-
Gradient Inversion of Federated Diffusion Models
Authors:
Jiyue Huang,
Chi Hong,
Lydia Y. Chen,
Stefanie Roos
Abstract:
Diffusion models are becoming defector generative models, which generate exceptionally high-resolution image data. Training effective diffusion models require massive real data, which is privately owned by distributed parties. Each data party can collaboratively train diffusion models in a federated learning manner by sharing gradients instead of the raw data. In this paper, we study the privacy l…
▽ More
Diffusion models are becoming defector generative models, which generate exceptionally high-resolution image data. Training effective diffusion models require massive real data, which is privately owned by distributed parties. Each data party can collaboratively train diffusion models in a federated learning manner by sharing gradients instead of the raw data. In this paper, we study the privacy leakage risk of gradient inversion attacks. First, we design a two-phase fusion optimization, GIDM, to leverage the well-trained generative model itself as prior knowledge to constrain the inversion search (latent) space, followed by pixel-wise fine-tuning. GIDM is shown to be able to reconstruct images almost identical to the original ones. Considering a more privacy-preserving training scenario, we then argue that locally initialized private training noise $ε$ and sampling step t may raise additional challenges for the inversion attack. To solve this, we propose a triple-optimization GIDM+ that coordinates the optimization of the unknown data, $ε$ and $t$. Our extensive evaluation results demonstrate the vulnerability of sharing gradient for data protection of diffusion models, even high-resolution images can be reconstructed with high quality.
△ Less
Submitted 30 May, 2024;
originally announced May 2024.
-
SFDDM: Single-fold Distillation for Diffusion models
Authors:
Chi Hong,
Jiyue Huang,
Robert Birke,
Dick Epema,
Stefanie Roos,
Lydia Y. Chen
Abstract:
While diffusion models effectively generate remarkable synthetic images, a key limitation is the inference inefficiency, requiring numerous sampling steps. To accelerate inference and maintain high-quality synthesis, teacher-student distillation is applied to compress the diffusion models in a progressive and binary manner by retraining, e.g., reducing the 1024-step model to a 128-step model in 3…
▽ More
While diffusion models effectively generate remarkable synthetic images, a key limitation is the inference inefficiency, requiring numerous sampling steps. To accelerate inference and maintain high-quality synthesis, teacher-student distillation is applied to compress the diffusion models in a progressive and binary manner by retraining, e.g., reducing the 1024-step model to a 128-step model in 3 folds. In this paper, we propose a single-fold distillation algorithm, SFDDM, which can flexibly compress the teacher diffusion model into a student model of any desired step, based on reparameterization of the intermediate inputs from the teacher model. To train the student diffusion, we minimize not only the output distance but also the distribution of the hidden variables between the teacher and student model. Extensive experiments on four datasets demonstrate that our student model trained by the proposed SFDDM is able to sample high-quality data with steps reduced to as little as approximately 1%, thus, trading off inference time. Our remarkable performance highlights that SFDDM effectively transfers knowledge in single-fold distillation, achieving semantic consistency and meaningful image interpolation.
△ Less
Submitted 23 May, 2024;
originally announced May 2024.
-
Payout Races and Congested Channels: A Formal Analysis of Security in the Lightning Network
Authors:
Ben Weintraub,
Satwik Prabhu Kumble,
Cristina Nita-Rotaru,
Stefanie Roos
Abstract:
The Lightning Network, a payment channel network with a market cap of over 192M USD, is designed to resolve Bitcoin's scalability issues through fast off-chain transactions. There are multiple Lightning Network client implementations, all of which conform to the same textual specifications known as BOLTs. Several vulnerabilities have been manually discovered, but to-date there have been few works…
▽ More
The Lightning Network, a payment channel network with a market cap of over 192M USD, is designed to resolve Bitcoin's scalability issues through fast off-chain transactions. There are multiple Lightning Network client implementations, all of which conform to the same textual specifications known as BOLTs. Several vulnerabilities have been manually discovered, but to-date there have been few works systematically analyzing the security of the Lightning Network.
In this work, we take a foundational approach to analyzing the security of the Lightning Network with the help of formal methods. Based on the BOLTs' specifications, we build a detailed formal model of the Lightning Network's single-hop payment protocol and verify it using the Spin model checker. Our model captures both concurrency and error semantics of the payment protocol. We then define several security properties which capture the correct intermediate operation of the protocol, ensuring that the outcome is always certain to both channel peers, and using them we re-discover a known attack previously reported in the literature along with a novel attack, referred to as a Payout Race. A Payout Race consists of a particular sequence of events that can lead to an ambiguity in the protocol in which innocent users can unwittingly lose funds. We confirm the practicality of this attack by reproducing it in a local testbed environment.
△ Less
Submitted 3 May, 2024;
originally announced May 2024.
-
Game-Theoretic Analysis of (Non-)Refundable Fees in the Lightning Network
Authors:
Satwik Prabhu Kumble,
Dick Epema,
Stefanie Roos
Abstract:
In PCNs, nodes that forward payments between a source and a receiver are paid a small fee if the payment is successful. The fee is a compensation for temporarily committing funds to the payment. However, payments may fail due to insufficient funds or attacks, often after considerable delays of up to several days, leaving a node without compensation. Furthermore, attackers can intentionally cause f…
▽ More
In PCNs, nodes that forward payments between a source and a receiver are paid a small fee if the payment is successful. The fee is a compensation for temporarily committing funds to the payment. However, payments may fail due to insufficient funds or attacks, often after considerable delays of up to several days, leaving a node without compensation. Furthermore, attackers can intentionally cause failed payments, e.g., to infer private information (like channel balances), without any cost in fees. In this paper, we first use extensive form games to formally characterize the conditions that lead to rational intermediaries refusing (or agreeing) to forward payments. A decision made by an intermediary to forward or not depends on the probability of failure, which they approximate based on past experience. We then propose and analyze an alternative fee model that allows the sender to determine and pay a fraction of the fee to intermediaries in a non refundable manner. A rational sender chooses the fraction such that the intermediaries' utility for forwarding the payment exceeds their utility for not forwarding. Our simulation study, based on real world Lightning snapshots, confirms that our novel mechanism can increase the probability of successful payments by 12 percent and decrease routing fees for senders by about 6 percent if all nodes behave rationally. Furthermore, previously cost free probing attacks now require that the attacker pays 1500 satoshis for every 1 million satoshis inferred. Finally, we propose a modification to the Hash Time Locked Contract to enable secure payments of the non refundable fees.
△ Less
Submitted 6 October, 2023;
originally announced October 2023.
-
ALBERTI, a Multilingual Domain Specific Language Model for Poetry Analysis
Authors:
Javier de la Rosa,
Álvaro P��rez Pozo,
Salvador Ros,
Elena González-Blanco
Abstract:
The computational analysis of poetry is limited by the scarcity of tools to automatically analyze and scan poems. In a multilingual settings, the problem is exacerbated as scansion and rhyme systems only exist for individual languages, making comparative studies very challenging and time consuming. In this work, we present \textsc{Alberti}, the first multilingual pre-trained large language model f…
▽ More
The computational analysis of poetry is limited by the scarcity of tools to automatically analyze and scan poems. In a multilingual settings, the problem is exacerbated as scansion and rhyme systems only exist for individual languages, making comparative studies very challenging and time consuming. In this work, we present \textsc{Alberti}, the first multilingual pre-trained large language model for poetry. Through domain-specific pre-training (DSP), we further trained multilingual BERT on a corpus of over 12 million verses from 12 languages. We evaluated its performance on two structural poetry tasks: Spanish stanza type classification, and metrical pattern prediction for Spanish, English and German. In both cases, \textsc{Alberti} outperforms multilingual BERT and other transformers-based models of similar sizes, and even achieves state-of-the-art results for German when compared to rule-based systems, demonstrating the feasibility and effectiveness of DSP in the poetry domain.
