Search

We present a Bayesian algorithm to identify generators of open quantum system dynamics, described by a Lindblad master equation, that are compatible with measured experimental data. The algorithm, based on a Markov Chain Monte Carlo approach, assumes the energy levels of the system are known and outputs a ranked list of interpretable master equation models that produce predicted measurement traces that closely match experimental data. We benchmark our algorithm on quantum optics experiments performed on single and pairs of quantum emitters. The latter case opens the possibility of cooperative emission effects and additional complexity due to the possible interplay between photon and phonon influences on the dynamics. Our algorithm retrieves various minimal models that are consistent with the experimental data, and which can provide a closer fit to measured data than previously suggested and physically expected approximate models. Our results represent an important step towards automated systems characterisation with an approach that is capable of working with diverse and tomographically incomplete input data. This may help with the development of theoretical models for unknown quantum systems as well as providing scientists with alternative interpretations of the data that they might not have originally envisioned and enabling them to challenge their original hypotheses., Comment: 11 pages, 5 figures + Supplementary Information. Comments and suggestions are welcome!

We report the measurement of collective emission from a pair of independently tuneable InAs quantum dots embedded in a nanophotonic waveguide. A split diode structure allows independent electrical control of the quantum dot transition energies over a wide range with minimal loss in waveguide coupling efficiency. We utilise this to systematically map out the transition from collective to independent emission. We perform both lifetime as well as Hanbury Brown-Twiss measurements on the device, observing anti-dips in the photon coincidences indicating collective emission while at the same time observing a drop in lifetime around zero detuning, indicating superradiant behaviour. Performing both measurement types allows us to investigate detuning regions which show both superradiant rate enhancement and inter-emitter coherence, as well as regions in which correlations persist in the absence of rate enhancement.

Cooperative optical effects are enabled and controlled by interactions between molecular dipoles, meaning that their mutual orientation is of paramount importance to, for example, superabsorbing light-harvesting antennas. Here we show how to move beyond the possibilities of simple geometric tailoring, demonstrating how a metallic sphere placed within a ring of parallel dipoles engineers an effective Hamiltonian that generates "guide-sliding" states within the ring system. This allows steady-state superabsorption in noisy room temperature environments, outperforming previous designs while being significantly simpler to implement. As exemplified by this showcase, our approach represents a powerful design paradigm for tailoring cooperative light-matter effects in molecular structures that extends beyond superabsorbing systems, to a huge array of quantum energy transport systems.

Quantum networks will provide unconditional security for communication, computation and distributed sensing tasks. We report on an experimental demonstration of private parameter estimation, which allows a global phase to be evaluated without revealing the constituent local phase values. This is achieved by sharing a Greenberger-Horne-Zeilinger (GHZ) state among three users who first verify the shared state before performing the sensing task. We implement the verification protocol, based on stabilizer measurements, and measure an average failure rate of 0.038(5) which we use to establish the security and privacy parameters. We validate the privacy conditions established by the protocol by evaluating the quantum Fisher information of the experimentally prepared GHZ states., Comment: 14 pages, 8 figures. Updated

We derive the transition rates, dephasing rates, and Lamb shifts for a system consisting of many molecules collectively coupled to a resonant cavity mode. Using a variational polaron master equation, we show that strong vibrational interactions inherent to molecules give rise to multi phonon processes and suppress the light--matter coupling. In the strong light--matter coupling limit, multi-phonon contributions to the transition and dephasing rates strongly dominate over single phonon contributions for typical molecular parameters. This leads to novel dependencies of the rates and spectral line widths on the number of molecules in the cavity. We also find that vibrational Lamb shifts can substantially modify the polariton energies in the strong light--matter coupling limit., Comment: 15 + 6 pages, 6 + 1 figures

