This thesis addresses the challenge of chromatic dispersion compensation (CDC) in coherent optical communication systems, focusing on high-speed, finite-length impulse response (FIR) filters. Chromatic dispersion, a significant limiting factor in long-distance, high data-rate fiber-optic networks, requires efficient filtering techniques to mitigate its impact on signal quality with low power dissipation. Although frequency-domain filtering has been well studied, most implementations have considered software for processor-based systems making them primarily suitable for off-line processing or limited data rates. High-speed real-time streaming implementations exist, but most only consider a small part of the available design space.

The primary contributions focus on design space exploration of frequency-domain FFT-based architectural approaches for CDC, considering both fully parallel and time-multiplexed configurations, implemented on FPGA and ASIC platforms. These designs, implemented for 16-QAM 60 GBaud systems, corresponding to 400 Gbps data rates, demonstrate an extended design space compared to earlier works. Especially, the FFT-length trade-off is considered, showing that estimates based on the number of multiplications are not fully relevant and that it is in general not possible to rely on only power-of-two-length FFTs. The results show that the extended design space provides better power efficiency. As part of the work, finite word-length effects are studied and novel blocks for performing overlap-save and overlap-add FFT-based filtering proposed.

Furthermore, a CDC filter design method for overlap-save processing that enhances filtering performance is proposed. Experimental results confirm that the proposed method significantly improves bit error rate (BER) performance without increasing latency or computational load when compared to optimal time-domain design.

The findings underscore the importance of optimizing both algorithmic and architectural aspects to optain energy efficient implementations of CDC filters in coherent optical communication systems. The presented works provide a scalable foundation for future high-speed, power-efficient static CDC.