△ Less
Submitted 3 July, 2023;
originally announced July 2023.
-
Blockchain Nodes are Heterogeneous and Your P2P Overlay Should be Too: PODS
Authors:
Naqib Zarin,
Isaac Sheff,
Stefanie Roos
Abstract:
At the core of each blockchain system, parties communicate through a peer-to-peer (P2P) overlay. Unfortunately, recent evidence suggests these P2P overlays represent a significant bottleneck for transaction throughput and scalability. Furthermore, they enable a number of attacks. We argue that these performance and security problems arise because current P2P overlays cannot fully capture the compl…
▽ More
At the core of each blockchain system, parties communicate through a peer-to-peer (P2P) overlay. Unfortunately, recent evidence suggests these P2P overlays represent a significant bottleneck for transaction throughput and scalability. Furthermore, they enable a number of attacks. We argue that these performance and security problems arise because current P2P overlays cannot fully capture the complexity of a blockchain system as they do not offer flexibility to accommodate node heterogeneity. We propose a novel approach to address these issues: P2P Overlay Domains with Sovereignty (PODS), which allows nodes in a single overlay to belong to multiple heterogeneous groups, called domains. Each domain features its own set of protocols, tailored to the characteristics and needs of its nodes. To demonstrate the effectiveness of PODS, we design and implement two novel node discovery protocols: FedKad and SovKad. Using a custom simulator, we show that node discovery using PODS (SovKad) architecture outperforms both single overlay (Kademlia) and multi-overlay (FedKad) architectures in terms of hop count and success rate, though FedKad requires slightly less bandwidth.
△ Less
Submitted 28 June, 2023;
originally announced June 2023.
-
LyricSIM: A novel Dataset and Benchmark for Similarity Detection in Spanish Song LyricS
Authors:
Alejandro Benito-Santos,
Adrián Ghajari,
Pedro Hernández,
Víctor Fresno,
Salvador Ros,
Elena González-Blanco
Abstract:
In this paper, we present a new dataset and benchmark tailored to the task of semantic similarity in song lyrics. Our dataset, originally consisting of 2775 pairs of Spanish songs, was annotated in a collective annotation experiment by 63 native annotators. After collecting and refining the data to ensure a high degree of consensus and data integrity, we obtained 676 high-quality annotated pairs t…
▽ More
In this paper, we present a new dataset and benchmark tailored to the task of semantic similarity in song lyrics. Our dataset, originally consisting of 2775 pairs of Spanish songs, was annotated in a collective annotation experiment by 63 native annotators. After collecting and refining the data to ensure a high degree of consensus and data integrity, we obtained 676 high-quality annotated pairs that were used to evaluate the performance of various state-of-the-art monolingual and multilingual language models. Consequently, we established baseline results that we hope will be useful to the community in all future academic and industrial applications conducted in this context.
△ Less
Submitted 2 June, 2023;
originally announced June 2023.
-
A Deployment-First Methodology to Mechanism Design and Refinement in Distributed Systems
Authors:
Martijn de Vos,
Georgy Ishmaev,
Johan Pouwelse,
Stefanie Roos
Abstract:
Catalyzed by the popularity of blockchain technology, there has recently been a renewed interest in the design, implementation and evaluation of decentralized systems. Most of these systems are intended to be deployed at scale and in heterogeneous environments with real users and unpredictable workloads. Nevertheless, most research in this field evaluates such systems in controlled environments th…
▽ More
Catalyzed by the popularity of blockchain technology, there has recently been a renewed interest in the design, implementation and evaluation of decentralized systems. Most of these systems are intended to be deployed at scale and in heterogeneous environments with real users and unpredictable workloads. Nevertheless, most research in this field evaluates such systems in controlled environments that poorly reflect the complex conditions of real-world environments. In this work, we argue that deployment is crucial to understanding decentralized mechanisms in a real-world environment and an enabler to building more robust and sustainable systems. We highlight the merits of deployment by comparing this approach with other experimental setups and show how our lab applied a deployment-first methodology. We then outline how we use Tribler, our peer-to-peer file-sharing application, to deploy and monitor decentralized mechanisms at scale. We illustrate the application of our methodology by describing a deployment trial in experimental tokenomics. Finally, we summarize four lessons learned from multiple deployment trials where we applied our methodology.
△ Less
Submitted 11 January, 2023;
originally announced January 2023.
-
SyncPCN/PSyncPCN: Payment Channel Networks without Blockchain Synchrony
Authors:
Oğuzhan Ersoy,
Jérémie Decouchant,
Satwik Prabhu Kimble,
Stefanie Roos
Abstract:
Payment channel networks (PCNs) enhance the scalability of blockchains by allowing parties to conduct transactions off-chain, i.e, without broadcasting every transaction to all blockchain participants. To conduct transactions, a sender and a receiver can either establish a direct payment channel with a funding blockchain transaction or leverage existing channels in a multi-hop payment. The securit…
▽ More
Payment channel networks (PCNs) enhance the scalability of blockchains by allowing parties to conduct transactions off-chain, i.e, without broadcasting every transaction to all blockchain participants. To conduct transactions, a sender and a receiver can either establish a direct payment channel with a funding blockchain transaction or leverage existing channels in a multi-hop payment. The security of PCNs usually relies on the synchrony of the underlying blockchain, i.e., evidence of misbehavior needs to be published on the blockchain within a time limit. Alternative payment channel proposals that do not require blockchain synchrony rely on quorum certificates and use a committee to register the transactions of a channel. However, these proposals do not support multi-hop payments, a limitation we aim to overcome. In this paper, we demonstrate that it is in fact impossible to design a multi-hop payment protocol with both network asynchrony and faulty channels, i.e., channels that may not correctly follow the protocol. We then detail two committee-based multi-hop payment protocols that respectively assume synchronous communications and possibly faulty channels, or asynchronous communication and correct channels. The first protocol relies on possibly faulty committees instead of the blockchain to resolve channel disputes, and enforces privacy properties within a synchronous network. The second one relies on committees that contain at most f faulty members out of 3f+1 and successively delegate to each other the role of eventually completing a multi-hop payment. We show that both protocols satisfy the security requirements of a multi-hop payment and compare their communication complexity and latency.
△ Less
Submitted 4 August, 2022; v1 submitted 23 July, 2022;
originally announced July 2022.
