Experimental Comparison of B-Ppm and Nrz Coding Formats in A 1550nm Trans-Atmospheric Optical Communication Link under Deep Fading Conditions

B-PPM formatting for trans-atmospheric optical communication is compared experimentally to OOK (NRZ) at a single channel data rate of 1.25Gbps in deep fading conditions. Unlike low data rate transmission using Wary PPM formatting, high-speed B-PPM formatting does not benefit from the theoretical improvement that has been realized at low data rate. Although B-PPM can indeed benefit from a threshold set to near-zero, the high speed transmission precludes the implementation of a traditional Maximum Likelihood Detection circuit that compares the integrated power of each slot. At high speed, one has to rely on signal strength alone within the bit period which degrades the contrast between a "one" and a "zero." Moreover, the need for twice the bandwidth for B-PPM significantly limits available components such as APDs. More important, however, is the fact that during deep fades clock resynchronization dominates at high data rate. The primary question to be addressed is: Does B-PPM formatting really provide sufficient margin compared to NRZ to merit its use in deep fading atmospheric conditions? By building a special dual transceiver system, we have been able to propagate both B-PPM and NRZ formatted signals co-linearly on two C-band wavelengths centered close to 1550nm. Under field testing we measured the BER, including signal resynchronization, using special InGaAs, high-speed, multimode pigtailed, APD-based detectors in the receiver. The data were collected on fully instrumented horizontal paths of I km and 500m with C-n(2) [m(-2/3)] ranging from 10(-15) to 10(-13).