Fiber-optic switched digital video networks conduct educational lessons
Distance-learning applications over fiber-optic networks link online instructors and educational resources to dispersed students in communities, schools and businesses
Earl W. Philmon
Broadband technologies inc.
Employing a videoconferencing approach, education in schools, colleges and hospitals is being enhanced by fiber optics technology, in general, and by switched digital video networks, in particular. The impact of these networks is driving the deployment of short- and long-haul distance-learning networks and a variety of communications equipment into classrooms.
In fact, Kessler Marketing Intelligence Corp., a Newport, RI-based fiber optics marketing consultancy, reports that nearly 100 distance-learning systems using fiber optics technology have been set up, mostly in the eastern United States. In the early 1990s, states KMI, these types of networks generally interconnected as many as 20 institutions, but this year, the latest networks are connecting hundreds.
These networks are structured to provide students at one location with an instructor at another location using video-based capabilities. Colleges, schools and hospitals are supporting the delivery of an expanding library of multimedia educational programming services. Highly interactive programs featuring graphical displays and user navigation capabilities have conclusively demonstrated that student learning is improved by making the educational process more visual, enjoyable and entertaining.
Switched digital video networks used for distance learning satisfy the need for highly interactive emerging services and incorporate the advantages of fiber optics transport technology. The basic technology used in such networks is an offspring of the full-service network, fiber-to-the-curb products developed as an access platform for the forthcoming universal broadband video dialtone deployments. Applying this technology to distance learning allows educational networks to realize cost savings associated with this architecture while assuring an upgrade path as network service requirements expand.
The fiber-optic network technology commonly deployed for full-motion video in the educational environment employs a 45-megabit-per-second per-channel algorithm. Encoding technology for the transmission and receiver sites can be deployed at a cost of just hundreds of dollars per video channel, compared to a cost of tens of thousands of dollars for Motion Picture Experts Group, or MPEG, digital encoding. Therefore, even though the 45-Mbit/sec technology requires higher bandwidth, it is favored over MPEG technology because of the cost savings. As MPEG technology matures, and further integration and large-scale device manufacturing occur, this cost disparity is expected to diminish.
Increased channel capacity
With the introduction of lower bit rate algorithms comes a large increase in channel capacity. For example, with MPEG-2 streams running at 6 Mbits/sec, channel capacity increases by a factor of six. Migration becomes uncomplicated, however, because asynchronous transfer mode streams carrying variable bit rate information can be transported through the existing switched digital video platform. Fortunately, software and hardware for this upgrade currently exist and are operating in full-service network configurations.
Once the economics of compression technology support the propagation into large-scale deployment, higher capacity can then be realized. For example, high-definition television, or HDTV, standards requiring approximately 20 Mbits/sec can also be supported. Sufficient digital bandwidth currently exists to support HDTV service requirements, and when affordable technology for digital encoding and decoding become available these transitions will occur.
The critical factor for distance-learning services is not bandwidth, but cost. To upgrade from 45 Mbits/sec to MPEG-2 or from MPEG-2 to HDTV technology involves simply changing encoding and decoding equipment and downloading new system software to support the increased content capacity.
Network architecture
The switched digital network architecture for distance learning comprises five basic network elements: supertrunk equipment, a host digital terminal, an optical network unit, a set-top box and a video administration module.
The supertrunk provides an economical method of transporting digital bit streams in a point-to-point and a point-to-multipoint network. A 1310-nanometer Fabry-Perot laser diode provides the supertrunk transmitter output. Transmitter cards support either one or four laser outputs. Thus, a single digitized signal can be transmitted onto four fibers without the associated splitting losses.
Regenerators can be used within the supertrunk link if distances exceed 45 kilometers or if the link budget is exhausted before reaching the termination point. Regenerators are available in one- and four-laser versions, as well. Moreover, the cost of the lasers is approximately the same as the cost of splitters. However, the multiple laser configuration supplies more optical power. Therefore, the multiple laser version of both the supertrunk transmitter and the regenerator is often used in place of single-laser versions combined with splitters.
The use of a fiber-optic passive optical network, in conjunction with an ample link budget allowance between the supertrunk and the host digital terminal, and the use of regenerators, allows a network reach of hundreds of miles from a single headend site. Multiple headend sites will likely evolve as communities of interest form.
Essentially, the supertrunk network links the local access area to the local interoffice facilities. This supplementary link increases bandwidth within a local video service hub area. Typically, three or four fibers are used between the central offices within the hub. Existing DS-3 facilities, operating at 44.736 Mbits/sec, can be implemented when connecting network hubs.
The optical fiber from the supertrunk terminates at a receiver card in the host digital terminal. The receiver uses a germanium avalanche photodiode and, in conjunction with the supertrunk transmitter, generally furnishes a 26.0-decibel link budget. Links between regenerators and the receiver support a 22.5-dB link budget. This optical link, therefore, allows many host digital terminals to be reached from one supertrunk transmitter at the hub center.
