Showing 1–24 of 24 results for author: Fay, D

Dynamic Privacy Allocation for Locally Differentially Private Federated Learning with Composite Objectives

Authors: Jiaojiao Zhang, Dominik Fay, Mikael Johansson

Abstract: This paper proposes a locally differentially private federated learning algorithm for strongly convex but possibly nonsmooth problems that protects the gradients of each worker against an honest but curious server. The proposed algorithm adds artificial noise to the shared information to ensure privacy and dynamically allocates the time-varying noise variance to minimize an upper bound of the opti… ▽ More This paper proposes a locally differentially private federated learning algorithm for strongly convex but possibly nonsmooth problems that protects the gradients of each worker against an honest but curious server. The proposed algorithm adds artificial noise to the shared information to ensure privacy and dynamically allocates the time-varying noise variance to minimize an upper bound of the optimization error subject to a predefined privacy budget constraint. This allows for an arbitrarily large but finite number of iterations to achieve both privacy protection and utility up to a neighborhood of the optimal solution, removing the need for tuning the number of iterations. Numerical results show the superiority of the proposed algorithm over state-of-the-art methods. △ Less

Submitted 2 August, 2023; originally announced August 2023.

Personalized Privacy Amplification via Importance Sampling

Authors: Dominik Fay, Sebastian Mair, Jens Sjölund

Abstract: We examine the privacy-enhancing properties of importance sampling. In importance sampling, selection probabilities are heterogeneous and each selected data point is weighted by the reciprocal of its selection probability. Due to the heterogeneity of importance sampling, we express our results within the framework of personalized differential privacy. We first consider the general case where an ar… ▽ More We examine the privacy-enhancing properties of importance sampling. In importance sampling, selection probabilities are heterogeneous and each selected data point is weighted by the reciprocal of its selection probability. Due to the heterogeneity of importance sampling, we express our results within the framework of personalized differential privacy. We first consider the general case where an arbitrary personalized differentially private mechanism is subsampled with an arbitrary importance sampling distribution and show that the resulting mechanism also satisfies personalized differential privacy. This constitutes an extension of the established privacy amplification by subsampling result to importance sampling. Then, for any fixed mechanism, we derive the sampling distribution that achieves the optimal sampling rate subject to a worst-case privacy constraint. Empirically, we evaluate the privacy, efficiency, and accuracy of importance sampling on the example of k-means clustering. △ Less

Submitted 15 February, 2024; v1 submitted 5 July, 2023; originally announced July 2023.

Comments: Preprint. Under review

Future Directions of the Cyberinfrastructure for Sustained Scientific Innovation (CSSI) Program

Authors: Ritu Arora, Xiaosong Li, Bonnie Hurwitz, Daniel Fay, Dhabaleswar K. Panda, Edward Valeev, Shaowen Wang, Shirley Moore, Sunita Chandrasekaran, Ting Cao, Holly Bik, Matthew Curry, Tanzima Islam

Abstract: The CSSI 2019 workshop was held on October 28-29, 2019, in Austin, Texas. The main objectives of this workshop were to (1) understand the impact of the CSSI program on the community over the last 9 years, (2) engage workshop participants in identifying gaps and opportunities in the current CSSI landscape, (3) gather ideas on the cyberinfrastructure needs and expectations of the community with resp… ▽ More The CSSI 2019 workshop was held on October 28-29, 2019, in Austin, Texas. The main objectives of this workshop were to (1) understand the impact of the CSSI program on the community over the last 9 years, (2) engage workshop participants in identifying gaps and opportunities in the current CSSI landscape, (3) gather ideas on the cyberinfrastructure needs and expectations of the community with respect to the CSSI program, and (4) prepare a report summarizing the feedback gathered from the community that can inform the future solicitations of the CSSI program. The workshop brought together different stakeholders interested in provisioning sustainable cyberinfrastructure that can power discoveries impacting the various fields of science and technology and maintaining the nation's competitiveness in the areas such as scientific software, HPC, networking, cybersecurity, and data/information science. The workshop served as a venue for gathering the community-feedback on the current state of the CSSI program and its future directions. △ Less

Submitted 15 October, 2020; originally announced October 2020.

