Unbiased finite-memory digital phase-locked loop

Abstract

Digital phase-locked loops (DPLLs) have been commonly used to estimate phase information. However, they exhibit poor performance or, occasionally, a divergence phenomenon, if noise information is incorrect or if there are quantization effects. To overcome the weaknesses of existing DPLLs, we propose a new DPLL with a finite-memory structure called the unbiased finite-memory DPLL (UFMDPLL). The UFMDPLL is independent of noise covariance information, and it shows intrinsic robustness properties against incorrect noise information and quantization effects due to the finite-memory structure. Through numerical simulations, we show that the proposed DPLL is more robust against incorrect noise information and quantization effects than the conventional DPLLs are.