Measurement and analysis of high energetic particles for scientific, medical or industrial applications is a complex procedure, requiring the design of sophisticated detector and data processing systems. The development of adaptive and differentiable software pipelines using a combination of conventional and machine learning algorithms is therefore getting ever more important to optimize and operate the system efficiently while maintaining end-to-end (E2E) differentiability. We propose for the application of charged particle tracking an E2E differentiable decision-focused learning scheme using graph neural networks with combinatorial components solving a linear assignment problem for each detector layer. We demonstrate empirically that including differentiable variations of discrete assignment operations allows for efficient network optimization, working better or on par with approaches that lack E2E differentiability. In additional studies, we dive deeper into the optimization process and provide further insights from a loss landscape perspective. We demonstrate that while both methods converge into similar performing, globally well-connected regions, they suffer under substantial predictive instability across initialization and optimization methods, which can have unpredictable consequences on the performance of downstream tasks such as image reconstruction. We also point out a dependency between the interpolation factor of the gradient estimator and the prediction stability of the model, suggesting the choice of sufficiently small values. Given the strong global connectivity of learned solutions and the excellent training performance, we argue that E2E differentiability provides, besides the general availability of gradient information, an important tool for robust particle tracking to mitigate prediction instabilities by favoring solutions that perform well on downstream tasks.

We have applied an operator-overloading forward-mode algorithmic differentiation tool to the Monte-Carlo particle simulation toolkit Geant4. Our differentiated version of Geant4 allows computing mean pathwise derivatives of user-defined outputs of Geant4 applications with respect to user-defined inputs. This constitutes a major step towards enabling gradient-based optimization techniques in high-energy physics, as well as other application domains of Geant4. This is a preliminary report on the technical aspects of applying operator-overloading AD to Geant4, as well as a first analysis of some results obtained by our differentiated Geant4 prototype. We plan to follow up with a more refined analysis.

We describe a general-purpose computational toolkit for simulating open quantum systems, which provides numerically exact solutions for composites of zero-dimensional quantum systems that may be strongly coupled to multiple, quite general non-Markovian environments. It is based on process tensor matrix product operators (PT-MPOs), which efficiently encapsulate environment influences. The code features implementations of several PT-MPO algorithms, in particular, Automated Compression of Environments (ACE) for general environments comprised of independent modes as well as schemes for generalized spin boson models. The latter includes a divide-and-conquer scheme for periodic PT-MPOs, which enable million time step simulations for realistic models. PT-MPOs can be precalculated and reused for efficiently probing different time-dependent system Hamiltonians. They can also be stacked together and combined to provide numerically complete solutions of small networks of open quantum systems. The code is written in C++ and is fully controllable by configuration files, for which we have developed a versatile and compact human-readable format.

The assembly and co-evolution of supermassive black holes (SMBH) and their host galaxy stellar population is a key open questions in galaxy evolution. Stellar mass ($M_\star$) and star formation rate (SFR), are inferred by modeling the spectral energy distribution (SED). For galaxies triggering SMBH activity, the active galactic nucleus (AGN) contaminates the light at all wavelengths, hampering the inference of galaxy parameters. Incomplete AGN templates can lead to systematic overestimates of the stellar mass, biasing our understanding of AGN-galaxy co-evolution. This challenge has gained further impetus with the advent of sensitive wide-area surveys with millions of luminous AGN, including by eROSITA, Euclid and LSST. We aim to estimate the accuracy and bias of AGN host galaxy parameters and improve upon existing techniques. This work makes two contributions: 1) a new SED fitting code, GRAHSP, with a flexible, empirically motivated AGN model including a power law continuum emission lines, a FeII forest and a flexible infrared torus. We verify that our model reproduces published X-ray to infrared SEDs of AGN to better than 20\% accuracy. A fully Bayesian fit with nested sampling includes uncertainties in the model and the data, making the inference highly robust. 2) we created a benchmark photometric dataset where pure quasars are merged with non-AGN pure galaxies into a hybrid (Chimera) object but with known galaxy and AGN properties. Comparing the true and retrieved $M_\star$, SFR and AGN luminosities shows that previous codes systematically over-estimate $M_\star$ and SFR by 0.5 dex with a wide scatter of 0.7 dex, at AGN luminosities above 10^44 erg/s. In contrast, GRAHSP shows no bias on $M_\star$ and SFR. GRAHSP also estimates more realistic uncertainties. GRAHSP enables characterization of the environmental conditions conducive to black hole growth. (abridged), Comment: accepted in A&A