-
Combining Visual Saliency Methods and Sparse Keypoint Annotations to Providently Detect Vehicles at Night
Authors:
Lukas Ewecker,
Lars Ohnemus,
Robin Schwager,
Stefan Roos,
Tim Brühl,
Sascha Saralajew
Abstract:
Provident detection of other road users at night has the potential for increasing road safety. For this purpose, humans intuitively use visual cues, such as light cones and light reflections emitted by other road users to be able to react to oncoming traffic at an early stage. This behavior can be imitated by computer vision methods by predicting the appearance of vehicles based on emitted light r…
▽ More
Provident detection of other road users at night has the potential for increasing road safety. For this purpose, humans intuitively use visual cues, such as light cones and light reflections emitted by other road users to be able to react to oncoming traffic at an early stage. This behavior can be imitated by computer vision methods by predicting the appearance of vehicles based on emitted light reflections caused by the vehicle's headlights. Since current object detection algorithms are mainly based on detecting directly visible objects annotated via bounding boxes, the detection and annotation of light reflections without sharp boundaries is challenging. For this reason, the extensive open-source dataset PVDN (Provident Vehicle Detection at Night) was published, which includes traffic scenarios at night with light reflections annotated via keypoints. In this paper, we explore the potential of saliency-based approaches to create different object representations based on the visual saliency and sparse keypoint annotations of the PVDN dataset. For that, we extend the general idea of Boolean map saliency towards a context-aware approach by taking into consideration sparse keypoint annotations by humans. We show that this approach allows for an automated derivation of different object representations, such as binary maps or bounding boxes so that detection models can be trained on different annotation variants and the problem of providently detecting vehicles at night can be tackled from different perspectives. With that, we provide further powerful tools and methods to study the problem of detecting vehicles at night before they are actually visible.
△ Less
Submitted 27 September, 2022; v1 submitted 25 April, 2022;
originally announced April 2022.
-
Fabricated Flips: Poisoning Federated Learning without Data
Authors:
Jiyue Huang,
Zilong Zhao,
Lydia Y. Chen,
Stefanie Roos
Abstract:
Attacks on Federated Learning (FL) can severely reduce the quality of the generated models and limit the usefulness of this emerging learning paradigm that enables on-premise decentralized learning. However, existing untargeted attacks are not practical for many scenarios as they assume that i) the attacker knows every update of benign clients, or ii) the attacker has a large dataset to locally tr…
▽ More
Attacks on Federated Learning (FL) can severely reduce the quality of the generated models and limit the usefulness of this emerging learning paradigm that enables on-premise decentralized learning. However, existing untargeted attacks are not practical for many scenarios as they assume that i) the attacker knows every update of benign clients, or ii) the attacker has a large dataset to locally train updates imitating benign parties. In this paper, we propose a data-free untargeted attack (DFA) that synthesizes malicious data to craft adversarial models without eavesdropping on the transmission of benign clients at all or requiring a large quantity of task-specific training data. We design two variants of DFA, namely DFA-R and DFA-G, which differ in how they trade off stealthiness and effectiveness. Specifically, DFA-R iteratively optimizes a malicious data layer to minimize the prediction confidence of all outputs of the global model, whereas DFA-G interactively trains a malicious data generator network by steering the output of the global model toward a particular class. Experimental results on Fashion-MNIST, Cifar-10, and SVHN show that DFA, despite requiring fewer assumptions than existing attacks, achieves similar or even higher attack success rate than state-of-the-art untargeted attacks against various state-of-the-art defense mechanisms. Concretely, they can evade all considered defense mechanisms in at least 50% of the cases for CIFAR-10 and often reduce the accuracy by more than a factor of 2. Consequently, we design REFD, a defense specifically crafted to protect against data-free attacks. REFD leverages a reference dataset to detect updates that are biased or have a low confidence. It greatly improves upon existing defenses by filtering out the malicious updates and achieves high global model accuracy
△ Less
Submitted 2 August, 2023; v1 submitted 7 February, 2022;
originally announced February 2022.
-
A Sybil-Resistant and Decentralized Market Place
Authors:
Naqib Zarin,
Dirk van Bokkem,
Justin Segond,
Stefanie Roos
Abstract:
Existing centralised market places such as Ebay enable companies to gather large amounts of personal data that can be used to manipulate users. Furthermore, users can frequently perform fraud without severe consequence. Reputation systems only solve this problem partially as malicious users can re-join the network with a new identity if their reputation is too low. By performing a Sybil attack, i.…
▽ More
Existing centralised market places such as Ebay enable companies to gather large amounts of personal data that can be used to manipulate users. Furthermore, users can frequently perform fraud without severe consequence. Reputation systems only solve this problem partially as malicious users can re-join the network with a new identity if their reputation is too low. By performing a Sybil attack, i.e., joining with multiple seemingly distinct identities, malicious participants can further boost their own reputation. In this paper, we present MarketPalace. MarketPalace relies on a peer-to-peer infrastructure to realize a decentralized market place during trading. Only when registering, users communicate with a central server to verify that they are not Sybils. More concretely, our system leverages self-sovereign identity to detect and undermine repeated joins by the same user. We implemented MarketPalace and demonstrated its feasibility for small regional markets.
△ Less
Submitted 25 January, 2022;
originally announced January 2022.
-
Attacks and Defenses for Free-Riders in Multi-Discriminator GAN
Authors:
Zilong Zhao,
Jiyue Huang,
Stefanie Roos,
Lydia Y. Chen
Abstract:
Generative Adversarial Networks (GANs) are increasingly adopted by the industry to synthesize realistic images. Due to data not being centrally available, Multi-Discriminator (MD)-GANs training framework employs multiple discriminators that have direct access to the real data. Distributedly training a joint GAN model entails the risk of free-riders, i.e., participants that aim to benefit from the…
▽ More
Generative Adversarial Networks (GANs) are increasingly adopted by the industry to synthesize realistic images. Due to data not being centrally available, Multi-Discriminator (MD)-GANs training framework employs multiple discriminators that have direct access to the real data. Distributedly training a joint GAN model entails the risk of free-riders, i.e., participants that aim to benefit from the common model while only pretending to participate in the training process. In this paper, we conduct the first characterization study of the impact of free-riders on MD-GAN. Based on two production prototypes of MD-GAN, we find that free-riders drastically reduce the ability of MD-GANs to produce images that are indistinguishable from real data, i.e., they increase the FID score -- the standard measure to assess the quality of generated images. To mitigate the model degradation, we propose a defense strategy against free-riders in MD-GAN, termed DFG. DFG distinguishes free-riders and benign participants through periodic probing and clustering of discriminators' responses based on a reference response of free-riders, which then allows the generator to exclude the detected free-riders from the training. Furthermore, we extend our defense, termed DFG+, to enable discriminators to filter out free-riders at the variant of MD-GAN that allows peer exchanges of discriminators networks. Extensive evaluation on various scenarios of free-riders, MD-GAN architecture, and three datasets show that our defenses effectively detect free-riders. With 1 to 5 free-riders, DFG and DFG+ averagely decreases FID by 5.22% to 11.53% for CIFAR10 and 5.79% to 13.22% for CIFAR100 in comparison to an attack without defense. In a shell, the proposed DFG(+) can effectively defend against free-riders without affecting benign clients at a negligible computation overhead.
△ Less
Submitted 24 January, 2022;
originally announced January 2022.
-
RIS-Aware Indoor Network Planning: The Rennes Railway Station Case
Authors:
Antonio Albanese,
Guillermo Encinas-Lago,
Vincenzo Sciancalepore,
Xavier Costa-Pérez,
Dinh-Thuy Phan-Huy,
Stéphane Ros
Abstract:
Future generations of wireless networks will offer newfangled performance via unprecedented solutions: the metasurface innovation will drive such a revolution by posing control onto the surrounding propagation environment, always portrayed as a tamper-proof black-box. The RIS technology, envisioned as the discrete version of a metasurface, can be dynamically configured to alter the propagation pro…
▽ More
Future generations of wireless networks will offer newfangled performance via unprecedented solutions: the metasurface innovation will drive such a revolution by posing control onto the surrounding propagation environment, always portrayed as a tamper-proof black-box. The RIS technology, envisioned as the discrete version of a metasurface, can be dynamically configured to alter the propagation properties of the impinging signals by, e.g., steering the corresponding beams towards defined directions. This will unlock new application opportunities and deliver advanced end-user services. However, this fascinating solution comes at not negligible costs: RIS require ad-hoc design, deployment and management operations to be fully exploited. In this paper, we tackle the RIS placement problem from a theoretical viewpoint, showcasing a large-scale solution on synthetic topologies to improve communication performance while solving the "dead-zone" problem. Additionally, our mathematical framework is empirically validated within a realistic indoor scenario, the Rennes railway station, showing how a complex indoor propagation environment can be fully disciplined by an advanced RIS installation.
