The close of the 1980s saw many trials and applications of fiber in the loop or FITL. Most delivered traditional telephony (i.e., narrowband) services to the home, but a few provided both narrowband and entertainment video (i.e., broadband) services. During that time, the local-exchange carriers (LECs) also formulated their objectives and strategies for deploying FITL. As they gained a better understanding of the marketplace and the political realities of delivering video services, the LECs adopted an FITL-deployment strategy based on strict cost-effectiveness for the delivery of traditional telephone services. However, any fiber-optic access architecture they adopt has to support broadband services in the future. To address the cost and service challenges of FITL, the LECs and vendors must continually evaluate many alternative access architectures to identify potential advantages that can help fiber access achieve cost parity with copper access. Total system costs must be considered including electrical and optical components, powering system, fiber and cable components, and life-cycle costs (i.e., administration, maintenance, assignment, and provisioning operations).

Fiber-in-the-loop(FITL) is currently being deployed to deliver narrowband services such as voice-only telephony.[1] [2] It is essential to demonstrate that this fiber network, which will become the embedded base of the future, can be upgraded to support broadband services. Since these systems often provide bidirectional narrowband service for many users over a single fiber to reduce the cost/user, it may be difficult to provide this broadband upgrade without introducing large system impairments. Here, we demonstrate the upgrade of a fiber used in a current, low-speed FITL system[1] to provide broadband-ISDN service(622 Mb/s outbound and 155.5 Mb/s return) to each of four subscribers. Our system uses a Passive Optical Network(PON),[3] [4] and, because it simultaneously utilizes wavelength-division multiplexing(WDM), time-division multiplexing(TDM), and subcarrier multiplexing(SCM) on a single fiber, it operates with essentially no interchannel interference. We believe this is the first report of this type of multiplexing. In addition, we observe for the first time penalties associated with coherent interference of separate subcarrier lasers.[5]