Comments: This report was submitted in April 2020 to the National Science Foundation (NSF)

Decentralized Differentially Private Segmentation with PATE

Authors: Dominik Fay, Jens Sjölund, Tobias J. Oechtering

Abstract: When it comes to preserving privacy in medical machine learning, two important considerations are (1) kee** data local to the institution and (2) avoiding inference of sensitive information from the trained model. These are often addressed using federated learning and differential privacy, respectively. However, the commonly used Federated Averaging algorithm requires a high degree of synchroniz… ▽ More When it comes to preserving privacy in medical machine learning, two important considerations are (1) kee** data local to the institution and (2) avoiding inference of sensitive information from the trained model. These are often addressed using federated learning and differential privacy, respectively. However, the commonly used Federated Averaging algorithm requires a high degree of synchronization between participating institutions. For this reason, we turn our attention to Private Aggregation of Teacher Ensembles (PATE), where all local models can be trained independently without inter-institutional communication. The purpose of this paper is thus to explore how PATE -- originally designed for classification -- can best be adapted for semantic segmentation. To this end, we build low-dimensional representations of segmentation masks which the student can obtain through low-sensitivity queries to the private aggregator. On the Brain Tumor Segmentation (BraTS 2019) dataset, an Autoencoder-based PATE variant achieves a higher Dice coefficient for the same privacy guarantee than prior work based on noisy Federated Averaging. △ Less

Submitted 9 April, 2020; originally announced April 2020.

Comments: Under review for MICCAI 2020

Privacy of Real-Time Pricing in Smart Grid

Authors: Mahrokh GhoddousiBoroujeni, Dominik Fay, Christos Dimitrakakis, Maryam Kamgarpour

Abstract: Installing smart meters to publish real-time electricity rates has been controversial while it might lead to privacy concerns. Dispatched rates include fine-grained data on aggregate electricity consumption in a zone and could potentially be used to infer a household's pattern of energy use or its occupancy. In this paper, we propose Blowfish privacy to protect the occupancy state of the houses co… ▽ More Installing smart meters to publish real-time electricity rates has been controversial while it might lead to privacy concerns. Dispatched rates include fine-grained data on aggregate electricity consumption in a zone and could potentially be used to infer a household's pattern of energy use or its occupancy. In this paper, we propose Blowfish privacy to protect the occupancy state of the houses connected to a smart grid. First, we introduce a Markov model of the relationship between electricity rate and electricity consumption. Next, we develop an algorithm that perturbs electricity rates before publishing them to ensure users' privacy. Last, the proposed algorithm is tested on data inspired by household occupancy models and its performance is compared to an alternative solution. △ Less

Submitted 18 December, 2019; originally announced December 2019.

A self-organising eigenspace map for time series clustering

Authors: Donya Rahmani, Damien Fay, Jacek Brodzki

Abstract: This paper presents a novel time series clustering method, the self-organising eigenspace map (SOEM), based on a generalisation of the well-known self-organising feature map (SOFM). The SOEM operates on the eigenspaces of the embedded covariance structures of time series which are related directly to modes in those time series. Approximate joint diagonalisation acts as a pseudo-metric across these… ▽ More This paper presents a novel time series clustering method, the self-organising eigenspace map (SOEM), based on a generalisation of the well-known self-organising feature map (SOFM). The SOEM operates on the eigenspaces of the embedded covariance structures of time series which are related directly to modes in those time series. Approximate joint diagonalisation acts as a pseudo-metric across these spaces allowing us to generalise the SOFM to a neural network with matrix input. The technique is empirically validated against three sets of experiments; univariate and multivariate time series clustering, and application to (clustered) multi-variate time series forecasting. Results indicate that the technique performs a valid topologically ordered clustering of the time series. The clustering is superior in comparison to standard benchmarks when the data is non-aligned, gives the best clustering stage for when used in forecasting, and can be used with partial/non-overlap** time series, multivariate clustering and produces a topological representation of the time series objects. △ Less

Submitted 14 May, 2019; originally announced May 2019.