△ Less
Submitted 23 March, 2022; v1 submitted 19 January, 2022;
originally announced January 2022.
-
Topology Inference of Networks utilizing Rooted Spanning Tree Embeddings
Authors:
Martin Byrenheid,
Stefanie Roos,
Thorsten Strufe
Abstract:
Due to its high efficiency, routing based on greedy embeddings of rooted spanning trees is a promising approach for dynamic, large-scale networks with restricted topologies. Friend-to-friend (F2F) overlays, one key application of embedding-based routing, aim to prevent disclosure of their participants to malicious members by restricting exchange of messages to mutually trusted nodes. Since embeddi…
▽ More
Due to its high efficiency, routing based on greedy embeddings of rooted spanning trees is a promising approach for dynamic, large-scale networks with restricted topologies. Friend-to-friend (F2F) overlays, one key application of embedding-based routing, aim to prevent disclosure of their participants to malicious members by restricting exchange of messages to mutually trusted nodes. Since embeddings assign a unique integer vector to each node that encodes its position in a spanning tree of the overlay, attackers can infer network structure from knowledge about assigned vectors. As this information can be used to identify participants, an evaluation of the scale of leakage is needed. In this work, we analyze in detail which information malicious participants can infer from knowledge about assigned vectors. Also, we show that by monitoring packet trajectories, malicious participants cannot unambiguously infer links between nodes of unidentified participants. Using simulation, we find that the vector assignment procedure has a strong impact on the feasibility of inference. In F2F overlay networks, using vectors of randomly chosen numbers for routing decreases the mean number of discovered individuals by one order of magnitude compared to the popular approach of using child enumeration indexes as vector elements.
△ Less
Submitted 15 November, 2021; v1 submitted 9 August, 2021;
originally announced August 2021.
-
How Lightning's Routing Diminishes its Anonymity
Authors:
Satwik Prabhu Kumble,
Dick Epema,
Stefanie Roos
Abstract:
The system shows the error of "Bad character(s) in field Abstract" for no reason. Please refer to manuscript for the full abstract
The system shows the error of "Bad character(s) in field Abstract" for no reason. Please refer to manuscript for the full abstract
△ Less
Submitted 21 July, 2021;
originally announced July 2021.
-
Is Shapley Value fair? Improving Client Selection for Mavericks in Federated Learning
Authors:
Jiyue Huang,
Chi Hong,
Lydia Y. Chen,
Stefanie Roos
Abstract:
Shapley Value is commonly adopted to measure and incentivize client participation in federated learning. In this paper, we show -- theoretically and through simulations -- that Shapley Value underestimates the contribution of a common type of client: the Maverick. Mavericks are clients that differ both in data distribution and data quantity and can be the sole owners of certain types of data. Sele…
▽ More
Shapley Value is commonly adopted to measure and incentivize client participation in federated learning. In this paper, we show -- theoretically and through simulations -- that Shapley Value underestimates the contribution of a common type of client: the Maverick. Mavericks are clients that differ both in data distribution and data quantity and can be the sole owners of certain types of data. Selecting the right clients at the right moment is important for federated learning to reduce convergence times and improve accuracy. We propose FedEMD, an adaptive client selection strategy based on the Wasserstein distance between the local and global data distributions. As FedEMD adapts the selection probability such that Mavericks are preferably selected when the model benefits from improvement on rare classes, it consistently ensures the fast convergence in the presence of different types of Mavericks. Compared to existing strategies, including Shapley Value-based ones, FedEMD improves the convergence of neural network classifiers by at least 26.9% for FedAvg aggregation compared with the state of the art.
△ Less
Submitted 20 June, 2021;
originally announced June 2021.
-
A Dataset for Provident Vehicle Detection at Night
Authors:
Sascha Saralajew,
Lars Ohnemus,
Lukas Ewecker,
Ebubekir Asan,
Simon Isele,
Stefan Roos
Abstract:
In current object detection, algorithms require the object to be directly visible in order to be detected. As humans, however, we intuitively use visual cues caused by the respective object to already make assumptions about its appearance. In the context of driving, such cues can be shadows during the day and often light reflections at night. In this paper, we study the problem of how to map this…
▽ More
In current object detection, algorithms require the object to be directly visible in order to be detected. As humans, however, we intuitively use visual cues caused by the respective object to already make assumptions about its appearance. In the context of driving, such cues can be shadows during the day and often light reflections at night. In this paper, we study the problem of how to map this intuitive human behavior to computer vision algorithms to detect oncoming vehicles at night just from the light reflections they cause by their headlights. For that, we present an extensive open-source dataset containing 59746 annotated grayscale images out of 346 different scenes in a rural environment at night. In these images, all oncoming vehicles, their corresponding light objects (e.g., headlamps), and their respective light reflections (e.g., light reflections on guardrails) are labeled. In this context, we discuss the characteristics of the dataset and the challenges in objectively describing visual cues such as light reflections. We provide different metrics for different ways to approach the task and report the results we achieved using state-of-the-art and custom object detection models as a first benchmark. With that, we want to bring attention to a new and so far neglected field in computer vision research, encourage more researchers to tackle the problem, and thereby further close the gap between human performance and computer vision systems.
△ Less
Submitted 11 August, 2021; v1 submitted 27 May, 2021;
originally announced May 2021.
-
Provident Vehicle Detection at Night: The PVDN Dataset
Authors:
Lars Ohnemus,
Lukas Ewecker,
Ebubekir Asan,
Stefan Roos,
Simon Isele,
Jakob Ketterer,
Leopold Müller,
Sascha Saralajew
Abstract:
For advanced driver assistance systems, it is crucial to have information about oncoming vehicles as early as possible. At night, this task is especially difficult due to poor lighting conditions. For that, during nighttime, every vehicle uses headlamps to improve sight and therefore ensure safe driving. As humans, we intuitively assume oncoming vehicles before the vehicles are actually physically…
▽ More
For advanced driver assistance systems, it is crucial to have information about oncoming vehicles as early as possible. At night, this task is especially difficult due to poor lighting conditions. For that, during nighttime, every vehicle uses headlamps to improve sight and therefore ensure safe driving. As humans, we intuitively assume oncoming vehicles before the vehicles are actually physically visible by detecting light reflections caused by their headlamps. In this paper, we present a novel dataset containing 59746 annotated grayscale images out of 346 different scenes in a rural environment at night. In these images, all oncoming vehicles, their corresponding light objects (e.g., headlamps), and their respective light reflections (e.g., light reflections on guardrails) are labeled. This is accompanied by an in-depth analysis of the dataset characteristics. With that, we are providing the first open-source dataset with comprehensive ground truth data to enable research into new methods of detecting oncoming vehicles based on the light reflections they cause, long before they are directly visible. We consider this as an essential step to further close the performance gap between current advanced driver assistance systems and human behavior.
△ Less
Submitted 23 January, 2021; v1 submitted 30 December, 2020;
originally announced December 2020.
