Abstract

Bit-error-rate (BER) measurements using pseudorandom sequence generators and swept-frequency measurements using broadband laser diodes connected to scalar network analyzers are well-established characterization methods for the performance of optoelectronic receivers. However, while the BER test closely simulates the digital system environment of the receivers, it averages over both the system noise and amplitude and phase response. Therefore, it does not allow isolation of component-specific problems. Swept-frequency measurements using broadband laser diodes, on the other hand, are expensive as they require multiple broadband components (sweeper, laser, scalar analyzer). In many cases they require compensation for the laser-frequency response. Furthermore, they do not provide information on receiver sensitivity.

© 1988 Optical Society of America