Detailed Descriptions of the Options
Option 1, expand the GTD4600 to meet the requests for new lines
until the end of the original Certificates of Participation repayment date on
September 1, 2000 (the need is projected to be at 6,345 lines by that date) and
then replace the system, would provide somewhat of a "business as usual"
solution to the problem for the short term. The reality of the situation is
that this option represents a significant investment in a technology which has
been stagnant since 1989, when GTE sold their telephone switch manufacturing
business to Fujitsu. Since that time all product development has ceased and
parts availability is limited to used equipment.
There is no reason to believe that the current system will be able to
provide basic, reliable service until the year 2000, let alone meet the requests
for the newer types of enhanced services which are rapidly evolving. The number
of technicians and software support personnel who are familiar with this system
are dwindling rapidly at both GTE and Fujitsu. GTEL's ability to provide
quality maintenance service is already deteriorating and we may experience total
failures of the switch for extended periods in the future unless we augment the
existing maintenance capabilities of GTEL. If we continue to use the system
beyond the end of the annual maintenance agreement on October, 1997, we would
have to train multiple UCSB staff in maintenance and repair, without having the
benefit of a manufacturer's training class, and should maintain an adequate
level of spare parts on campus.
Option 2, acquire additional equipment to expand the GTD4600 to
meet the requests for new lines until July, 1997 (the current end of
amortization) and then replace the system, provides a business as usual
solution to the problem without the risk of a long term investment in obsolete
technology. This option would provide service to the campus during an interval
which would be used to analyze customer needs assessment survey data, consult
with the campus community, competitively bid a replacement system and, if
needed, construct a new switch room.
The risk in exercising this option is that there may be service
interruptions as the result of the software changes to the data base of the
GTD4600. Because there has been no continuing software development for this
system, the changes made to the system in recent years have been software "patches".
These patches are apt to be lost during the expansion of the system, and will
have to be hand entered as their loss is discovered through customer trouble
reports. This potential risk can be mitigated, but probably not eliminated, by
careful scrutiny of the software when it is returned from Fujitsu. Since the
system expansion could be planned for the 1995 holidays in December, there would
be an opportunity to test the system with a minimum of impact upon the campus
Option 3, move some telephone equipment (e.g., modems and
emergency telephones) off of the GTD4600 onto GTE business lines in an effort to
extend the useful life of the current configuration until July, 1996 and then
replace the system in July, 1997, is designed to eliminate any future
capital investment in the GTD4600 by recovering lines such as the modem pool
lines, emergency telephone lines, and lines on "temporary disconnect".
The recovered lines would be use to satisfy requests for additional telephone
service from our customers. The modem pool and emergency telephones would be
connected to GTE business telephone lines. We would lose an extremely important
feature of the emergency telephone system called Automatic Location
Identification (ALI). ALI automatically provides Police Dispatchers with the
physical location of the person using the emergency telephone, negating the need
for the person to describe their location. Elimination of this feature may be a
violation of California Public Utility Commission regulations, but this has not
yet been verified.
If the modem pool lines are moved off of the GTD4600 and on to GTE business
lines, it would be necessary to change the telephone numbers used to access the
three modem pools.
The lines on temporary disconnect would have to be carefully managed and
tracked, but eventually the GTD4600 switch will run out of lines in
approximately June, 1996 since this strategy only recovers 347 lines.
Although this is a relatively cost effective solution on the surface, it
only delays the inevitable for about 6 months, which is not enough time to
select and implement a replacement telephone service. If a new system is to be
installed on campus, a new switch room will not be available for occupancy until
November, 1996, at the earliest; installation, testing and implementation of a
new system would add another 7 months, which means 12 months of "no growth"
for telephone service. Under this option, requests for new telephone lines
between July, 1996 and July, 1997 could not be satisfied.
Option 4, convert an unused portion of the 893 prefix to
Centrex, provide a hybrid service to meet the requests for line growth until
July, 1997 and then replace the system, is based upon the concept of
combining a rented service with our own telephone switch to provide line growth.
CentraNet, (GTE's trademarked name for Centrex service), can be rented one line
at a time on a month to month basis. By taking unused numbers in our 893
telephone numbering plan and assigning them to CentraNet service, additional
CentraNet lines could be provided with features similar to those which exist on
the GTD4600, providing a relatively transparent, hybrid service with features of
both CentraNet and the GTD4600.
