Abstract
Deep learning methods have been shown to be competitive solutions for modulation classification tasks, but suffer from being computationally expensive, limiting their use on embedded devices. We propose a new deep neural network architecture which employs known structures, depth-wise separable convolution and residual connections, as well as a compression methodology, which combined lead to a tiny and fast algorithm for modulation classification. Our compressed model won the first place in ITU's AI/ML in 5G Challenge 2021, achieving 61.73? compression over the challenge baseline and being over 2.6? better than the second best submission. The source code of this work is publicly available at github.com/ITU-AI- ML-in-5G-Challenge/ITU-ML5G-PS-007-BacalhauNet.

Abstract

Abstract
In this work, an in-depth analysis concerning the transmission performance of IEEE802.11g/n WiFi signals in a radio-over-fiber system is presented. Low-cost optical/electrical transceivers based on 850 nm vertical cavity surface emitting lasers (VCSELs) and PIN photodiodes are considered. System modelling includes the impact of noise generated in the optical path, such as relative intensity noise (RIN), shot noise, photodetector thermal noise, clipping and intermodulation distortion. Analytic results based on Volterra series analysis for the performance of the system in terms of SNR and EVM for several optical modulation index values are obtained. The theoretical analysis is also compared with experimental results. Among several conclusions, it is observed that the laser intermodulation distortion, clipping and RIN are the most relevant factors.

Abstract
We discuss recent developments in signal processing techniques for post-compensation of fiber dispersion and nonlinear distortion in both coherent and radio-over-fiber optical systems.

Abstract
This paper presents a performance evaluation of a multiband-orthogonal frequency division multiplexing (MBOFDM) ultra-wideband (UWB) signal transmission over two types of perfluorinated graded-index polymer optical fibers (PFGI-POFs) with diameters of 62.5 µm and 120 µm, using a lowcost optical transceiver. Experimental measurements of packet error rate (PER) and minimum transmitted powers to achieve the maximum allowed PER show that it is possible to have a viable transmission at data rates of 480 Mbps, 200 Mbps and 53.3Mbps over 100, 150 and 200 meters of PF-GI-POF, respectively, preceded by a 1 meter wireless link. ©2010 IEEE.

Abstract
We assess the performance of a coherent optical system employing a phase estimation algorithm combined with a least mean square decision-directed equalizer in a feedback configuration and compare it to the constant modulus algorithm with an independent carrier phase estimation approach for 16-quadratic-amplitude-modulation transmission. Both series and parallel implementations are discussed. Hardware issues regarding a recursive implementation of the estimation filter are also reported.