Modelling quantum light-matter interactions in waveguide-QED with retardation and a time-delayed feedback: matrix product states versus a space-discretized waveguide model

TL;DR

This paper compares two advanced computational methods, matrix product states and space-discretized waveguide models, for simulating complex non-Markovian quantum light-matter interactions in waveguide-QED systems, highlighting their respective strengths and limitations.

Contribution

It introduces and compares two novel approaches for modeling non-Markovian waveguide-QED interactions, providing insights into their efficiency, scalability, and ease of implementation.

Findings

MPS scales better for multi-photon and long delay dynamics

SDW model effectively includes retardation and dissipation effects

Both methods accurately model non-Markovian regimes in waveguide QED

Abstract

We present two different methods for modelling non-Markovian quantum light-matter interactions in waveguide QED systems, using matrix product states (MPSs) and a space-discretized waveguide (SDW) model. After describing the general theory and implementation of both approaches, we compare and contrast these methods directly on three topical problems of interest in waveguide-QED, including (i) a two-level system (TLS) coupled to an infinite (one-dimensional) waveguide, (ii) a TLS coupled to a terminated waveguide with a time-delayed coherent feedback, and (iii) two spatially separated TLSs coupled within an infinite waveguide. Both approaches are shown to efficiently model multi-photon nonlinear dynamics in highly non-Markovian regimes, and we highlight the advantages and disadvantages of these methods for modelling waveguide QED interactions, including their implementation in Python, computational run times, and ease of conceptual understanding. We explore both vacuum dynamics as well as regimes of strong optical pumping, where a weak excitation approximation cannot be applied. The MPS approach scales better when modelling multi-photon dynamics and long delay times, and explicitly includes non-Markovian memory effects. In contrast, the SDW model accounts for non-Markovian effects through space discretization, and solves Markovian equations of motion, yet rigorously includes the effects of retardation. The SDW model, based on an extension of recent collisional pictures in quantum optics, is solved through quantum trajectory techniques, and can more easily add in additional dissipation processes, including off-chip decay and TLS pure dephasing. The impact of these processes is shown directly on feedback-induced population trapping and TLS entanglement between spatially separated TLSs.