Digital Signal Processing Algorithms for Long-Reach DSL Deployment and Speech and Audio Enhancement

Korte inhoud:The PhD explores advancements in digital signal processing (DSP) in two core areas: broadband digital communication, specifically digital subscriber line (DSL) technology, and audio and speech processing, focusing on subband system identification. These domains, though distinct in application, share a common reliance on narrowband signal processing through filter bank techniques In broadband communication, the research addresses key challenges in next-generation DSL networks, particularly MGfast—a multi-gigabit, ultra-broadband DSL standard. Central to this work is the optimization of resource-sharing strategies for point-to-multipoint (P2MP) distribution, crucial for sustaining high data rates and quality of service within dynamic, high-bandwidth networks. Additionally, the research develops channel shortening techniques critical for long-reach DSL networks, such as LR-VDSL2 and LR-G.fast, to support the International Telecommunication Union's (ITU-T) goal of a unified DSL standard. In audio and speech processing, the work addresses accurate subband system identification in acoustic channels with long impulse responses. A novel weighted overlap-add (WOLA) filter bank design is proposed to enhance modeling accuracy while effectively managing computational complexity. Overall, the research presented in this PhD enhances DSP applications by optimizing resource-sharing and channel shortening in DSL networks and advancing subband adaptive filtering techniques for audio and speech processing. These contributions underscore the critical role of DSP in driving innovation across digital communication and audio processing technologies.

Jaar van publicatie:2024

Toegankelijkheid:Open