Completely-positive non-signalling non-Markovian dynamics

This paper defines and characterizes completely positive, non-signalling non-Markovian quantum dynamics as an integro-differential equation extending the Lindblad formalism, enabling the exact calculation of multi-time correlations and the derivation of memory-encoded spectral features without relying on the regression theorem.

The Gist

Imagine you are trying to predict the weather. In a simple, "Markovian" world, the weather tomorrow depends only on the weather today. If you know it's raining right now, you don't need to know if it rained last week or last year to make a good guess. The past is forgotten; only the present matters.

But the real world is rarely that simple. Often, the weather today depends on a long history of storms, pressure systems, and temperature shifts from the past. This is called a non-Markovian process. It's like trying to predict a person's mood: their current happiness might depend not just on what happened five minutes ago, but on a conversation they had three days ago, or a bad night's sleep from a week prior.

This paper by Serhii Kryhin and Vivishek Sudhir tackles a very difficult problem in quantum physics: How do we mathematically describe a quantum system (like an atom or a qubit) when its future depends on its entire past, not just the present?

Here is the breakdown of their work using simple analogies:

1. The Problem: The "Amnesia" of Standard Physics

For decades, physicists have used a famous set of rules (called the GKSL equation) to describe how quantum systems change. Think of these rules as a "perfect amnesia" machine. They assume that as soon as a quantum system interacts with its environment (like air molecules or heat), the system instantly forgets everything that happened before.

This works great for simple situations. But in the real world, environments often have "memory." A quantum system might be interacting with a noisy background that remembers its past interactions. When this happens, the standard "amnesia" rules break down. Previous attempts to fix this were often messy, made up rules, or only worked for specific types of noise. There was no single, clean rulebook for this "memory-filled" quantum world.

2. The Solution: A New Rulebook with Two Guardrails

The authors created a new, rigorous mathematical framework to describe these memory-filled quantum systems. To ensure their new rules make physical sense, they built them on two strict "guardrails":

• Guardrail 1: Complete Positivity (The "No Negative Probabilities" Rule): In quantum mechanics, probabilities must always be positive numbers (0% to 100%). You can't have a -10% chance of something happening. Their new math guarantees that no matter how complex the history is, the system will never produce impossible, negative probabilities.
• Guardrail 2: Non-Signalling (The "No Telepathy" Rule): This ensures that the system's behavior doesn't allow for impossible communication. For example, if you have a bag of marbles, the way you mix them shouldn't magically change the color of a marble in a bag across the room instantly. Their math ensures that the system's history doesn't allow for "spooky" shortcuts that violate the laws of physics.

3. The Result: The "Memory Equation"

By combining these two guardrails, the authors derived a new equation.

• The Old Way: The standard equation is like a car driving forward where the steering wheel only reacts to the road right now.
• The New Way: Their new equation is like a car with a "ghost driver" who remembers every turn you've ever taken. The equation has two parts:
  1. The Present: It reacts to what is happening right now (just like the old rules).
  2. The Memory Integral: It adds a "memory term." This is a mathematical sum that looks back at every previous state of the system and weighs how much that past state influences the present.

This allows them to describe quantum systems exposed to any type of noise (as long as the noise isn't infinitely wild), without having to make up approximations or guesswork.

4. Measuring the Unmeasurable

Usually, to predict what happens when you measure a quantum system, physicists use a "regression theorem" (a shortcut that links past measurements to future ones). But this shortcut breaks down when the system has memory.

The authors showed how to calculate measurement results without this broken shortcut. They developed a method to track the "state" of the system even after a measurement happens.

• Analogy: Imagine you are watching a movie. In a normal movie, if you pause it, the story stops. In this new framework, even if you pause the movie (measure the system), the "story" (the quantum state) continues to evolve in a way that remembers the pause. Their math tells you exactly how to calculate the next scene based on that pause.

5. The Proof: The "Mollow Triplet" with a Twist

To prove their theory works, they applied it to a classic physics experiment: a two-level atom (like a simple light switch) being driven by a laser while sitting in a noisy environment.

• The Standard Result: In a normal, memory-less world, the light emitted by this atom forms a pattern called the "Mollow Triplet" (three distinct peaks of light).
• The New Result: When they applied their new "memory" equation, the three peaks were still there, but they changed shape. The "width" of each peak (how blurry it is) became dependent on the frequency of the light.
• The Meaning: This "blurriness" is a direct fingerprint of the environment's memory. The math successfully encoded the history of the noise into the shape of the light, proving that their framework can capture the "ghosts of the past" in a quantum system.

Summary

This paper provides the first clean, mathematically rigorous "rulebook" for quantum systems that remember their past. It ensures that these systems behave logically (no negative probabilities, no magic communication) and gives physicists a way to predict what these systems will do and how they will look when measured, even when they are surrounded by complex, noisy environments that refuse to let them forget.