The algorithm "automated compression of environments" (ACE) [Nat. Phys. 18, 662 (2022)] provides a versatile way of simulating an extremely broad class of open quantum systems. This is achieved by encapsulating the influence of the environment, which is determined by the interaction Hamiltonian(s) and initial states, into compact process tensor matrix product operator (PT-MPO) representations. The generality of the ACE method comes at high numerical cost. Here, we demonstrate that orders-of-magnitude improvement of ACE is possible by changing the order of PT-MPO contraction from a sequential to a tree-like scheme. The problem of combining two partial PT-MPOs with large inner bonds is solved by a preselection approach. The drawbacks of the preselection approach are that the MPO compression is suboptimal and that it is more prone to error accumulation than sequential combination and compression. We therefore also identify strategies to mitigate these disadvantages by fine-tuning compression parameters. This results in a scheme that is similar in compression efficiency and accuracy to the original ACE algorithm, yet is significantly faster. Our numerical experiments reach similar conclusions for bosonic and fermionic test cases, suggesting that our findings are characteristic of the combination of PT-MPOs more generally.

Among the well-known methods to approximate derivatives of expectancies computed by Monte-Carlo simulations, averages of pathwise derivatives are often the easiest one to apply. Computing them via algorithmic differentiation typically does not require major manual analysis and rewriting of the code, even for very complex programs like simulations of particle-detector interactions in high-energy physics. However, the pathwise derivative estimator can be biased if there are discontinuities in the program, which may diminish its value for applications. This work integrates algorithmic differentiation into the electromagnetic shower simulation code HepEmShow based on G4HepEm, allowing us to study how well pathwise derivatives approximate derivatives of energy depositions in a sampling calorimeter with respect to parameters of the beam and geometry. We found that when multiple scattering is disabled in the simulation, means of pathwise derivatives converge quickly to their expected values, and these are close to the actual derivatives of the energy deposition. Additionally, we demonstrate the applicability of this novel gradient estimator for stochastic gradient-based optimization in a model example., Comment: 12 pages, 11 figures, 2 tables

In shared-memory parallel automatic differentiation, inputs that are shared among simultaneous thread-local preaccumulations lead to data races if Jacobians are accumulated with a single, shared vector of adjoint variables. In this work, we discuss the benefits and tradeoffs of re-enabling such preaccumulations by a transition to suitable local adjoints. We propose different vector- and map-based approaches for storing local adjoint variables and analyze them with respect to memory consumption, memory allocation, and adjoint variable access times in the context of simultaneous preaccumulations in multiple threads. We implement the approaches in CoDiPack and benchmark them in parallel discrete adjoint computations in the multiphysics simulation suite SU2., Comment: 12 pages, 5 figures. Updated and extended all parts of the paper

The open-source multiphysics suite SU2 features discrete adjoints by means of operator overloading automatic differentiation (AD). While both primal and discrete adjoint solvers support MPI parallelism, hybrid parallelism using both MPI and OpenMP has only been introduced for the primal solvers so far. In this work, we enable hybrid parallel discrete adjoint solvers. Coupling SU2 with OpDiLib, an add-on for operator overloading AD tools that extends AD to OpenMP parallelism, marks a key step in this endeavour. We identify the affected parts of SU2's advanced AD workflow and discuss the required changes and their tradeoffs. Detailed performance studies compare MPI parallel and hybrid parallel discrete adjoints in terms of memory and runtime and unveil key performance characteristics. We showcase the effectiveness of performance optimizations and highlight perspectives for future improvements. At the same time, this study demonstrates the applicability of OpDiLib in a large code base and its scalability on large test cases, providing valuable insights for future applications both within and beyond SU2., Comment: 28 pages, 9 figures, 2 listings; new layout, revised section structure, polishing and small updates