The host digital terminal includes a circuit switch capable of handling crossconnections for incoming video signals and directing them to the proper optical network unit. In fact, the intelligent host digital terminal switch can be provisioned to support channel switching based on user commands from the service site or from a preprogrammed database initiated from the video administrative module, via an X.25 connection.
In addition, the modular host digital terminal allows linecards to be added to accommodate growth. Therefore, additions to existing networks become incremental and can be deferred until service sites are integrated. Each host digital terminal linecard can accommodate as many as 24 video channels (based on a 45-Mbit/sec coding algorithm) over a fiber-optic link transmitting at 1.118 Gbits/sec. The output device is normally a Fabry-Perot laser diode operating at 1310 nm.
Delivered over singlemode fiber, the signals are transported to the optical network unit, where an indium gallium arsenide PIN, or positive intrinsic negative, diode receives the signal. The optical network unit is typically located either in or close to the service site. The bandwidth of the fiber trunk between the host digital terminal and the optical network unit delivers 24 unique channels to each unit simultaneously.
For a typical network deployment, where the initial service requirement is three to four receive channels per service site, services can be expanded when the demand for more channels arises. In this network, three fibers are needed to connect the central office containing the serving host digital terminal and the service site.
In addition, the linecard can support a passive optical link between the host digital terminal and the optical network unit. Typically, this link is not shared; the full bandwidth is dedicated to each service site. However, if fiber availability is an issue, fibers from the host digital terminal to the optical network unit can be shared by as many as four of these units. This sharing minimizes the fiber count required in the local access area. Those optical network units connected on a link with other units can share the channel capacity of that link, thus reducing both the fiber requirements and the cost for network electronics.
The video administration module--a computer-based network system controller--can be connected to 64 host digital terminals, and the channels can be controlled from a central location. Multiple communities of interest can coexist on one host digital terminal, and channel security is maintained by both the video administrative module and the host digital terminal.
Each set-top box incorporated into the network has a unique identification code. Services are provisioned based on the code and the box location in the network. Consequently, video information can be delivered only to a set-top box that has the proper identification and only if that code is reported from the proper location in the network. Therefore, the system provides a secure environment for mixing different types of video service applications, such as those in education and medicine.
Each environment is independently defined by the video administration module, and individual channels are allocated to each environment. This network control allows electronic devices to be cost-efficiently shared across services, while maintaining a secure transport platform.
Deployment mission
The advantages of the flexible, modular, switched digital video architecture become apparent as the network needs to be expanded and the service requirements become complex. For example, initial network service capabilities might include videoconferencing, data connections for remote testing and Internet access ports. If this network is deployed for distance-learning applications, other services can easily be added in increments.
Implementation of high-quality, full-motion video connections among all service sites lets users realize the advantages of a real-time distance-learning environment. High-quality conference links produce video images that attract and hold viewers` attention. Available data capacity can be delivered to the service site through a 56-kilobit-per-second interface.
Future enhancements that could also be added easily include integration of multimedia servers and high-bandwidth local area and wide area network connections. Designed for full-service applications, the switched digital network is compatible with common gateway protocols. This compatibility allows multiple servers to be accessed via user commands. Furthermore, complicated network navigation can be simplified using a graphical interface similar to proven Internet navigators. In this manner, distance-learning students could browse through the available applications and study at their own pace. Once linked to the full-service network, the students could also interact with the same applications at home, in the dormitory or in a computer room.
Even though multiple copies of a program can be distributed to achieve similar results, the advantages of having server-based training abound. For example, teacher instructions can be integrated into the application, updates can be controlled and centralized with copy protection, and centralized testing and scoring can be offered.
Although it seems complex, the switched digital video network provides cost savings due to sharing of electronics by multiple users, high integration employed in the user network elements and equipment operations savings. Other savings derive from the host digital terminal switching functionality being shared by 64 or more service sites.
In a typical distance-learning network deployment, one host digital terminal is shared among 4 to 6 optical network units. Therefore, the cost of the terminal is shared by 4 to 6 customers. The components not shared by multiple users are therefore designed for low cost. For example, optical network units and set-top boxes, even though they perform as intelligent devices, perform minimal processing. Thus, they can be designed to defer functionality, or processing, to the host digital terminal, where the cost is shared with other optical network units.
The switched digital network can also support standard telephony interfaces such as T1, at 1.554 Mbits/sec, and DS-3. In fact, dispersed network deployments can be linked via a telephony infrastructure because growth is not restricted to contiguous areas. For example, deployment clusters can be connected using an existing telephone company inter-office network. u
Earl W. Philmon is product line manager at Broadband Technologies Inc. in Research Triangle Park, NC.