Comments: 16 pages-27 figures

Automated Adaptation Strategies for Stream Learning

Authors: Rashid Bakirov, Bogdan Gabrys, Damien Fay

Abstract: Automation of machine learning model development is increasingly becoming an established research area. While automated model selection and automated data pre-processing have been studied in depth, there is, however, a gap concerning automated model adaptation strategies when multiple strategies are available. Manually develo** an adaptation strategy can be time consuming and costly. In this pap… ▽ More Automation of machine learning model development is increasingly becoming an established research area. While automated model selection and automated data pre-processing have been studied in depth, there is, however, a gap concerning automated model adaptation strategies when multiple strategies are available. Manually develo** an adaptation strategy can be time consuming and costly. In this paper we address this issue by proposing the use of flexible adaptive mechanism deployment for automated development of adaptation strategies. Experimental results after using the proposed strategies with five adaptive algorithms on 36 datasets confirm their viability. These strategies achieve better or comparable performance to the custom adaptation strategies and the repeated deployment of any single adaptive mechanism. △ Less

Submitted 30 April, 2021; v1 submitted 27 December, 2018; originally announced December 2018.

Implementing Adaptive Separable Convolution for Video Frame Interpolation

Authors: Mart Kartašev, Carlo Rapisarda, Dominik Fay

Abstract: As Deep Neural Networks are becoming more popular, much of the attention is being devoted to Computer Vision problems that used to be solved with more traditional approaches. Video frame interpolation is one of such challenges that has seen new research involving various techniques in deep learning. In this paper, we replicate the work of Niklaus et al. on Adaptive Separable Convolution, which cla… ▽ More As Deep Neural Networks are becoming more popular, much of the attention is being devoted to Computer Vision problems that used to be solved with more traditional approaches. Video frame interpolation is one of such challenges that has seen new research involving various techniques in deep learning. In this paper, we replicate the work of Niklaus et al. on Adaptive Separable Convolution, which claims high quality results on the video frame interpolation task. We apply the same network structure trained on a smaller dataset and experiment with various different loss functions, in order to determine the optimal approach in data-scarce scenarios. The best resulting model is still able to provide visually pleasing videos, although achieving lower evaluation scores. △ Less

Submitted 20 September, 2018; originally announced September 2018.

Comments: All authors contributed equally

An exploration of fetish social networks and communities

Authors: Damien Fay, Hamed Haddadi, Michael C. Seto, Han Wang, Christoph Carl Kling

Abstract: Online Social Networks (OSNs) provide a venue for virtual interactions and relationships between individuals. In some communities, OSNs also facilitate arranging online meetings and relationships. FetLife, the worlds largest anonymous social network for the BDSM, fetish and kink communities, provides a unique example of an OSN that serves as an interaction space, community organizing tool, and sex… ▽ More Online Social Networks (OSNs) provide a venue for virtual interactions and relationships between individuals. In some communities, OSNs also facilitate arranging online meetings and relationships. FetLife, the worlds largest anonymous social network for the BDSM, fetish and kink communities, provides a unique example of an OSN that serves as an interaction space, community organizing tool, and sexual market. In this paper, we present a first look at the characteristics of European members of Fetlife, comprising 504,416 individual nodes with 1,912,196 connections. We looked at user characteristics in terms of gender, sexual orientation, and preferred role. We further examined the topological and structural properties of groups, as well as the type of interactions and relations between their members. Our results suggest there are important differences between the FetLife community and conventional OSNs. The network can be characterised by complex gender based interactions both from a sexual market and platonic viewpoint which point to a truly fascinating social network. △ Less

Submitted 4 November, 2015; originally announced November 2015.

Comments: 16 pages

Predictive partitioning for efficient BFS traversal in social networks

Authors: Damien Fay

Abstract: In this paper we show how graph structure can be used to drastically reduce the computational bottleneck of the Breadth First Search algorithm (the foundation of many graph traversal techniques). In particular, we address parallel implementations where the bottleneck is the number of messages between processors emitted at the peak iteration. First, we derive an expression for the expected degree d… ▽ More In this paper we show how graph structure can be used to drastically reduce the computational bottleneck of the Breadth First Search algorithm (the foundation of many graph traversal techniques). In particular, we address parallel implementations where the bottleneck is the number of messages between processors emitted at the peak iteration. First, we derive an expression for the expected degree distribution of vertices in the frontier of the algorithm which is shown to be highly skewed. Subsequently, we derive an expression for the expected message along an edge in a particular iteration. This skew suggests a weighted, iteration based, partition would be advantageous. Employing the METIS algorithm we then show empirically that such partitions can reduce the message overhead by up to 50% in some particular instances and in the order of 20% on average. These results have implications for graph processing in multiprocessor and distributed computing environments. △ Less

Submitted 15 October, 2015; originally announced October 2015.

