be confused with VOIP trunking. The latter is a way of saving long-distance costs by sending a call over an IP datastream for part or all of its route. This is supported by our current Ericsson software and there are very few technical drawbacks to it. A cost benefit analysis usually determines if it should be done or not. The more expensive the per-minute charges for voice calls, the easier it is to justify VOIP trunking.

The voice mail vendors are now switching to selling Unified Messaging products. These optional enhancements to a voice mail system allow voice messages to be retrieved and stored as e-mail attachments, and e-mail messages to be read out by a synthesized voice. At present this is may be a solution searching for a problem as there has not been much of a demand for unified messaging on campus. Unified messaging is particularly challenging when multiple e-mail platforms need to be supported, but improvements are being made in this area. Another factor counting against unified messaging is that it is becoming irrelevant because of the new portable devices that provide both voice and data capabilities.

Ericsson has announced future plans for upgrading their MD110 PBX. Their BC14 upgrade, in three years, will eliminate the group switch, which will be replaced by IP trunking between the LIMs.

• Replace our existing Pulsepoint system at some stage in the next three years

  Although we are uncertain if and when Unified Messaging will become widespread, we should replace it with a system which can be enhanced to provide unified messaging, so that we have the option of offering it in the future.

• Upgrade the PBX to BC14 when available, although we do not need to be one of the first 'guinea pig' sites to do so

• Install a larger (30 to 40 user) test site of Ericsson VOIP, and a similar sized test of Cisco VOIP

  We need to stay current on developments in the VOIP market, and understand all of the pricing issues. Over the next few years VOIP will becoming increasingly attractive for new sites both on and off the campus.

• Schedule additional cross training, so that the data engineers are more familiar with the voice services and vice versa, as it will be important for future VOIP installations

Some residential carriers are offering flat-rate services, which include unlimited long distance. Although vendors are not currently offering this to large business customers, we will investigate including flat-rate long distance service as part of the NGN when the model is recast. Due to the possibility of abuse and loss, we should only introduce domestic calling in the NGN fee if we can negotiate flat-rate trunks with our providers. Even if we move to an increased NGN rate to cover domestic calling, we will still need to collect call record details for international and they may be necessary for audits, security, etc. It is unlikely that international flat rate trunks will be available within the next five years, although they may be available to certain countries such as Canada and the United Kingdom.

We should work with CENIC to investigate trunked VOIP to other UC campuses and sites connected to the CENIC network. Inter-campus calls could then be made at no cost to the campus (after some set-up investment).

The increased use of cell phones has had and will continue to have an impact on our revenue, and the costs that we cover through long distance may need to be transferred to other businesses. This may be a challenge as we have recently reduced our long distance recharge rates.

BC14 will cost in the range of $400,000. This significant sum will extend the useful life of the Ericsson PBX for several more years, and should be considered part of the future Next Generation Network costs. It is unlikely that the Ericsson system will need to be replaced in the forseeable future. If it is, any new system will have a very different architecture than a traditional PBX. We should be aware that the time will come when we will need to spend significant money in order to continue providing voice services to the campus.