Process tensor matrix product operators (PT-MPOs) enable numerically exact simulations for an unprecedentedly broad range of open quantum systems. By representing environment influences in MPO form, they can be efficiently compressed using established algorithms. The dimensions of inner bonds of the compressed PT-MPO may be viewed as an indicator of the complexity of the environment. Here, we show that the inner bonds themselves, not only their dimensions, have a concrete physical meaning: They represent the subspace of the full environment Liouville space which hosts environment excitations that may influence the subsequent open quantum systems dynamics the most. This connection can be expressed in terms of lossy linear transformations, whose pseudoinverses facilitate the extraction of environment observables. We demonstrate this by extracting the environment spin of a central spin problem, the current through a quantum system coupled to two leads, the number of photons emitted from quantum emitters into a structured environment, and the distribution of the total absorbed energy in a driven non-Markovian quantum system into system, environment, and interaction energy terms. Numerical tests further indicate that different PT-MPO algorithms compress environments to similar subspaces. Thus, the physical interpretation of inner bonds of PT-MPOs both provides a conceptional understanding and it enables new practical applications.

Cooperative effects of multiple quantum emitters are characterized by transitions via delocalized collective states with altered emission properties due to the existence of inter-emitter coherences. When realized with excitonic condensed matter nanostructures, these effects are significantly affected by the presence of strong emitter-phonon coupling, which leads to the formation of polarons. We show that, while for single-emitter emission into free space this formation has no impact on its radiative lifetime, the same is not true for superradiant emission. Considering the case of two indistinguishable quantum emitters, we analyse how polaron dressing affects collective photon emission by mixing bright and dark Dicke states. Our numerical simulations show that this mixing crucially depends on the circumstances of the excitation of the system: Depending on the pulse length of an exciting laser, one can choose to either prepare polaronic Dicke states, or bare electronic Dicke states, changing the superradiant decay characteristics of the system. Additionally, we derive analytic expressions for these limiting cases, which match the results of numerically exact calculations.

Differentiable models of physical systems provide a powerful platform for gradient-based algorithms, with particular impact on parameter estimation and optimal control. Quantum systems present a particular challenge for such characterisation and control, owing to their inherently stochastic nature and sensitivity to environmental parameters. To address this challenge, we present a versatile differentiable quantum master equation solver, and incorporate this solver into a framework for device characterisation. Our approach utilises gradient-based optimisation and Bayesian inference to provide estimates and uncertainties in quantum device parameters. To showcase our approach, we consider steady state charge transport through electrostatically defined quantum dots. Using simulated data, we demonstrate efficient estimation of parameters for a single quantum dot, and model selection as well as the capability of our solver to compute time evolution for a double quantum dot system. Our differentiable solver stands to widen the impact of physics-aware machine learning algorithms on quantum devices for characterisation and control.

The eROSITA instrument aboard the Spectrum Roentgen Gamma (SRG) satellite has performed its first all-sky survey between December 2019 and June 2020. This paper presents the resulting hard X-ray (2.3-5 keV) sample, the first created from an all-sky imaging survey in the 2-8 keV band, for sources within western galactic sky. The 5466 hard X-ray selected sources detected with eROSITA are presented and discussed. The Bayesian statistics-based code NWAY is used to identify the counterparts for the X-ray sources. These sources are classified based on their multiwavelength properties, and the literature is searched to identify spectroscopic redshifts, which further inform the source classification. A total of 2547 sources are found to have good-quality counterparts, and 111 of these are detected only in the hard band. Comparing with other hard X-ray selected surveys, the eROSITA hard sample covers a larger redshift range and probes dimmer sources, providing a complementary and expanded sample as compared to Swift-BAT. Examining the column density distribution of missed and detected eROSITA sources present in the follow-up catalog of Swift BAT 70 month sources, it is demonstrated that eROSITA can detect obscured sources with column densities $>10^{24}$ cm$^{-2}$, but that the completeness drops rapidly after $10^{23}$ cm$^{-2}$. A sample of hard-only sources, many of which are likely to be heavily obscured AGN, is also presented and discussed. X-ray spectral fitting reveals that these sources have extremely faint soft X-ray emission and their optical images suggest that they are found in more edge-on galaxies with lower b/a. The resulting X-ray catalog is demonstrated to be a powerful tool for understanding AGN, in particular heavily obscured AGN found in the hard-only sample., Comment: 20 pages, 21 figures. Submitted to A&A. Associated with eROSITA data release. Associated catalogs will be released after journal publication

