Free-space optical wireless communication links for next-generation heterogeneous green access networks: Study of atmospheric effects

End-users data communication capacity needs have been increasing exponentially over the past several years, thus driving the existing optical network-architecture for even higher-capacity and reliable-links. This poses several challenges to minimize the connectivity gap in optical networks, which exist due to geographical constraints in optical-fiber deployment. In this regard, free-space-optics (FSO) communication has been suggested as a practical solution for lass-miles access, thus providing seamless connectivity to remote areas such as cities, villages located in desert regions, for instance, of Saudi Arabia, with minimum capital and operational-expenditure. Furthermore, FSO guarantees long-range, high-throughput, and secure-connection between remote areas and cities and a complementing solution to scale down the challenges of next-generation communication networks. Recently, the complex structured spatial-light-beam modes, such as Laguerre-Gaussian (LG), Hermite- Gaussian, Bessel-Gaussian, etc., have garnered attention for FSO communication. This allows increasing fixed FSO communication link capacity by exploiting space-division-multiplexing (SDM) where various spatial modes patterns to be utilized as information carriers are used to build M-ary pattern coding systems. However, the phase-fronts of these modes are severely affected by the outdoor FSO environment. In particular, weather conditions such as rain, fog, smoke, dust, etc., adversely scatter the optical signal, thus complicating the detection process and affecting the communication performance. In this context, we aim to investigate an FSO fog-channel by generating LG spatial mode beams in the unconventional L-band wavelength window rather than classical C-band, utilizing InAs/InP quantum-dash semiconductor laser diode (Qdash-LD). This stems from the fact that next-generation optical networks are considering engaging an L-band wavelength window for operation and wavelength-dependent optical signal scattering. We plan to generate 16-LG modes at 1610-nm using a spatial-light modulator and then transmit them over a controlled fog-chamber, which will emulate the effect of the fog environment and be captured by the camera at the receiver. The fog-channel characteristics will be evaluated in terms of visibility. Lastly, computer vision and machine-learning (ML) techniques will be engaged to identify spatial-modes under fog conditions correctly. Identification accuracies will be investigated using convolutional neural networks (CNN), and the sensed mode patterns and regression to predict the visibility of the foggy channel.