On Joint Diagonalisation for Dynamic Network Analysis

Authors: Damien Fay, Jérôme Kunegis, Eiko Yoneki

Abstract: Joint diagonalisation (JD) is a technique used to estimate an average eigenspace of a set of matrices. Whilst it has been used successfully in many areas to track the evolution of systems via their eigenvectors; its application in network analysis is novel. The key focus in this paper is the use of JD on matrices of spanning trees of a network. This is especially useful in the case of real-world c… ▽ More Joint diagonalisation (JD) is a technique used to estimate an average eigenspace of a set of matrices. Whilst it has been used successfully in many areas to track the evolution of systems via their eigenvectors; its application in network analysis is novel. The key focus in this paper is the use of JD on matrices of spanning trees of a network. This is especially useful in the case of real-world contact networks in which a single underlying static graph does not exist. The average eigenspace may be used to construct a graph which represents the `average spanning tree' of the network or a representation of the most common propagation paths. We then examine the distribution of deviations from the average and find that this distribution in real-world contact networks is multi-modal; thus indicating several \emph{modes} in the underlying network. These modes are identified and are found to correspond to particular times. Thus JD may be used to decompose the behaviour, in time, of contact networks and produce average static graphs for each time. This may be viewed as a mixture between a dynamic and static graph approach to contact network analysis. △ Less

Submitted 6 October, 2011; originally announced October 2011.

Beyond Node Degree: Evaluating AS Topology Models

Authors: Hamed Haddadi, Damien Fay, Almerima Jamakovic, Olaf Maennel, Andrew W. Moore, Richard Mortier, Miguel Rio, Steve Uhlig

Abstract: Many models have been proposed to generate Internet Autonomous System (AS) topologies, most of which make structural assumptions about the AS graph. In this paper we compare AS topology generation models with several observed AS topologies. In contrast to most previous works, we avoid making assumptions about which topological properties are important to characterize the AS topology. Our analysi… ▽ More Many models have been proposed to generate Internet Autonomous System (AS) topologies, most of which make structural assumptions about the AS graph. In this paper we compare AS topology generation models with several observed AS topologies. In contrast to most previous works, we avoid making assumptions about which topological properties are important to characterize the AS topology. Our analysis shows that, although matching degree-based properties, the existing AS topology generation models fail to capture the complexity of the local interconnection structure between ASs. Furthermore, we use BGP data from multiple vantage points to show that additional measurement locations significantly affect local structure properties, such as clustering and node centrality. Degree-based properties, however, are not notably affected by additional measurements locations. These observations are particularly valid in the core. The shortcomings of AS topology generation models stems from an underestimation of the complexity of the connectivity in the core caused by inappropriate use of BGP data. △ Less

Submitted 13 July, 2008; originally announced July 2008.

ACM Class: C.2.1; I.6.4

Thermal conductivity in the vortex state of the superconductor UPd_2Al_3

Authors: L. Tewordt, D. Fay

Abstract: The magneto-thermal conductivity kappa is calculated for the vortex state of UPd_2Al_3 by assuming horizontal gap nodes. The Green's function method we employed takes into account the effects of supercurrent flow and Andreev scattering on the quasiparticles due to Abrikosov's vortex lattice order parameter. The calculated angular dependence of kappa_{yy} for field rotation theta_0 in the ac-plan… ▽ More The magneto-thermal conductivity kappa is calculated for the vortex state of UPd_2Al_3 by assuming horizontal gap nodes. The Green's function method we employed takes into account the effects of supercurrent flow and Andreev scattering on the quasiparticles due to Abrikosov's vortex lattice order parameter. The calculated angular dependence of kappa_{yy} for field rotation theta_0 in the ac-plane depends strongly on field strength H, impurity scattering, anisotropy of the Fermi velocity, and temperature. For finite temperatures and the clean unitary scattering limit we get qualitative agreement with recent experiments for all four proposed gap functions having horizontal line nodes at ck_z = 0, pi/4, and pi/2. △ Less

Submitted 3 May, 2005; originally announced May 2005.

