Search

This paper describes the results of an experi ment where an FSO link was established between a ground terminal and an airborne terminal on a tethered aerostat. In the first part of the experiment, four 10 Gb/s WDM channels were transmitted from the aerostat to the ground terminal, demonstrating error -free performance over many 5 minute windows. In the second part of the experiment, both one and two WDM channels, each running at a native 40 Gb/s rate, were transmitted over the FSO link, demonstrating error free performance over short (30 second) timeframes, and bit error rates in the 10 -6 range over longer term operation. During all tests, a 1 Gb/s optical Ethernet channel was run on a separate WDM channel. The paper will describe the test con figuration as well as the results when running both single and multiple channels. The im pact of the overall system design used for this experiment on the link integrity will be analyzed and discussed. It was found that the spatial dyna mics of the aerostat as a function of time of day had an impact on the quality of the link – this will also be di scussed. Modeling of poten tial penalty sources will also be discussed. Keywords: Free-space optical communications, wavelength division multiplexing, optical amplifier, noise figure

A free-space optical (FSO) communication demonstration was conducted with JHU/APL and AOptix at the TCOM Test Facility in Elizabeth City, NC in May 2006. The primary test objective was to evaluate the performance of an FSO link from a fiber-tethered aerostat to a ground platform at effective data rates approaching 100 Gigabits/sec using wavelength division multiplexing (WDM) techniques. (Multiple optical channels operating near 1550 nm were modulated at data rates of 1, 10 and 40 Gbps). The test was conducted with a 38 meter aerostat raised to an altitude of 1 km and a ground platform located 1.2 km from the aerostat (limited by property boundary). Error free data transfers of 1.2 Terabits in 30 seconds at 40 Gbps were demonstrated. The total data transferred during the test was greater than 30 Terabits with an average BER of 10 -6 without any forward error correction (FEC) coding.