Fast Self-Adaptive Generic Digital Linearization for Analog Microwave Photonic Systems

Developing generic and self-adaptive methods to alleviate the nonlinearity issues remain of high importance for various microwave photonic (MWP) systems. Here, we propose and experimentally demonstrate a new digital linearization strategy to meet the requirements above. The key to our proposal lies on revealing that the dominant third-order nonlinear distortions (ND3s) have a ∼π phase deviation with respect to the fundamental signals in analog MWP systems (e.g., links, filters, mixers). By leveraging this peculiarity, the cost-function possessing a closed-form solution is achieved for the rapid self-adaptive acquirement of optimal linearization coefficients. The proposed digital approach is distinguished by the feature of fast self-adaptivity to various MWP systems with diverse input signals, with no need of prior system or signal parameters, complex training, or iterative optimization processes. Experimental investigations regarding the applications of our proposal in several typical MWP systems involving transmission link, filter and mixer are carried out. Achieved experimental results validate the feasibility and extensive applicability of our linearization strategy, in the sense of effective ND3s suppression for different MWP systems. Furthermore, the proposed linearization approach towards the simultaneous suppression of second and third-order distortions (ND2s
$\& $
ND3s) is evaluated, in a low-biasing Mach-Zehnder modulator based MWP link. The combination of self-adaptive ND2s
$\& $
ND3s mitigations and low-biasing technique experimentally shows a viable solution for highly required high-performance MWP links, in terms of high RF gain, low noise figure, wide operating bandwidth and large dynamic range.