Comments: 12 pages, 3 figures with several parts

Journal ref: Physical Review B 72, 014502 (2005)

Specific heat and thermal conductivity in Sr_2RuO_4 for rotating in-plane magnetic field

Authors: L. Tewordt, D. Fay

Abstract: We calculate the density of states $N$ and the the thermal conductivity $κ$ in layered superconductors for rotating in-plane magnetic field by using approximate analytical expressions for Abrikosov's vortex lattice state. For a gap with d$_{xy} $- orbital symmetry we find that the differences $ΔN$ and $Δκ$ between field directions parallel to the antinodes and parallel to the nodes of the gap ar… ▽ More We calculate the density of states $N$ and the the thermal conductivity $κ$ in layered superconductors for rotating in-plane magnetic field by using approximate analytical expressions for Abrikosov's vortex lattice state. For a gap with d$_{xy} $- orbital symmetry we find that the differences $ΔN$ and $Δκ$ between field directions parallel to the antinodes and parallel to the nodes of the gap are positive for lower fields and change sign for higher fields near $H_{c2} $. For increasing impurity scattering, $ΔN$ and $Δκ$ decrease due to a broadening of certain peaks in the angle-resolved density of states. The frequency dependencies of the amplitudes $ΔN$ and $Δκ$ exhibit similar crossover behavior at low frequencies which give rise to a sign change of $Δκ$ at low $T/T_c $. We conclude that variations of the specific heat and thermal conductivity can be observed only in the clean and low temperature limits. △ Less

Submitted 26 April, 2004; originally announced April 2004.

Comments: Submitted to Physical Review B

Thermal conductivity in the vortex state of YBa_2Cu_3O_7 and Sr_2RuO_4

Authors: L. Tewordt, D. Fay

Abstract: The thermal conductivity $κ$ is calculated for the Abrikosov vortex lattice with d-wave pairing. The Andreev scattering rate arises from the self energy term in Gorkov's integral equation for the corresponding Green's function. The impurity scattering rate is calculated self-consistently from this Green's function in the t-matrix approximation. Neither point-like impurity scattering with a phase… ▽ More The thermal conductivity $κ$ is calculated for the Abrikosov vortex lattice with d-wave pairing. The Andreev scattering rate arises from the self energy term in Gorkov's integral equation for the corresponding Green's function. The impurity scattering rate is calculated self-consistently from this Green's function in the t-matrix approximation. Neither point-like impurity scattering with a phase shift in the almost unitary limit nor in the Born limit can consistently explain recent zero-field and finite-field measurements of $κ/T$ at low $T$ in ultraclean YBCO. Our theory also applies to the spin triplet pairing states with vertical or horizontal line nodes in Sr$_2$RuO$_4$ △ Less

Submitted 23 February, 2004; v1 submitted 13 January, 2004; originally announced January 2004.

Comments: 4 pages, 3 figures, references added, discussion expanded

Two-band theory of specific heat and thermal conductivity in the mixed state of MgB_2

Authors: L. Tewordt, D. Fay

Abstract: We solve the coupled gap equations for the $σ$- and $π$-bands of MgB$_2$ in the vortex state and calculate the resulting field dependencies of the specific heat coefficient $γ$ and the thermal conductivity $κ$. The crucial parameters of the theory are the interband pairing interaction $λ_{πσ}$ and the ratio $s=ξ_σ/ξ_π$ of the coherence lengths. For reasonably small $λ_{πσ}$ and s, the small gap… ▽ More We solve the coupled gap equations for the $σ$- and $π$-bands of MgB$_2$ in the vortex state and calculate the resulting field dependencies of the specific heat coefficient $γ$ and the thermal conductivity $κ$. The crucial parameters of the theory are the interband pairing interaction $λ_{πσ}$ and the ratio $s=ξ_σ/ξ_π$ of the coherence lengths. For reasonably small $λ_{πσ}$ and s, the small gap $Δ_π$ decreases with increasing magnetic field $H$ much faster than the large gap $Δ_σ $. This gives rise to the observed rapid increase of $γ_π$ and $κ_π$ for small fields while $γ_σ$ and $κ_σ$ exhibit conventional field dependencies. Inclusion of intraband impurity scattering yields fairly good agreement with experiments for applied fields along the c axis. △ Less

Submitted 4 July, 2003; v1 submitted 13 May, 2003; originally announced May 2003.