Certain feature limitations would occur with this arrangement, the most
noticeable of which would be the loss of the "station camp-on with call
back" feature between the CentraNet and the GTD4600 (in both directions). "Call
pickup" would not work in an office which had a mixture of CentraNet and
GTD4600 lines, and call transfers from one system to the other would have to be
monitored until the other party answered (instead of transferring and hanging up
when the other party's line begins to ring). This requirement to stay on the
line when transferring would be noticed most by the UCSB operators and other
departmental telephone receptionists.
Another feature difference would manifest itself when emergency calls to "9-911"
from off-campus locations were placed. Currently, calls to "9-911"
from off-campus locations such as Purchasing on Aero Camino end up being routed
to the campus Police Department since the call really originates here on campus
via a line or extension which has been extended to an off-campus location. In a
CentraNet environment, the call would originate in GTE's central office via a
CentraNet line located on Aero Camino, go into the 911 network as a call from an
address on Aero Camino, and be routed to the Santa Barbara County Sheriff's
Office. This emergency call routing process is in conflict with the UCSB Police
Department's policy of being the first agency to be dispatched to an emergency
at a site occupied by UCSB staff. The only way to mitigate this problem would
be to set up a unique dialing plan for emergency calls, such as "8-911",
for emergency calls from CentraNet lines located at off-campus locations. This
would result in the calls being routed to the GTD4600 and subsequently to the
UCSB Police Department, but would also result in the loss of Automatic Number
Identification (ANI), since the calling number of the CentraNet line cannot
transit the GTD4600 on its way into the 911 network. The loss of ANI would
result in the loss of ALI, which means that the 911 dispatcher would have to get
the correct address from any caller who was reporting an emergency from a
CentraNet line , regardless of whether or not the caller was located on or off
campus, and regardless of whether the caller dialed "8-911" or "9-911".
From an administrative point of view, difficulties could arise when lines on
the GTD4600 became available as the result of a request for a disconnection, and
these lines became available for reassignment to new customers. Policies would
have to be developed to determine who was entitled to a line with no feature
limitations and who would receive a CentraNet line.
In order to provide voice mail service to customers with CentraNet lines, it
would be necessary to partition the UCSB Voice Server into "domains".
Although the technical literature on the UCSB Voice Server indicates that the
feature limitations between domains is controlled by the system administrator,
it is unlikely that service between the two domains would be completely
transparent. Forwarding messages between domains, and creating broadcast lists
including message boxes from both domains could entail exceptionally complicated
procedures. One feature difference which cannot be mitigated is that the
message waiting indication for campus lines is "high tone", and for
CentraNet lines the message waiting indication is "stutter dial tone".
There would be other slight differences in the way that the two systems
behaved, but these differences could be mitigated with customer training when
the CentraNet lines were installed. These differences include such things as
different feature activation codes for speed call and three-way calling. Most
feature activation codes would be the same.
This option has the benefit of being risk-free in terms of capital
commitment, and it provides the time necessary for needs assessment,
consultation, and competitive procurement. It's single largest liability would
be lack of customer acceptance based upon the feature limitations, which in
certain circumstances could be conceived as more than just an inconvenience, but
rather as an impediment to one's ability to perform their job. Based upon the
line growth requirement between January, 1996 and June, 1997, this service would
be provided to between 500 and 600 customers.
Option 5, a combination of options 3 and 4 above, attempts
to minimize the number of hybrid CentraNet lines being provided to the campus by
first migrating the modem and emergency telephone lines off of the GTD4600 and
then providing CentraNet as the last resort to customers. Both the migration of
lines and implementation of CentraNet would occur concurrently, but an effort
would be made to ascertain when CentraNet could be used for less personal
applications such as facsimile, modems, conference room phones, lobby phones and
similar applications. The early assignment of CentraNet lines for non personal
applications would in effect conserve the lines on the GTD4600 for traditional
applications for faculty and staff members. It would be difficult to predict
how many of the new lines would fall into this category of non personal usage,
but the net effect of this combined strategy of recovering lines, assigning new
lines by type of usage, and ultimately installing exclusively CentraNet lines,
would certainly result in fewer individual customers being assigned the
CentraNet lines with their inherent feature losses.
