Digital Signal Processing Techniques Applied to Radio over Fiber Systems

The dissertation aims to analyze different Radio over Fiber systems for the front-haul applications. Particularly, analog radio over fiber (A-RoF) are simplest and suffer from nonlinearities, therefore, mitigating such nonlinearities through digital predistortion are studied. In particular for the long haul A-RoF links, direct digital predistortion technique (DPDT) is proposed which can be applied to reduce the impairments of A-RoF systems due to the combined effects of frequency chirp of the laser source and chromatic dispersion of the optical channel. Then, indirect learning architecture (ILA) based structures namely memory polynomial (MP), generalized memory polynomial (GMP) and decomposed vector rotation (DVR) models are employed to perform adaptive digital predistortion with low complexities. Distributed feedback (DFB) laser and vertical capacity surface emitting lasers (VCSELs) in combination with single mode/multi-mode fibers have been linearized with different quadrature amplitude modulation (QAM) formats for single and multichannel cases. Finally, a feedback adaptive DPD compensation is proposed. Then, there is still a possibility to exploit the other realizations of RoF namely digital radio over fiber (D-RoF) system where signal is digitized and transmits the digitized bit streams via digital optical communication links. The proposed solution is robust and immune to nonlinearities up-to 70 km of link length. Lastly, in light of disadvantages coming from A-RoF and D-RoF, it is still possible to take only the advantages from both methods and implement a more recent form knows as Sigma Delta Radio over Fiber (S-DRoF) system. Second Order Sigma Delta Modulator and Multi-stAge-noise-SHaping (MASH) based Sigma Delta Modulator are proposed. The workbench has been evaluated for 20 MHz LTE signal with 256 QAM modulation. Finally, The 6x2 GSa/s sigma delta modulators are realized on FPGA to show a real time demonstration of S-DRoF system. The demonstration shows that S-DRoF is a competitive competitor for 5G sub-6GHz band applications