The challenging propagation environment, combined with the hardware limitations of mmWave systems, gives rise to the need for accurate initial access beam alignment strategies with low latency and high achievable beamforming gain. Much of the recent work in this area either focuses on one-sided beam alignment, or, joint beam alignment methods where both sides of the link perform a sequence of fixed channel probing steps. Codebook-based non-adaptive beam alignment schemes have the potential to allow multiple user equipment (UE) to perform initial access beam alignment in parallel whereas adaptive schemes are favourable in achievable beamforming gain. This work introduces a novel deep learning based joint beam alignment scheme that aims to combine the benefits of adaptive, codebook-free beam alignment at the UE side with the advantages of a codebook-sweep based scheme at the base station. The proposed end-to-end trainable scheme is compatible with current cellular standard signaling and can be readily integrated into the standard without requiring significant changes to it. Extensive simulations demonstrate superior performance of the proposed approach over purely codebook-based ones.

The field of nanoscale magnetic resonance imaging (NanoMRI) was started 30 years ago. It was motivated by the desire to image single molecules and molecular assemblies, such as proteins and virus particles, with near-atomic spatial resolution and on a length scale of 100 nm. Over the years, the NanoMRI field has also expanded to include the goal of useful high-resolution nuclear magnetic resonance (NMR) spectroscopy of molecules under ambient conditions, including samples up to the micron-scale. The realization of these goals requires the development of spin detection techniques that are many orders of magnitude more sensitive than conventional NMR and MRI, capable of detecting and controlling nanoscale ensembles of spins. Over the years, a number of different technical approaches to NanoMRI have emerged, each possessing a distinct set of capabilities for basic and applied areas of science. The goal of this roadmap article is to report the current state of the art in NanoMRI technologies, outline the areas where they are poised to have impact, identify the challenges that lie ahead, and propose methods to meet these challenges. This roadmap also shows how developments in NanoMRI techniques can lead to breakthroughs in emerging quantum science and technology applications.

We present the performance of a full-length prototype of the ALICE Forward Calorimeter (FoCal). The detector is composed of a silicon-tungsten electromagnetic sampling calorimeter with longitudinal and transverse segmentation (FoCal-E) of about 20$X_0$ and a hadronic copper-scintillating-fiber calorimeter (FoCal-H) of about 5$\lambda_{\rm int}$. The data were taken between 2021 and 2023 at the CERN PS and SPS beam lines with hadron (electron) beams up to energies of 350 (300) GeV. Regarding FoCal-E, we report a comprehensive analysis of its response to minimum ionizing particles across all pad layers. The longitudinal shower profile of electromagnetic showers is measured with a layer-wise segmentation of 1$X_0$. As a projection to the performance of the final detector in electromagnetic showers, we demonstrate linearity in the full energy range, and show that the energy resolution fulfills the requirements for the physics needs. Additionally, the performance to separate two-showers events was studied by quantifying the transverse shower width. Regarding FoCal-H, we report a detailed analysis of the response to hadron beams between 60 and 350 GeV. The results are compared to simulations obtained with a Geant4 model of the test beam setup, which in particular for FoCal-E are in good agreement with the data. The energy resolution of FoCal-E was found to be lower than 3% at energies larger than 100 GeV. The response of FoCal-H to hadron beams was found to be linear, albeit with a significant intercept that is about factor 2 larger than in simulations. Its resolution, which is non-Gaussian and generally larger than in simulations, was quantified using the FWHM, and decreases from about 16% at 100 GeV to about 11% at 350 GeV. The discrepancy to simulations, which is particularly evident at low hadron energies, needs to be further investigated., Comment: 57 pages (without acronyms), 45 captioned figures

We formulate and solve data-driven aerodynamic shape design problems with distributionally robust optimization (DRO) approaches. Building on the findings of the work \cite{gotoh2018robust}, we study the connections between a class of DRO and the Taguchi method in the context of robust design optimization. Our preliminary computational experiments on aerodynamic shape optimization in transonic turbulent flow show promising design results.

