| Summary: | 碩士 === 國立交通大學 === 光電系統研究所 === 102 === The Radio-over-Fiber system has been employed to transmit signals for a long time, but there are still many disadvantages with analog radio-over-fiber (ARoF) systems, such as tolerance to non-linearity effects, noise and dynamic range of the signal with increase fiber length. Recently, digital radio-over-fiber (DRoF) has been proposed as better candidate in order to solve these problems. In DRoF system, only two-level amplitudes for digital signal are used to send an RF signal, hence fundamentally, DRoF is robust to noise and any non-linear distortions caused by microwave and optical components. However, the data rate of DRoF systems depend on the sampling frequency and resolution bit of an analog-to-digital (ADC). In this thesis, we employed a band pass sampling frequency in order to reduce on the total bit-rate of the system and relax the bandwidth requirement of the components used in DRoF systems. In order to understand the fundamental benefit of DRoF system, some simulations and experiments has been carried out in our work. Simulation results show that a pre-aliasing band-pass filter (BPF) is needed for avoiding noise aliasing after sampling the radio frequency (RF) signal. In addition, it has been observed that the required bit-resolution in the ADC depends on the peak to average power ratio (PAPR) and signal amplitude input to the ADC. For multi-services or multi-band RF signals, an efficient and low complexity algorithm has been employed for finding a minimum sampling rate that avoid overlapping different RF signals after sampling. Experimental results show DRoF for OFDM modulated systems show a better performance compared to ARoF system over 50-km fiber transmission.
|
|---|