Comments: 4 pages, 3 figures, Accepted by Phys. Rev. B

Specific heat and thermal conductivity in the vortex state of the two-gap superconductor MgB_2

Authors: L. Tewordt, D. Fay

Abstract: The specific heat coefficient gamma_s(H) and the electronic thermal conductivity kappa_{es}(H) are calculated for Abrikosov's vortex lattice by taking into account the effects of supercurrent flow and Andreev scattering. First we solve the gap equation for the entire range of magnetic fields. We take into account vertex corrections due to impurity scattering calculated in the Born approximation.… ▽ More The specific heat coefficient gamma_s(H) and the electronic thermal conductivity kappa_{es}(H) are calculated for Abrikosov's vortex lattice by taking into account the effects of supercurrent flow and Andreev scattering. First we solve the gap equation for the entire range of magnetic fields. We take into account vertex corrections due to impurity scattering calculated in the Born approximation. The function gamma_s(H)/gamma_n increases from zero and becomes approximately linear above H/H_{c2} \sim 0.1. The dependence on impurity scattering is substantially reduced by the vertex corrections. The upward curvature of kappa_{es}(H)/kappa_{en}, which is caused by decreasing Andreev scattering for increasing field, is reduced for increasing impurity scattering. We also calculate the temperature dependence of the scattering rates 1/tau_{ps}(H) of a phonon and 1/tau_{es}(H) of a quasiparticle due to quasiparticle and phonon scattering, respectively. At low temperatures the ratio tau_{pn}/tau_{ps}(H) increases rapidly to one as H tends to H_{c2} which yields a rapid drop in the phononic thermal conductivity kappa_{ph}. Our results are in qualitative agreement with the experiments on the two-gap superconductor MgB_2. △ Less

Submitted 20 March, 2003; v1 submitted 19 December, 2002; originally announced December 2002.

Comments: 12 pages, 5 figures, additions to figures 1, 2, and 3. Accepted by Phys. Rev. B

Journal ref: Phys. Rev. B 67,134524 (2003)

Effect of spin orbit scattering on the magnetic and superconducting properties of nearly ferromagnetic metals: application to granular Pt

Authors: D. Fay, J. Appel

Abstract: We calculate the effect of scattering on the static, exchange enhanced, spin susceptibility and show that in particular spin orbit scattering leads to a reduction of the giant moments and spin glass freezing temperature due to dilute magnetic impurities. The harmful spin fluctuation contribution to the intra-grain pairing interaction is strongly reduced opening the way for BCS superconductivity.… ▽ More We calculate the effect of scattering on the static, exchange enhanced, spin susceptibility and show that in particular spin orbit scattering leads to a reduction of the giant moments and spin glass freezing temperature due to dilute magnetic impurities. The harmful spin fluctuation contribution to the intra-grain pairing interaction is strongly reduced opening the way for BCS superconductivity. We are thus able to explain the superconducting and magnetic properties recently observed in granular Pt as due to scattering effects in single small grains. △ Less

Submitted 8 August, 2002; originally announced August 2002.

Comments: 9 pages 3 figures, accepted for publication in Phys. Rev. Letters

Specific heat and thermal conductivity in the mixed state of MgB2

Authors: L. Tewordt, D. Fay

Abstract: The specific heat C and the electronic and phononic thermal conductivities kappa_e and kappa_{ph} are calculated in the mixed state for magnetic fields H near H_{c2}. The effects of supercurrent flow and Andreev scattering of the Abrikosov vortex lattice on the quasiparticles are taken into account. The resulting function C(H) is nearly linear while kappa_e(H) exhibits an upward curvature near H… ▽ More The specific heat C and the electronic and phononic thermal conductivities kappa_e and kappa_{ph} are calculated in the mixed state for magnetic fields H near H_{c2}. The effects of supercurrent flow and Andreev scattering of the Abrikosov vortex lattice on the quasiparticles are taken into account. The resulting function C(H) is nearly linear while kappa_e(H) exhibits an upward curvature near H_{c2}. The slopes decrease with impurity scattering which improves the agreement with the data on MgB_2. The ratio of phonon relaxation times tau_n/tau_s = g(omega_0,H) for phonon energy omega_0, which is nearly a step function at omega_0 = 2Delta for the BCS state, is smeared out and tends to one for increasing H. This leads to a rapid reduction of kappa_{ph}(H) in MgB_2 for relatively small fields due to the rapid suppression of the smaller energy gap. △ Less

Submitted 14 August, 2002; v1 submitted 22 April, 2002; originally announced April 2002.