In this document we describe a model of an array of water Cherenkov detectors proposed to study ultra-high-energy gamma rays in the southern hemisphere, and a continuous model of secondary particles produced on the ground from gamma and proton showers. We use the model of the detector and the parametrization of showers for the identification of the most promising configuration of detector elements, using a likelihood ratio test statistic to classify showers and a stochastic gradient descent technique to maximize a utility function describing the measurement precision on the gamma-ray flux., Comment: 47 pages, 15 figures

In this article we examine recent developments in the research area concerning the creation of end-to-end models for the complete optimization of measuring instruments. The models we consider rely on differentiable programming methods and on the specification of a software pipeline including all factors impacting performance -- from the data-generating processes to their reconstruction and the extraction of inference on the parameters of interest of a measuring instrument -- along with the careful specification of a utility function well aligned with the end goals of the experiment. Building on previous studies originated within the MODE Collaboration, we focus specifically on applications involving instruments for particle physics experimentation, as well as industrial and medical applications that share the detection of radiation as their data-generating mechanism., Comment: 70 pages, 17 figures. To be submitted to journal

The coherence of light has been proposed as a quantum-mechanical control for enhancing light-harvesting efficiency. In particular, optical coherence can be manipulated by changing either the polarization state or spectral phase of the light. Here, we show that, in weak light, light-harvesting efficiency cannot be controlled using any form of optical coherence in molecular light-harvesting systems and, more broadly, those comprising orientationally disordered sub-units and operating on longer-than-ultrafast timescales. Under those conditions, optical coherence does not affect light-harvesting efficiency, meaning that it cannot be used for control. Specifically, polarization-state control is lost in disordered samples or when the molecules reorient on the timescales of the light-harvesting, and spectral-phase control is lost when the efficiency is time-averaged for longer than the optical coherence time. In practice, efficiency is always averaged over long times, meaning that coherent optical control is only possible through polarisation in systems with orientational order., Comment: 8 + 5 pages, 5 + 1 figures. Replaced previous manuscript with the version accepted by The Journal of Physical Chemistry Letters

In operator overloading algorithmic differentiation, it can be beneficial to create custom derivative functions for some parts of the code base. For manual implementations of the derivative functions, it can be quite cumbersome to derive, implement, test, and maintain these. The process can be automated with source transformation algorithmic differentiation tools like Tapenade or compiler-based algorithmic differentiation tools like Enzyme. This eliminates most of the work required from a manual implementation but usually has the same efficiency with respect to timing and memory. We present a new helper in CoDiPack that allows Enzyme-generated derivative functions to be automatically added during the recording process of CoDiPack. The validity of the approach is demonstrated on a synthetic benchmark, which shows promising results.

For short-packet, low-latency communications over random access channels, piloting overhead significantly reduces spectral efficiency. Therefore, pilotless systems recently gained attraction. While blind phase estimation algorithms such as Viterbi-Viterbi Phase Estimation (VVPE) can correct a phase offset using only payload symbols, a phase ambiguity remains. We first show that the remaining phase rotations in a polar coded quadrature amplitude modulation (QAM) transmission with gray labeling are combinations of bit-flips and automorphisms. Therefore, the decoder is equivariant to such phase rotations and, by smartly selecting the frozen bits, one can jointly decode and resolve the phase ambiguity, without the need for pilot symbols or an outer code. Our proposed system outperforms pilot-assisted transmissions by up to 0.8 dB and 2 dB for quaternary phase shift keying (QPSK) and 16-QAM, respectively., Comment: 5 pages, 6 figures, submitted to IEEE for possible publication