Technical Summary

Technical Summary: Completely-Positive Non-Signalling Non-Markovian Dynamics

Problem Statement
While Markovian models offer simplicity, non-Markovian processes—where the current state depends on the entire history of past states—are ubiquitous in classical and quantum physics, appearing in phenomena ranging from "1/f" noise and glassy dynamics to precision sensing in mechanical oscillators. Existing approaches to describe non-Markovian quantum dynamics often rely on ad-hoc approximations, are specific to certain models, or are purely phenomenological. Crucially, there is a lack of a precise, general characterization of non-Markovian quantum processes analogous to the Gorini-Kossakowski-Sudarshan-Lindblad (GKSL) equation for Markovian processes. Specifically, a rigorous framework that defines the quantum state of a system at all times under general noise conditions, while ensuring complete positivity (CP) and non-signalling (NS), has been missing.

Methodology
The authors define a class of non-Markovian quantum processes on the space of density matrices and characterize their structure under two fundamental physical axioms:

1. Complete Positivity (CP): The maps must produce positive states even when the system is entangled with an ancilla.
2. Non-Signalling (NS): The maps must be convex-linear, ensuring that statistically equivalent ensembles cannot be distinguished, thereby preventing faster-than-light communication.

The derivation proceeds in two stages:

• Discrete-Time Formulation: The authors consider a sequence of states $\hat{\varrho}_0, \dots, \hat{\varrho}_n$ where the next state $\hat{\varrho}_{n+1}$ is determined by a multi-variable map $K_{n+1}$ acting on all previous states. They demonstrate that any CP map can be decomposed into homogeneous components. By imposing the NS condition (convex-linearity), they constrain these components to be linear in exactly one argument. This leads to a Kraus-like representation for the multi-variable map:
  $$\hat{\varrho}_{n+1} = \sum_{m=0}^n \sum_{\alpha} V_{\alpha nm}^\dagger \hat{\varrho}_m V_{\alpha nm}$$
  where the operators $V_{\alpha nm}$ satisfy specific trace-preserving conditions.
• Continuous-Time Limit: By taking the limit $\Delta t \to 0$ and analyzing the scaling of the operators $V_{\alpha nm}$, the authors derive a continuous-time integro-differential equation. The scaling requires that terms acting on the current state ($m=n$) scale as $\sqrt{\Delta t}$ (standard Markovian noise), while terms acting on past states ($m < n$) scale linearly with $\Delta t$ to ensure a finite contribution in the limit.

Key Contributions and Results

1. Non-Markovian Master Equation: The central result is a generalized master equation for the density matrix $\hat{\varrho}(t)$:
   $$\frac{\partial \hat{\varrho}}{\partial t} = (\mathcal{L}\hat{\varrho})(t) + \int_{t_0}^t dt' (\mathcal{R}\hat{\varrho})(t, t')$$

   • The term $(\mathcal{L}\hat{\varrho})(t)$ is the standard GKSL generator, describing Markovian dynamics.
   • The term $(\mathcal{R}\hat{\varrho})(t, t')$ is the non-Markovian memory kernel. It takes the form of a GKSL generator but acts asymmetrically: the "jump" term acts on the retarded state $\hat{\varrho}(t')$, while the anti-commutator term acts on the current state $\hat{\varrho}(t)$. This specific structure is a direct consequence of the CP and NS axioms.
   • The equation is valid for any noise with an integrable power spectral density, without further approximations.

2. Formalism for Multi-Time Correlations: The authors establish a method to evaluate multi-time correlations of measurement outcomes without relying on a regression theorem (which fails in non-Markovian regimes due to the lack of a composition law for the dynamical map). They define a "conditional state" $\hat{\varrho}_{J,t}(t')$ that evolves according to the master equation with specific initial conditions at the time of measurement. This allows the calculation of correlation functions $E[I(t+\tau)I(t)]$ for both instantaneous and continuous-time measurements.

3. Application to a Driven Two-Level System (TLS): The formalism is applied to a driven TLS coupled to a bosonic bath without the Markov approximation.

   • The bath correlation function $S_E(t)$ determines the memory kernel.
   • The authors derive the emission spectrum of the system. In the limit of small detuning and large pumping, the spectrum retains the Mollow triplet structure.
   • Key Finding: Unlike the standard Markovian Mollow triplet, the linewidths of the three peaks in the non-Markovian case become frequency-dependent. This frequency dependence, governed by the Fourier transform of the bath correlation function, explicitly encodes the memory of the bath.

Significance and Claims
The paper claims to provide a "rigorous yet transparent description of the quantum state of non-Markovian systems." Its primary significance lies in:

• Generalization of GKSL: It offers the first general characterization of non-Markovian dynamics that parallels the GKSL theorem for Markovian processes, valid for arbitrary integrable power spectral densities.
• State Definition: It enables the consistent assignment of a legitimate quantum state to a system at all times, even in the presence of memory effects, resolving ambiguities in phenomenological approaches.
• Operational Predictions: It provides a complete framework for extracting operationally meaningful predictions, such as measurement statistics and emission spectra, from the quantum state, bypassing the need for regression theorems.
• Physical Insight: The application to the driven TLS demonstrates that memory effects manifest physically as frequency-dependent linewidths in the emission spectrum, offering a concrete signature of non-Markovianity.

The authors note that extending this formalism to non-integrable power spectral densities (such as $1/f$ noise) and developing a full quantum trajectory description for state estimation and feedback control remain open problems.