The measured worst case loss was 5.2dB. The sensitivity or noise equivalent power (NEP) had to be adjusted as well.
In addition to the parameters of the MOST150 calculation we included an equalizer penalty, which is the effective insertion loss of the equalizer, and the gain of the FEC into the power budget calculation.
The result can be seen in table 1. Please note that we converted the transmitted and received power into OMA in order to calculate an SNR and compare the two power budgets. It can be seen that the proposed changes enable a POF length of 10m for a net data rate of 1.25Gbit/s. This is mainly facilitated through the higher fiber-coupled OMA and the application of forward error correction in the receiver.
We expect to see improvements in several components. The next generation of the driver will improve on the OMA over temperature. The coupling at the receiver will improve through better alignment of PD and lens or even some kind of concentrator. The TIA NEP could be lowered through an optimization for the bandwidth demand of the next-gen MOST.
We have reported experimental results and demonstrated in a power budget calculation that the next generation of MOST can keep its current cheap optoelectronics and fiber. The compensation for the lower performance of the optoelectronics was achieved through the application of signal processing.
The gap to achieve the same performance as MOST150 is small and we expect that it will be closed soon. This gained margin can then result in a higher fiber length or in reduced complexity of the signal processing.
About the authors:
Norbert Weber, Ph.D has been head of the project group optical sensors and communications at the Fraunhofer Institute for Integrated Circuits, Germany since 1999. Currently his research interests include high-speed circuit design and optical communications, especially with polymer optical fibers.
Conrad Zerna is working in the project group optical sensors