-
The Merchant: Avoiding Payment Channel Depletion through Incentives
Authors:
Yuup van Engelshoven,
Stefanie Roos
Abstract:
Payment channels networks drastically increase the throughput and hence scalability of blockchains by performing transactions \emph{off-chain}. In an off-chain payment, parties deposit coins in a channel and then perform transactions without invoking the global consensus mechanism of the blockchain. However, the transaction value is limited by the capacity of the channel, i.e., the amount of funds…
▽ More
Payment channels networks drastically increase the throughput and hence scalability of blockchains by performing transactions \emph{off-chain}. In an off-chain payment, parties deposit coins in a channel and then perform transactions without invoking the global consensus mechanism of the blockchain. However, the transaction value is limited by the capacity of the channel, i.e., the amount of funds available on a channel. These funds decrease when a transaction is sent and increase when a transaction is received on the channel. Recent research indicates that there is an imbalance between sending and receiving transactions, which leads to channel depletion in the sense that one of these operations becomes impossible over time due to the lack of available funds.
We incentivize the balanced use of payment channels through fees. Whereas the current fee model depends solely on the transaction value, our fee policies encourage transactions that have a positive effect on the balance in a channel and discourage those that have a negative effect. This paper first defines necessary properties of fee strategies. Then, it introduces two novel fees strategies that provably satisfy all necessary properties. Our extensive simulation study reveals that these incentives increase the effectiveness of payments by $8\%$ to $19\%$.
△ Less
Submitted 18 December, 2020;
originally announced December 2020.
-
Predicting metrical patterns in Spanish poetry with language models
Authors:
Javier de la Rosa,
Salvador Ros,
Elena González-Blanco
Abstract:
In this paper, we compare automated metrical pattern identification systems available for Spanish against extensive experiments done by fine-tuning language models trained on the same task. Despite being initially conceived as a model suitable for semantic tasks, our results suggest that BERT-based models retain enough structural information to perform reasonably well for Spanish scansion.
In this paper, we compare automated metrical pattern identification systems available for Spanish against extensive experiments done by fine-tuning language models trained on the same task. Despite being initially conceived as a model suitable for semantic tasks, our results suggest that BERT-based models retain enough structural information to perform reasonably well for Spanish scansion.
△ Less
Submitted 18 November, 2020;
originally announced November 2020.
-
An Exploratory Analysis on Users' Contributions in Federated Learning
Authors:
Jiyue Huang,
Rania Talbi,
Zilong Zhao,
Sara Boucchenak,
Lydia Y. Chen,
Stefanie Roos
Abstract:
Federated Learning is an emerging distributed collaborative learning paradigm adopted by many of today's applications, e.g., keyboard prediction and object recognition. Its core principle is to learn from large amount of users data while preserving data privacy by design as collaborative users only need to share the machine learning models and keep data locally. The main challenge for such systems…
▽ More
Federated Learning is an emerging distributed collaborative learning paradigm adopted by many of today's applications, e.g., keyboard prediction and object recognition. Its core principle is to learn from large amount of users data while preserving data privacy by design as collaborative users only need to share the machine learning models and keep data locally. The main challenge for such systems is to provide incentives to users to contribute high-quality models trained from their local data. In this paper, we aim to answer how well incentives recognize (in)accurate local models from honest and malicious users, and perceive their impacts on the model accuracy of federated learning systems. We first present a thorough survey on two contrasting perspectives: incentive mechanisms to measure the contribution of local models by honest users, and malicious users to deliberately degrade the overall model. We conduct simulation experiments to empirically demonstrate if existing contribution measurement schemes can disclose low-quality models from malicious users. Our results show there exists a clear tradeoff among measurement schemes in terms of the computational efficiency and effectiveness to distill the impact of malicious participants. We conclude this paper by discussing the research directions to design resilient contribution incentives.
△ Less
Submitted 13 November, 2020;
originally announced November 2020.
-
Structural Attacks on Local Routing in Payment Channel Networks
Authors:
Ben Weintraub,
Cristina Nita-Rotaru,
Stefanie Roos
Abstract:
Payment channel networks (PCN) enable scalable blockchain transactions without fundamentally changing the underlying distributed ledger algorithm. However, routing a payment via multiple channels in a PCN requires locking collateral for potentially long periods of time. Adversaries can abuse this mechanism to conduct denial-of-service attacks. Previous work focused on source routing, which is unli…
▽ More
Payment channel networks (PCN) enable scalable blockchain transactions without fundamentally changing the underlying distributed ledger algorithm. However, routing a payment via multiple channels in a PCN requires locking collateral for potentially long periods of time. Adversaries can abuse this mechanism to conduct denial-of-service attacks. Previous work focused on source routing, which is unlikely to remain a viable routing approach as these networks grow.
In this work, we examine the effectiveness of attacks in PCNs that use routing algorithms based on local knowledge, where compromised intermediate nodes can delay or drop transactions to create denial-of-service. We focus on SpeedyMurmurs as a representative of such protocols. We identify two attacker node selection strategies; one based on the position in the routing tree, and the other on betweenness centrality. Our simulation-driven study shows that while they are both effective, the centrality-based attack approaches near-optimal effectiveness. We also show that the attacks are ineffective in less centralized networks and discuss incentives for the participants in PCNs to create less centralized topologies through the payment channels they establish among themselves.
△ Less
Submitted 7 September, 2021; v1 submitted 17 July, 2020;
originally announced July 2020.
-
DISCO PAL: Diachronic Spanish Sonnet Corpus with Psychological and Affective Labels
Authors:
Alberto Barbado,
Víctor Fresno,
Ángeles Manjarrés Riesco,
Salvador Ros
Abstract:
Nowadays, there are many applications of text mining over corpora from different languages. However, most of them are based on texts in prose, lacking applications that work with poetry texts. An example of an application of text mining in poetry is the usage of features derived from their individual words in order to capture the lexical, sublexical and interlexical meaning, and infer the General…
▽ More
Nowadays, there are many applications of text mining over corpora from different languages. However, most of them are based on texts in prose, lacking applications that work with poetry texts. An example of an application of text mining in poetry is the usage of features derived from their individual words in order to capture the lexical, sublexical and interlexical meaning, and infer the General Affective Meaning (GAM) of the text. However, even though this proposal has been proved as useful for poetry in some languages, there is a lack of studies for both Spanish poetry and for highly-structured poetic compositions such as sonnets. This article presents a study over an annotated corpus of Spanish sonnets, in order to analyse if it is possible to build features from their individual words for predicting their GAM. The purpose of this is to model sonnets at an affective level. The article also analyses the relationship between the GAM of the sonnets and the content itself. For this, we consider the content from a psychological perspective, identifying with tags when a sonnet is related to a specific term. Then, we study how GAM changes according to each of those psychological terms.
The corpus used contains 274 Spanish sonnets from authors of different centuries, from 15th to 19th. This corpus was annotated by different domain experts. The experts annotated the poems with affective and lexico-semantic features, as well as with domain concepts that belong to psychology. Thanks to this, the corpus of sonnets can be used in different applications, such as poetry recommender systems, personality text mining studies of the authors, or the usage of poetry for therapeutic purposes.
△ Less
Submitted 20 June, 2021; v1 submitted 9 July, 2020;
originally announced July 2020.