While several numerical techniques are available for predicting the dynamics of non-Markovian open quantum systems, most struggle with simulations for very long memory and propagation times, e.g., due to superlinear scaling with the number of time steps $n$. Here, we introduce a numerically exact algorithm to calculate process tensors -- compact representations of environmental influences -- which provides a scaling advantage over previous algorithms by leveraging self-similarity of the tensor networks that represent Gaussian environments. Based on a divide-and-conquer strategy, our approach requires only $\mathcal{O}(n\log n)$ singular value decompositions for environments with infinite memory. Where the memory can be truncated after $n_c$ time steps, a scaling $\mathcal{O}(n_c\log n_c)$ is found, which is independent of $n$. This improved scaling is enabled by identifying process tensors with repeatable blocks. To demonstrate the power and utility of our approach we provide three examples. (1) We calculate the fluorescence spectra of a quantum dot under both strong driving and strong dot-phonon couplings, a task requiring simulations over millions of time steps, which we are able to perform in minutes. (2) We efficiently find process tensors describing superradiance of multiple emitters. (3) We explore the limits of our algorithm by considering coherence decay with a very strongly coupled environment. The algorithm we present here not only significantly extends the scope of numerically exact techniques to open quantum systems with long memory times, but also has fundamental implications for simulation complexity.

We present a first-order method for solving constrained optimization problems. The method is derived from our previous work, a modified search direction method inspired by singular value decomposition. In this work, we simplify its computational framework to a ``gradient descent akin'' method (GDAM), i.e., the search direction is computed using a linear combination of the negative and normalized objective and constraint gradient. We give fundamental theoretical guarantees on the global convergence of the method. This work focuses on the algorithms and applications of GDAM. We present computational algorithms that adapt common strategies for the gradient descent method. We demonstrate the potential of the method using two engineering applications, shape optimization and sensor network localization. When practically implemented, GDAM is robust and very competitive in solving the considered large and challenging optimization problems.

We investigate the applicability of deep reinforcement learning algorithms to the adaptive initial access beam alignment problem for mmWave communications using the state-of-the-art proximal policy optimization algorithm as an example. In comparison to recent unsupervised learning based approaches developed to tackle this problem, deep reinforcement learning has the potential to address a new and wider range of applications, since, in principle, no (differentiable) model of the channel and/or the whole system is required for training, and only agent-environment interactions are necessary to learn an algorithm (be it online or using a recorded dataset). We show that, although the chosen off-the-shelf deep reinforcement learning agent fails to perform well when trained on realistic problem sizes, introducing action space shaping in the form of beamforming modules vastly improves the performance, without sacrificing much generalizability. Using this add-on, the agent is able to deliver competitive performance to various state-of-the-art methods on simulated environments, even under realistic problem sizes. This demonstrates that through well-directed modification, deep reinforcement learning may have a chance to compete with other approaches in this area, opening up many straightforward extensions to other/similar scenarios.

Broadband noise reduction is a significant problem in aerospace and industrial applications. Specifically, the noise generated from the trailing edge of an airfoil poses a challenging problem with various proposed solutions. This study investigates the porous trailing edge treatment. We use surrogate-based gradient-free optimization and an empirical noise model to efficiently explore the design space and find the optimal porosity distribution. As a result, a predicted 8-10 dB reduction in the broadband 300-5000 Hz was achieved. Furthermore, the optimal design emphasizes the design space's complexity and global exploration's difficulty. Further, the optimal design presents a low porous solution while constituting significant noise reduction., Comment: 5 pages, 6 figures, added affiliation, fixed typos, homog. citation style

Tools for algorithmic differentiation (AD) provide accurate derivatives of computer-implemented functions for use in, e. g., optimization and machine learning (ML). However, they often require the source code of the function to be available in a restricted set of programming languages. As a step towards making AD accessible for code bases with cross-language or closed-source components, we recently presented the forward-mode AD tool Derivgrind. It inserts forward-mode AD logic into the machine code of a compiled program using the Valgrind dynamic binary instrumentation framework. This work extends Derivgrind, adding the capability to record the real-arithmetic evaluation tree, and thus enabling operator overloading style reverse-mode AD for compiled programs. We maintain the high level of correctness reported for Derivgrind's forward mode, failing the same few testcases in an extensive test suite for the same well-understood reasons. Runtime-wise, the recording slows down the execution of a compiled 64-bit benchmark program by a factor of about 180., Comment: 17 pages, 5 figures, 1 listing

