الفهرس | Only 14 pages are availabe for public view |
Abstract Serial links have been dramatically important along the past ten years, partic- ularly with the emergence of high-performance communications requirements due to the exponential rise in necessary data rates in applications like cloud computing, data centres, and automotive. The increasing need for such high data rates in chip-to-chip communications paired with bandwidth-restricted channels places a heavy burden on the serial link’s equalisation capabilities. Both the transmitter and the receiver have the ability to simultaneously implement channel equalisation. While the Continuous Time Linear Equalizer (CTLE) and Decision Feedback Equalizer (DFE) are in charge of achieving the equalisation requirements on the receiver side, the Feed Forward Equalizer (FFE) is the main contributor to the equalisation requirements in the transmitter. In this thesis, an adaptive continuous-time linear equalizer is presented. The adaptation algorithm is configured using sign-sign least-mean square algorithm. This algorithms is characterized by the ease of hardware implementation, this is why it is preferred over the conventional least-mean square algorithm that may necessitate a high-resolution Analog-to-Digital Converter (ADC) and replaced by a one-bit sampler. This adaptive CTLE can support many channels with different impulse responses. The system presented can detect the characteristics of the channel and based upon that, it can modify the coefficients of the CTLE to achieve the suitable CTLE peaking value and correspondingly the suitable eye-opening. |