-
How to profit from payments channels
Authors:
Oguzhan Ersoy,
Stefanie Roos,
Zekeriya Erkin
Abstract:
Payment channel networks like Bitcoin's Lightning network are an auspicious approach for realizing high transaction throughput and almost-instant confirmations in blockchain networks. However, the ability to successfully make payments in such networks relies on the willingness of participants to lock collateral in the network. In Lightning, the key financial incentive is to lock collateral are sma…
▽ More
Payment channel networks like Bitcoin's Lightning network are an auspicious approach for realizing high transaction throughput and almost-instant confirmations in blockchain networks. However, the ability to successfully make payments in such networks relies on the willingness of participants to lock collateral in the network. In Lightning, the key financial incentive is to lock collateral are small fees for routing payments for other participants. While users can choose these fees, currently, they mainly stick to the default fees. By providing insights on beneficial choices for fees, we aim to incentivize users to lock more collateral and improve the effectiveness of the network.
In this paper, we consider a node $\mathbf{A}$ that given the network topology and the channel details selects where to establish channels and how much fee to charge such that its financial gain is maximized. We formalize the optimization problem and show that it is NP-hard. We design a greedy algorithm to approximate the optimal solution. In each step, our greedy algorithm selects a node which maximizes the total reward concerning the number of shortest paths passing through $\mathbf{A}$ and channel fees. Our simulation study leverages real-world data set to quantify the impact of our gain optimization and indicates that our strategy is at least a factor two better than other strategies.
△ Less
Submitted 25 November, 2019; v1 submitted 20 November, 2019;
originally announced November 2019.
-
The Chiral Anomaly of the Free Fermion in Functorial Field Theory
Authors:
Matthias Ludewig,
Saskia Roos
Abstract:
When trying to cast the free fermion in the framework of functorial field theory, its chiral anomaly manifests in the fact that it assigns the determinant of the Dirac operator to a top-dimensional closed spin manifold, which is not a number as expected, but an element of a complex line. In functorial field theory language, this means that the theory is twisted, which gives rise to an anomaly theo…
▽ More
When trying to cast the free fermion in the framework of functorial field theory, its chiral anomaly manifests in the fact that it assigns the determinant of the Dirac operator to a top-dimensional closed spin manifold, which is not a number as expected, but an element of a complex line. In functorial field theory language, this means that the theory is twisted, which gives rise to an anomaly theory. In this paper, we give a detailed construction of this anomaly theory, as a functor that sends manifolds to infinite-dimensional Clifford algebras and bordisms to bimodules.
△ Less
Submitted 18 February, 2022; v1 submitted 9 September, 2019;
originally announced September 2019.
-
Attack-resistant Spanning Tree Construction in Route-Restricted Overlay Networks
Authors:
Martin Byrenheid,
Stefanie Roos,
Thorsten Strufe
Abstract:
Nodes in route-restricted overlays have an immutable set of neighbors, explicitly specified by their users. Popular examples include payment networks such as the Lightning network as well as social overlays such as the Dark Freenet. Routing algorithms are central to such overlays as they enable communication between nodes that are not directly connected. Recent results show that algorithms based o…
▽ More
Nodes in route-restricted overlays have an immutable set of neighbors, explicitly specified by their users. Popular examples include payment networks such as the Lightning network as well as social overlays such as the Dark Freenet. Routing algorithms are central to such overlays as they enable communication between nodes that are not directly connected. Recent results show that algorithms based on spanning trees are the most promising provably efficient choice. However, all suggested solutions fail to address how distributed spanning tree algorithms can deal with active denial of service attacks by malicious nodes. In this work, we design a novel self-stabilizing spanning tree construction algorithm that utilizes cryptographic signatures and prove that it reduces the set of nodes affected by active attacks. Our simulations substantiate this theoretical result with concrete values based on real-world data sets. In particular, our results indicate that our algorithm reduces the number of affected nodes by up to 74% compared to state-of-the-art attack-resistant spanning tree constructions.
△ Less
Submitted 9 August, 2019; v1 submitted 9 January, 2019;
originally announced January 2019.
-
Scalar curvature and the multiconformal class of a direct product Riemannian manifold
Authors:
Nobuhiko Otoba,
Saskia Roos
Abstract:
For a closed, connected direct product Riemannian manifold $(M, g)=(M_1\times\cdots\times M_l, g_1\oplus\cdots\oplus g_l)$, we define its multiconformal class $ [\![ g ]\!]$ as the totality $\{f_1^2g_1\oplus \cdots\oplus f_l^2g_l\}$ of all Riemannian metrics obtained from multiplying the metric $g_i$ of each factor $M_i$ by a function $f_i^2>0$ on the total space $M$. A multiconformal class…
▽ More
For a closed, connected direct product Riemannian manifold $(M, g)=(M_1\times\cdots\times M_l, g_1\oplus\cdots\oplus g_l)$, we define its multiconformal class $ [\![ g ]\!]$ as the totality $\{f_1^2g_1\oplus \cdots\oplus f_l^2g_l\}$ of all Riemannian metrics obtained from multiplying the metric $g_i$ of each factor $M_i$ by a function $f_i^2>0$ on the total space $M$. A multiconformal class $ [\![ g ]\!]$ contains not only all warped product type deformations of $g$ but also the whole conformal class $[\tilde{g}]$ of every $\tilde{g}\in [\![ g ]\!]$. In this article, we prove that $ [\![ g ]\!]$ carries a metric of positive scalar curvature if and only if the conformal class of some factor $(M_i, g_i)$ does, under the technical assumption $\dim M_i\ge 2$. We also show that, even in the case where every factor $(M_i, g_i)$ has positive scalar curvature, $ [\![ g ]\!]$ carries a metric of scalar curvature constantly equal to $-1$ and with arbitrarily large volume, provided $l\ge 2$ and $\dim M\ge 3$. In this case, such negative scalar curvature metrics within $ [\![ g ]\!]$ for $l=2$ cannot be of any warped product type.
△ Less
Submitted 22 October, 2018; v1 submitted 20 August, 2018;
originally announced August 2018.
-
The Dirac operator under collapse to a smooth limit space
Authors:
Saskia Roos
Abstract:
Let $(M_i, g_i)_{i \in \mathbb{N}}$ be a sequence of spin manifolds with uniform bounded curvature and diameter that converges to a lower dimensional Riemannian manifold $(B,h)$ in the Gromov-Hausdorff topology. Lott showed that the spectrum converges to the spectrum of a certain first order elliptic differential operator $\mathcal{D}$ on $B$. In this article we give an explicit description of…
▽ More
Let $(M_i, g_i)_{i \in \mathbb{N}}$ be a sequence of spin manifolds with uniform bounded curvature and diameter that converges to a lower dimensional Riemannian manifold $(B,h)$ in the Gromov-Hausdorff topology. Lott showed that the spectrum converges to the spectrum of a certain first order elliptic differential operator $\mathcal{D}$ on $B$. In this article we give an explicit description of $\mathcal{D}^B$. We conclude that $\mathcal{D}^B$ is self-adjoint and characterize the special case where $\mathcal{D}^B$ is the Dirac operator on $B$.
△ Less
Submitted 7 May, 2019; v1 submitted 2 February, 2018;
originally announced February 2018.