The benefits of this option are essentially the same as option 4, and while
the liabilities remain the same for this option as for the previous option, they
impact fewer customers. Assuming that 50 lines could be assigned to facsimile,
etc., this option would result in Centranet service ultimately being assigned to
only 150 to 200 customers.
Option 6, move some telephone equipment (e.g., modems) off of
the GTD4600 onto a new stand-alone telephone system, use the stand-alone system
to support emergency telephones, the modem pool, and campus line growth until
July, 1997 and then replace the system, is nearly identical to option 5,
except that the line growth which would have been provided by CentraNet would
instead be provided from a rented, stand-alone telephone system, the
same system which would be used to support the modem lines and emergency phones.
This approach has the benefit of being risk-free in terms of capital
commitment, and it provides the time necessary for needs assessment,
consultation, and competitive procurement. It also obviates the need to have
GTE involved in the provisioning process for new lines, since new lines could be
provided by Communications Services directly from the rented stand-alone system
(this would reduce installation intervals, repair intervals, and service order
costs). The loss of certain features and subtle differences in using other
features would be the nearly the same for the rented stand-alone system as it
would be for CentraNet service. In almost every respect, the stand-alone
telephone system would be analogous to having CentraNet service, except that the
supplementary source of dial tone would come from a rented telephone system
located side-by-side with the GTD4600, rather that dial tone rented from GTE's
off-campus Central Office.
Regarding the feature differences, a stand-alone telephone system would
suffer from the loss of the camp-on, call pickup, and transferring without
monitoring features. Emergency calls from lines on the stand-alone system,
whether they were located on-campus or off-campus, would be accomplished by
dialing "9-911", and these calls would be routed to the Campus Police
Department. The ANI and ALI would be lost, however, since the calling line
number would not transit the 911 network.
The campus emergency phones would be connected to the stand-alone system,
which in turn would be connected to the GTD4600 by means of multiple tie lines.
This arrangement would allow for the ALI feature to be used by the Campus Police
Department so that the location of the campus emergency phones would be known by
the police dispatcher on a call by call basis. These multiple tie lines would
also be used for the modem pool connections so that it would not be necessary to
change the pilot numbers of the modem pools. Calls to the modem pools would
come into the GTD4600 and they do today, and be switched to the stand-alone
system via the tie lines.
In order to provide voice mail service to customers with lines from the
stand-alone system, it would be necessary to partition the UCSB Voice Server
into "domains". Although the technical literature on the UCSB Voice
Server indicates that the feature limitations between domains is controlled by
the system administrator, it is unlikely that service between the two domains
would be completely transparent. Forwarding messages between domains, and
creating broadcast lists including message boxes from both domains could entail
exceptionally complicated procedures.
Ultimately, approximately 150 to 200 customers would be provided service
from the rented stand-alone system.
Option 7, replace the GTD4600 with leased service from GTE's
Central Office in Ellwood, no later than the first quarter of 1996, explores
the possibility of simply renting CentraNet service from GTE as an alternative
to the GTD4600, and converting the entire campus to CentraNet to eliminate the
feature differences and limitations. Since this represents a significant
capital investment on the part of GTE in an expansion of the Ellwood Central
Office and a major reconstruction of the conduit system between Ellwood Station
Road and the UCSB switch room, GTE would require a seven (7) year contract for
the CentraNet service. Because of the lead time necessary for the line addition
in GTE's Central Office and the reconstruction of the conduit system, GTE has
stated that it could be 4 to 6 months from the signing of a CentraNet contract
before the campus could be converted to Centranet.
This option has the effect of eliminating the time associated with needs
assessment, consultation, and competitive procurement, since, at this time, GTE
is the only telephone service provider who can provide UCSB with leased service
(this condition changes in January, 1997). This option also requires no
capital, since the investment would be made by GTE and recovered in their 7 year
The liability of this option is that it eliminates all the benefits of needs
assessment, consultation, and competitive procurement, and the campus would be
restricted from adding new lines during the 4 to 6 month interval immediately
after the contract is signed. In short, the campus would be committed to 7
years of telephone service which may or may not meet present and future
technological and feature requirements, at a cost which, without the benefit of
competitive procurement, would significantly raise the rates for telephone
service to the campus.