Understanding energy transport in quantum systems is crucial for an understanding of light-harvesting in nature, and for the creation of new quantum technologies. Open quantum systems theory has been successfully applied to predict the existence of environmental noise-assisted quantum transport (ENAQT) as a widespread phenomenon occurring in biological and artificial systems. That work has been primarily focused on several `canonical' structures, from simple chains, rings and crystals of varying dimensions, to well-studied light-harvesting complexes. Studying those particular systems has produced specific assumptions about ENAQT, including the notion of a single, ideal, range of environmental coupling rates that improve energy transport. In this paper we show that a consistent subset of physically modelled transport networks can have at least two ENAQT peaks in their steady state transport efficiency., Comment: Accepted Manuscript, 11 pages, 16 figures

Characterising the time over which quantum coherence survives is critical for any implementation of quantum bits, memories and sensors. The usual method for determining a quantum system's decoherence rate involves a suite of experiments probing the entire expected range of this parameter, and extracting the resulting estimation in post-processing. Here we present an adaptive multi-parameter Bayesian approach, based on a simple analytical update rule, to estimate the key decoherence timescales ($T_1$, $T_2^*$ and $T_2$) and the corresponding decay exponent of a quantum system in real time, using information gained in preceding experiments. This approach reduces the time required to reach a given uncertainty by a factor up to an order of magnitude, depending on the specific experiment, compared to the standard protocol of curve fitting. A further speed-up of a factor $\sim 2$ can be realised by performing our optimisation with respect to sensitivity as opposed to variance., Comment: made a change of word in the title ("Online" to "Real-time") and rearranged figures and text to improve the readability of the manuscript

When operating massive multiple-input multiple-output (MIMO) systems with uplink (UL) and downlink (DL) channels at different frequencies (frequency division duplex (FDD) operation), acquisition of channel state information (CSI) for downlink precoding is a major challenge. Since, barring transceiver impairments, both UL and DL CSI are determined by the physical environment surrounding transmitter and receiver, it stands to reason that, for a static environment, a mapping from UL CSI to DL CSI may exist. First, we propose to use various neural network (NN)-based approaches that learn this mapping and provide baselines using classical signal processing. Second, we introduce a scheme to evaluate the performance and quality of generalization of all approaches, distinguishing between known and previously unseen physical locations. Third, we evaluate all approaches on a real-world indoor dataset collected with a 32-antenna channel sounder.

We consider bio-inspired ring systems as photovoltaic circuits to explore the advantage of `optical ratcheting', a process whereby the arrangement of coupled optical dipoles enables delocalised excitonic states that are protected against radiative decay whilst permitting the absorption of further photons. We explore how the performance of a ratcheting antenna scales with system size when excitons are incoherently or coherently extracted from the antenna to an associated trap site. In both instances we also move to the polaron frame in order to more closely model realistic systems where the coupling to vibrational modes can generally be assumed to be strong. We find a multifaceted and nuanced dependence of the predicted performance on an interplay between geometrical arrangement, extraction mechanism, and vibrational coupling strength. Certain regimes support substantial performance improvements in the power extracted per site., Comment: 20 pages, 15 figures

Algorithmic differentiation (AD) is a set of techniques that provide partial derivatives of computer-implemented functions. Such a function can be supplied to state-of-the-art AD tools via its source code, or via an intermediate representation produced while compiling its source code. We present the novel AD tool Derivgrind, which augments the machine code of compiled programs with forward-mode AD logic. Derivgrind leverages the Valgrind instrumentation framework for a structured access to the machine code, and a shadow memory tool to store dot values. Access to the source code is required at most for the files in which input and output variables are defined. Derivgrind's versatility comes at the price of scaling the run-time by a factor between 30 and 75, measured on a benchmark based on a numerical solver for a partial differential equation. Results of our extensive regression test suite indicate that Derivgrind produces correct results on GCC- and Clang-compiled programs, including a Python interpreter, with a small number of exceptions. While we provide a list of scenarios that Derivgrind does not handle correctly, nearly all of them are academic counterexamples or originate from highly optimized math libraries. As long as differentiating those is avoided, Derivgrind can be applied to an unprecedentedly wide range of cross-language or partially closed-source software with little integration efforts., Comment: 21 pages, 3 figures, 3 tables, 5 listings