Comments: 8 pages, 4 figures, accepted for publication in Phys. Rev. Letters

Ultrasonic attenuation in magnetic fields for superconducting states with line nodes in Sr2RuO4

Authors: L. Tewordt, D. Fay

Abstract: We calculate the ultrasonic attenuation in magnetic fields for superconducting states with line nodes vertical or horizontal relative to the RuO_2 planes. This theory, which is valid for fields near Hc2 and not too low temperatures, takes into account the effects of supercurrent flow and Andreev scattering by the Abrikosov vortex lattice. For rotating in-plane field H(theta) the attenuation alph… ▽ More We calculate the ultrasonic attenuation in magnetic fields for superconducting states with line nodes vertical or horizontal relative to the RuO_2 planes. This theory, which is valid for fields near Hc2 and not too low temperatures, takes into account the effects of supercurrent flow and Andreev scattering by the Abrikosov vortex lattice. For rotating in-plane field H(theta) the attenuation alpha(theta)exhibits variations of fourfold symmetry in the rotation angle theta. In the case of vertical nodes, the transverse T100 sound mode yields the weakest(linear)H and T dependence of alpha, while the longitudinal L100 mode yields a stronger (quadratic) H and T dependence. This is in strong contrast to the case of horizontal line nodes where alpha is the same for the T100 and L100 modes (apart from a shift of pi/4 in field direction) and is roughly a quadratic function of H and T. Thus we conclude that measurements of alpha in in-plane magnetic fields for different in-plane sound modes may be an important tool for probing the nodal structure of the gap in Sr_2RuO_4. △ Less

Submitted 13 December, 2001; v1 submitted 7 August, 2001; originally announced August 2001.

Comments: 5 pages, 6 figures, replaced in non-preprint form, to appear in Phys. Rev. B

Vertex corrections in nearly ferroelectric superconductors

Authors: D. Fay, M. Weger

Abstract: We investigate the effect of an incipient ferrolectric transition on vertex corrections to the superconducting pairing interaction. The vertex corrections for small momentum transfers are large independent of the type of Boson responsible for the superconducting transition. The electron-phonon interaction is found to be enhanced by a nearly ferroelectric medium. We discuss application to the cup… ▽ More We investigate the effect of an incipient ferrolectric transition on vertex corrections to the superconducting pairing interaction. The vertex corrections for small momentum transfers are large independent of the type of Boson responsible for the superconducting transition. The electron-phonon interaction is found to be enhanced by a nearly ferroelectric medium. We discuss application to the cuprate superconductors. △ Less

Submitted 25 April, 2001; originally announced April 2001.

Comments: 7 pages, 4 figures, published in Phys. Rev. B 62, 15208 (2000)

Journal ref: Phys. Rev. B 62, 15208 (2000)

Thermal Conductivity near H_c2 for spin-triplet superconducting States with line nodes in Sr_2RuO_4

Authors: L. Tewordt, D. Fay

Abstract: We calculate the thermal conductivity kappa in magnetic fields near H_c2 for spin-triplet superconducting states with line nodes vertical and horizontal relative to the RuO_2-planes. The method for calculating the Green's functions takes into account the spatial variation of the order parameter and superconducting flow for the Abrikosov vortex lattice. For in-plane magnetic field we obtain varia… ▽ More We calculate the thermal conductivity kappa in magnetic fields near H_c2 for spin-triplet superconducting states with line nodes vertical and horizontal relative to the RuO_2-planes. The method for calculating the Green's functions takes into account the spatial variation of the order parameter and superconducting flow for the Abrikosov vortex lattice. For in-plane magnetic field we obtain variations of the in-plane kappa with two-fold symmetry as a function of rotation angle where the minima and maxima occur for field directions parallel and perpendicular to the heat flow. The amplitude of the variation decreases with increasing impurity scattering and temperature. At higher temperatures the minima and maxima of the variation are interchanged. Since the results for vertical and horizontal line nodes are almost the same we cannot say which of the two pairing models is more compatible with recent measurements of kappa in Sr_2RuO_4. The observed four-fold modulation of kappa in YBa_2Cu_3O_(7-δ) is obtained for d-wave pairing by taking into account the particular shape of the Fermi surface and the finite temperature effect. The results for kappa for the f-wave pairing state with horizontal line nodes disagree in some respects with the measurements on UPt_3. △ Less

