Noise Analysis and Transmission Coding Optimization for High-speed Interconnection
Tianfang Chen,Min Miao,Bo Han,Zhensong Li,Kai Zhao,Xiaole Cui
DOI: https://doi.org/10.1109/icept.2018.8480557
2018-01-01
Abstract:Signal integrity analysis is a critical part in high-speed interconnect design. The common practices to account for noise effect like placement, routing, impedance matching and shielding are usually complex and costly. As data rate increases and layout size reduces, there are huge challenges in high-speed interconnect design. This paper discusses a new method to overcome the restriction from the perspective of tolerance to noise. By looking into the noise feature of the interconnect system from transient simulation, the relationship between design parameters and noise effect is established. Based on the noise model obtained above, the Low Density Parity Check (LDPC) code is applied. And the calculation of prior information when decoding is modified to make coding more applicable for the interconnection system. The bit error rate (BER) of transmission using this optimized LDPC coding is compared with the conventional one by simulation. And the result shows that the case with this new coding has excellent BER performance, demonstrating the feasibility of performance enhancement realized by this new method.