-
Settling Payments Fast and Private: Efficient Decentralized Routing for Path-Based Transactions
Authors:
Stefanie Roos,
Pedro Moreno-Sanchez,
Aniket Kate,
Ian Goldberg
Abstract:
Path-based transaction (PBT) networks, which settle payments from one user to another via a path of intermediaries, are a growing area of research. They overcome the scalability and privacy issues in cryptocurrencies like Bitcoin and Ethereum by replacing expensive and slow on-chain blockchain operations with inexpensive and fast off-chain transfers. In the form of credit networks such as Ripple a…
▽ More
Path-based transaction (PBT) networks, which settle payments from one user to another via a path of intermediaries, are a growing area of research. They overcome the scalability and privacy issues in cryptocurrencies like Bitcoin and Ethereum by replacing expensive and slow on-chain blockchain operations with inexpensive and fast off-chain transfers. In the form of credit networks such as Ripple and Stellar, they also enable low-price real-time gross settlements across different currencies. For example, SilentWhsipers is a recently proposed fully distributed credit network relying on path-based transactions for secure and in particular private payments without a public ledger. At the core of a decentralized PBT network is a routing algorithm that discovers transaction paths between payer and payee. During the last year, a number of routing algorithms have been proposed. However, the existing ad hoc efforts lack either efficiency or privacy. In this work, we first identify several efficiency concerns in SilentWhsipers. Armed with this knowledge, we design and evaluate SpeedyMurmurs, a novel routing algorithm for decentralized PBT networks using efficient and flexible embedding-based path discovery and on-demand efficient stabilization to handle the dynamics of a PBT network. Our simulation study, based on real-world data from the currently deployed Ripple credit network, indicates that SpeedyMurmurs reduces the overhead of stabilization by up to two orders of magnitude and the overhead of routing a transaction by more than a factor of two. Furthermore, using SpeedyMurmurs maintains at least the same success ratio as decentralized landmark routing, while providing lower delays. Finally, SpeedyMurmurs achieves key privacy goals for routing in PBT networks.
△ Less
Submitted 13 December, 2017; v1 submitted 17 September, 2017;
originally announced September 2017.
-
Dirac operators with $W^{1,\infty}$-potential under codimension one collapse
Authors:
Saskia Roos
Abstract:
We study the behavior of the spectrum of the Dirac operator together with a symmetric $W^{1, \infty}$-potential on spin manifolds under a collapse of codimension one with bounded sectional curvature and diameter. If there is an induced spin structure on the limit space $N$ then there are convergent eigenvalues which converge to the spectrum of a first order differential operator $D$ on $N$ togethe…
▽ More
We study the behavior of the spectrum of the Dirac operator together with a symmetric $W^{1, \infty}$-potential on spin manifolds under a collapse of codimension one with bounded sectional curvature and diameter. If there is an induced spin structure on the limit space $N$ then there are convergent eigenvalues which converge to the spectrum of a first order differential operator $D$ on $N$ together with a symmetric $W^{1,\infty}$-potential. If $N$ is orientable and the dimension of the limit space is even then $D$ is the Dirac operator $D^N$ on $N$ and if the dimension of the limit space is odd, then $D = D^N \oplus -D^N$.
△ Less
Submitted 12 August, 2017; v1 submitted 3 July, 2017;
originally announced July 2017.
-
A characterization of codimension one collapse under bounded curvature and diameter
Authors:
Saskia Roos
Abstract:
Let $\mathcal{M}(n,D)$ be the space of closed $n$-dimensional Riemannian manifolds $(M,g)$ with $diam(M) \leq D$ and $| \sec^M | \leq 1$. In this paper we consider sequences $(M_i,g_i)$ in $\mathcal{M}(n,D)$ converging in the Gromov-Hausdorff topology to a compact metric space $Y$. We show on the one hand that the limit space of this sequence has at most codimension $1$ if there is a positive numb…
▽ More
Let $\mathcal{M}(n,D)$ be the space of closed $n$-dimensional Riemannian manifolds $(M,g)$ with $diam(M) \leq D$ and $| \sec^M | \leq 1$. In this paper we consider sequences $(M_i,g_i)$ in $\mathcal{M}(n,D)$ converging in the Gromov-Hausdorff topology to a compact metric space $Y$. We show on the one hand that the limit space of this sequence has at most codimension $1$ if there is a positive number $r$ such that the quotient $\frac{vol(B^{M_i}_r(x))}{inj^{M_i}(x)}$ can be uniformly bounded from below by a positive constant $C(n,r,Y)$ for all points $x \in M_i$. On the other hand, we show that if the limit space has at most codimension $1$ then for all positive $r$ there is a positive constant $C(n,r,Y)$ bounding the quotient $\frac{vol(B^{M_i}_r(x))}{inj^{M_i}(x)}$ uniformly from below for all $x \in M_i$. The proof uses results about the structure of collapse in $\mathcal{M}(n,D)$ by Cheeger, Fukaya and Gromov. In addition, we derive, for a submersion $M \rightarrow Y$ with uniformly bounded fundamental tensors, an upper bound on the injectivity radius of the fiber $F_p$, with $p \in Y$, which is proportional to the injectivity radius of $M$ at some $x \in F_p$, if the injectivity at $x$ is sufficiently small relative to the injectivity radius of $Y$. As a conclusion, we derive a uniform lower bound on the volume and a bound on the essential supremum of the sectional curvature for the closure of the space consisting of all manifolds in $\mathcal{M}(n,D)$ with $C \leq \min_{x \in M}\frac{vol(B^{M}_r(x))}{inj^{M}(x)}$ for fixed positive numbers $r$ and $C$.
△ Less
Submitted 18 July, 2017; v1 submitted 23 January, 2017;
originally announced January 2017.
-
Balanced Dynamic Content Addressing in Trees
Authors:
Stefanie Roos,
Martin Byrenheid,
Clemens Deusser,
Thorsten Strufe
Abstract:
Balancing the load in content addressing schemes for route-restricted networks represents a challenge with a wide range of applications. Solutions based on greedy embeddings maintain minimal state information and enable efficient routing, but any such solutions currently result in either imbalanced content addressing, overloading individual nodes, or are unable to efficiently account for network d…
▽ More
Balancing the load in content addressing schemes for route-restricted networks represents a challenge with a wide range of applications. Solutions based on greedy embeddings maintain minimal state information and enable efficient routing, but any such solutions currently result in either imbalanced content addressing, overloading individual nodes, or are unable to efficiently account for network dynamics. In this work, we propose a greedy embedding in combination with a content addressing scheme that provides balanced content addressing while at the same time enabling efficient stabilization in the presence of network dynamics. We point out the trade-off between stabilization complexity and maximal permitted imbalance when deriving upper bounds on both metrics for two variants of the proposed algorithms. Furthermore, we substantiate these bounds through a simulation study based on both real-world and synthetic data.
△ Less
Submitted 12 January, 2017;
originally announced January 2017.
-
Eigenvalue pinching on $\text{spin}^c$ manifolds
Authors:
Saskia Roos
Abstract:
We derive various pinching results for small Dirac eigenvalues using the classification of $\text{spin}^c$ and spin manifolds admitting nontrivial Killing spinors. For this, we introduce a notion of convergence for $\text{spin}^c$ manifolds which involves a general study on convergence of Riemannian manifolds with a principal $\mathbb{S}^1$-bundle. We also analyze the relation between the regulari…
▽ More
We derive various pinching results for small Dirac eigenvalues using the classification of $\text{spin}^c$ and spin manifolds admitting nontrivial Killing spinors. For this, we introduce a notion of convergence for $\text{spin}^c$ manifolds which involves a general study on convergence of Riemannian manifolds with a principal $\mathbb{S}^1$-bundle. We also analyze the relation between the regularity of the Riemannian metric and the regularity of the curvature of the associated principal $\mathbb{S}^1$-bundle on $\text{spin}^c$ manifolds with Killing spinors.
△ Less
Submitted 13 June, 2017; v1 submitted 6 April, 2016;
originally announced April 2016.
