Characterization and advanced communication techniques for free-space optical channels

Free-Space Optical (FSO) communication through the terrestrial atmospheric channel may offer many benefits in the wireless communications arena, like very high power efficiency; suitability for secure communications; absence of electromagnetic interference; and potentially very high bandwidth, to name a few. An optical beam propagating through the atmosphere is subject to optical turbulence. Optical turbulence is a random process that distorts the intensity and phase structure of a propagating optical beam and induces a varying signal at the receiver of an FSO communication link. This phenomenon (usually referred to as scintillation) degrades the performance of the FSO link by increasing the probability of error. In this dissertation we seek to characterize the effects of the scintillation-induced power fluctuations by determining the channel capacity of the optical link using numerical methods. We find that capacity decreases monotonically with increasing turbulence strength in weak turbulence conditions, but it is non-monotonic in strong turbulence conditions. We show that low-density parity-check (LDPC) codes provide strong error control capabilities in this channel if a perfect interleaver can be used. Multiple transmit optical beams can also be used to reduce scintillation. We characterize the spatial correlation of the atmospheric optical channel and determine a scintillation model for the multiple-bearn scheme. With this model we are able to predict the effective reduction in scintillation as a function of the system design parameters. A Multi-channel FSO communications system based on orbital angular momentum (OAM)-carrying beams is studied. We numerically analyze the effects of atmospheric turbulence on the system and find that turbulence induces attenuation and crosstalk among OAM channels. Based on a model in which the constituent channels are binary symmetric and crosstalk is a Gaussian noise source, we find optimal sets of OAM states at each turbulence condition studied, and determine the aggregate capacity of the multi-channel system at those conditions. At very high data rates the FSO channel shows some inter-symbol interference. We address the problem of joint sequence detection in partial-response (PR) channels and decoding of LDPC codes. We model the PR channel and the LDPC code as a combined inference problem. We derive the belief propagation equations that allow the simultaneous detection and decoding of a LDPC codeword in a PR channel.