Videoconferencing integrates analog, digital on microwave network - Washington Higher Education Telecommunication System

Communications News, Feb, 1993 by Dennis Ostraat

In 1984 Washington State University in Pullman expanded its distance learning facilities, the Washington Higher Education Telecommunication System (WHETS), to Spokane, Richland, Vancouver and Seattle, via a full duplex terrestrial analog microwave system.

WHETS uses two-way video, trying to achieve the same instructional environment for remote as well as on-campus students. The classroom has remote-controlled cameras, fixed lighting and sound pickup. Instructors and students adapt to the multiple classroom quickly and maintain a high degree of interaction.

In the 1989-90 academic year WSU operated 35 classes and 291 hours of teleconferences. In 1990-91 those numbers grew to 65 classes and 438 hours; by 1991-92 they were up to 82 classes and 617 hours. Each class averages three credits totaling 45 hours per course per semester.

This growth created major scheduling problems for circuit availability. With the increasing demand for data communications, it became obvious that it would be impossible to meet WSU's needs with more analog circuits in a finite frequency spectrum. To more efficiently use the frequency spectrum, WSU proposed a conversion to compressed video with a digital interconnection.

In 1990 the Washington legislature funded a modified proposal to replace one of the analog circuits with digital microwave to the branch campuses.

Original equipment estimates took for granted that digital radios could simply be added to existing analog waveguides and sites. After the RFP was released for the digital radio systems, it was apparent we would have to take an extra year and redesign the network to fit within budget. Four of the paths were 50 or more miles long. Although the analog microwave worked adequately over these paths, reliability calculations for digital microwave were marginal.

We eventually purchased radios from Farinon, in part because of their success with a project in Hawaii which added the Farinon digital radios to an existing analog microwave using paths 60 or more miles long.

Maintaining one analog circuit presented us with the dilemma of integrating analog and digital systems and managing network switching systems. In the RF world, integrating analog and digital is fairly straightforward. But with analog video the only common point available to manage the switching systems is at the analog signal level.

Our planning called for more conversion to digital so a parallel management system with a multipoint control unit (MCU) was purchased and integrated with the analog network management system.

Since analog traffic was to be maintained, selection of a codec which would provide high quality video was paramount. The CLI Rembrandt II VP was selected after evaluations of various products for 18 months. It operates in the CTX plus mode, 30 frames per second, at a DSI data rate. Codec output is crossconnected directly into the radio muldem (multiplexer-demultiplexer).

To date we have not seen a need for a CSU/DSU. Students and faculty have readily accepted the codec video and do not comment on differences between it and analog video.

Because of the one-year delay and a commitment to maintaining class schedules, the construction was compressed into the three summer months of 1992. The schedule was extremely tight--the last leg was completed only 36 hours before the start of classes.

Several major problems have yet to be resolved. We were aware that the audio would not be the same quality or bandwidth as with the analog system. We were not prepared, however, for the critical nature of operating audio with the echo cancelers used on the codec systems. With the analog system, our operators were used to changing microphone and speaker levels slightly if there were complaints. With echo cancelers this can create many problems.

We are revising our audio system. Some classrooms have as many as 20 microphones operating though a voice-activated mixer. Echo cancelers will not tolerate gated systems and we are having to reduce microphone count plus add acoustic treatment to some of the rooms which are more live. With echo cancelers, multiple site audio is much harder to deal with than point-to-point.

The lessons we have learned from this project are:

Leave adequate time for testing the new installation. There will never be enough time for resolving problems after facilities are scheduled for classes.

Research audio systems with an "on site" evaluation. The equipment is not used in the controlled environment of trade shows.

Acoustic treatment for the classrooms is a must.

The future of the WHETS network will be growth. We are looking at cascading multipoint units and at integrating a DACS (digital access crossconnect system) into the network management switching. The next stage for development is interfacing the WHETS network to the $17 million communications infrastructure system for voice, video and data that is being installed on the Pullman campus.