Submitted 23 April, 2001; originally announced April 2001.

Comments: 8 pages, 6 figures. To be published in Phys. Rev. B

Journal ref: Phys. Rev. B 64, 024528 (2001)

NMR and NQR Fluctuation Effects in Layered Superconductors

Authors: D. Fay, J. Appel, C. Timm, A. Zabel

Abstract: We study the effect of thermal fluctuations of the s-wave order parameter of a quasi two dimensional superconductor on the nuclear spin relaxation rate near the transition temperature Tc. We consider both the effects of the amplitude fluctuations and the Berezinskii-Kosterlitz-Thouless (BKT) phase fluctuations in weakly coupled layered superconductors. In the treatment of the amplitude fluctuati… ▽ More We study the effect of thermal fluctuations of the s-wave order parameter of a quasi two dimensional superconductor on the nuclear spin relaxation rate near the transition temperature Tc. We consider both the effects of the amplitude fluctuations and the Berezinskii-Kosterlitz-Thouless (BKT) phase fluctuations in weakly coupled layered superconductors. In the treatment of the amplitude fluctuations we employ the Gaussian approximation and evaluate the longitudinal relaxation rate 1/T1 for a clean s-wave superconductor, with and without pair breaking effects, using the static pair fluctuation propagator D. The increase in 1/T1 due to pair breaking in D is overcompensated by the decrease arising from the single particle Green's functions. The result is a strong effect on 1/T1 for even a small amount of pair breaking. The phase fluctuations are described in terms of dynamical BKT excitations in the form of pancake vortex-antivortex (VA) pairs. We calculate the effect of the magnetic field fluctuations caused by the translational motion of VA excitations on 1/T1 and on the transverse relaxation rate 1/T2 on both sides of the BKT transitation temperature T(BKT)<Tc. The results for the NQR relaxation rates depend strongly on the diffusion constant that governs the motion of free and bound vortices as well as the annihilation of VA pairs. We discuss the relaxation rates for real multilayer systems where the diffusion constant can be small and thus increase the lifetime of a VA pair, leading to an enhancement of the rates. We also discuss in some detail the experimental feasibility of observing the effects of amplitude fluctuations in layered s-wave superconductors such as the dichalcogenides and the effects of phase fluctuations in s- or d-wave superconductors such as the layered cuprates. △ Less

Submitted 6 December, 2000; originally announced December 2000.

Comments: 38 pages, 12 figures

Exchange Enhancement of the Electron-Phonon Pair Interaction

Authors: D. Manske, C. T. Rieck, D. Fay

Abstract: The critical temperature of high-$T_c$ superconductors is determined, at least in part, by the electron-phonon coupling. We include the effect of an exchange interaction between the electrons and calculate the renormalization of the bare phonon frequencies and the electron-phonon verticies in a random phase approximation and obtain a strongly enhanced attractive phonon-induced electron-electron… ▽ More The critical temperature of high-$T_c$ superconductors is determined, at least in part, by the electron-phonon coupling. We include the effect of an exchange interaction between the electrons and calculate the renormalization of the bare phonon frequencies and the electron-phonon verticies in a random phase approximation and obtain a strongly enhanced attractive phonon-induced electron-electron interaction. Using Fast Fourier Transform techniques, the weak-coupling selfconsistency equation for the order parameter is solved in the 2D first Brillouin zone for the Emery tight-binding band with different band fillings. The enhancement of $T_c$ arises primarily from the softening of the phonon frequencies rather than the vertex renormalization. △ Less

Submitted 1 July, 1994; originally announced July 1994.

Comments: (2 pages, postscript file, hardcopies available from the authors)