-
VOUTE-Virtual Overlays Using Tree Embeddings
Authors:
Stefanie Roos,
Martin Beck,
Thorsten Strufe
Abstract:
Friend-to-friend (F2F) overlays, which restrict direct communication to mutually trusted parties, are a promising substrate for privacy-preserving communication due to their inherent membership-concealment and Sybil-resistance. Yet, existing F2F overlays suffer from a low performance, are vulnerable to denial-of-service attacks, or fail to provide anonymity. In particular, greedy embeddings allow…
▽ More
Friend-to-friend (F2F) overlays, which restrict direct communication to mutually trusted parties, are a promising substrate for privacy-preserving communication due to their inherent membership-concealment and Sybil-resistance. Yet, existing F2F overlays suffer from a low performance, are vulnerable to denial-of-service attacks, or fail to provide anonymity. In particular, greedy embeddings allow highly efficient communication in arbitrary connectivity-restricted overlays but require communicating parties to reveal their identity. In this paper, we present a privacy-preserving routing scheme for greedy embeddings based on anonymous return addresses rather than identifying node coordinates. We prove that the presented algorithm are highly scalalbe, with regard to the complexity of both the routing and the stabilization protocols. Furthermore, we show that the return addresses provide plausible deniability for both sender and receiver. We further enhance the routing's resilience by using multiple embeddings and propose a method for efficient content addressing. Our simulation study on real-world data indicates that our approach is highly efficient and effectively mitigates failures as well as powerful denial-of-service attacks.
△ Less
Submitted 22 January, 2016;
originally announced January 2016.
-
Estimates for eigensections of Riemannian vector bundles
Authors:
Saskia Roos
Abstract:
We derive a bound on the $L^{\infty}$-norm of the covariant derivative of Laplace eigensections on general Riemannian vector bundles depending on the diameter, the dimension, the Ricci curvature of the underlying manifold, and the curvature of the Riemannian vector bundle. Our result implies that eigensections with small eigenvalues are almost parallel.
We derive a bound on the $L^{\infty}$-norm of the covariant derivative of Laplace eigensections on general Riemannian vector bundles depending on the diameter, the dimension, the Ricci curvature of the underlying manifold, and the curvature of the Riemannian vector bundle. Our result implies that eigensections with small eigenvalues are almost parallel.
△ Less
Submitted 13 June, 2017; v1 submitted 25 November, 2015;
originally announced November 2015.
-
A Lightweight Approach for Improving the Lookup Performance in Kademlia-type Systems
Authors:
Hani Salah,
Stefanie Roos,
Thorsten Strufe
Abstract:
Discovery of nodes and content in large-scale distributed systems is generally based on Kademlia, today. Understanding Kademlia-type systems to improve their performance is essential for maintaining a high service quality for an increased number of participants, particularly when those systems are adopted by latency-sensitive applications.
This paper contributes to the understanding of Kademlia…
▽ More
Discovery of nodes and content in large-scale distributed systems is generally based on Kademlia, today. Understanding Kademlia-type systems to improve their performance is essential for maintaining a high service quality for an increased number of participants, particularly when those systems are adopted by latency-sensitive applications.
This paper contributes to the understanding of Kademlia by studying the impact of \emph{diversifying} neighbours' identifiers within each routing table bucket on the lookup performance. We propose a new, yet backward-compatible, neighbour selection scheme that attempts to maximize the aforementioned diversity. The scheme does not cause additional overhead except negligible computations for comparing the diversity of identifiers. We present a theoretical model for the actual impact of the new scheme on the lookup's hop count and validate it against simulations of three exemplary Kademlia-type systems. We also measure the performance gain enabled by a partial deployment for the scheme in the real KAD system. The results confirm the superiority of the systems that incorporate our scheme.
△ Less
Submitted 27 July, 2014;
originally announced August 2014.
-
NextBestOnce: Achieving Polylog Routing despite Non-greedy Embeddings
Authors:
Stefanie Roos,
Thorsten Strufe
Abstract:
Social Overlays suffer from high message delivery delays due to insufficient routing strategies. Limiting connections to device pairs that are owned by individuals with a mutual trust relationship in real life, they form topologies restricted to a subgraph of the social network of their users. While centralized, highly successful social networking services entail a complete privacy loss of their u…
▽ More
Social Overlays suffer from high message delivery delays due to insufficient routing strategies. Limiting connections to device pairs that are owned by individuals with a mutual trust relationship in real life, they form topologies restricted to a subgraph of the social network of their users. While centralized, highly successful social networking services entail a complete privacy loss of their users, Social Overlays at higher performance represent an ideal private and censorship-resistant communication substrate for the same purpose.
Routing in such restricted topologies is facilitated by embedding the social graph into a metric space. Decentralized routing algorithms have up to date mainly been analyzed under the assumption of a perfect lattice structure. However, currently deployed embedding algorithms for privacy-preserving Social Overlays cannot achieve a sufficiently accurate embedding and hence conventional routing algorithms fail. Developing Social Overlays with acceptable performance hence requires better models and enhanced algorithms, which guarantee convergence in the presence of local optima with regard to the distance to the target.
We suggest a model for Social Overlays that includes inaccurate embeddings and arbitrary degree distributions. We further propose NextBestOnce, a routing algorithm that can achieve polylog routing length despite local optima. We provide analytical bounds on the performance of NextBestOnce assuming a scale-free degree distribution, and furthermore show that its performance can be improved by more than a constant factor when including Neighbor-of-Neighbor information in the routing decisions.
△ Less
Submitted 9 January, 2014;
originally announced January 2014.
-
Protecting Public OSN Posts from Unintended Access
Authors:
Frederik Armknecht,
Manuel Hauptmann,
Stefanie Roos,
Thorsten Strufe
Abstract:
The design of secure and usable access schemes to personal data represent a major challenge of online social networks (OSNs). State of the art requires prior interaction to grant access. Sharing with users who are not subscribed or previously have not been accepted as contacts in any case is only possible via public posts, which can easily be abused by automatic harvesting for user profiling, targ…
▽ More
The design of secure and usable access schemes to personal data represent a major challenge of online social networks (OSNs). State of the art requires prior interaction to grant access. Sharing with users who are not subscribed or previously have not been accepted as contacts in any case is only possible via public posts, which can easily be abused by automatic harvesting for user profiling, targeted spear-phishing, or spamming. Moreover, users are restricted to the access rules defined by the provider, which may be overly restrictive, cumbersome to define, or insufficiently fine-grained.
We suggest a complementary approach that can be easily deployed in addition to existing access control schemes, does not require any interaction, and includes even public, unsubscribed users. It exploits the fact that different social circles of a user share different experiences and hence encrypts arbitrary posts. Hence arbitrary posts are encrypted, such that only users with sufficient knowledge about the owner can decrypt.
Assembling only well-established cryptographic primitives, we prove that the security of our scheme is determined by the entropy of the required knowledge. We consequently analyze the efficiency of an informed dictionary attack and assess the entropy to be on par with common passwords. A fully functional implementation is used for performance evaluations, and available for download on the Web.
△ Less
Submitted 14 September, 2013;
originally announced September 2013.
-
Comprehending Kademlia Routing - A Theoretical Framework for the Hop Count Distribution
Authors:
Stefanie Roos,
Hani Salah,
Thorsten Strufe
Abstract:
The family of Kademlia-type systems represents the most efficient and most widely deployed class of internet-scale distributed systems. Its success has caused plenty of large scale measurements and simulation studies, and several improvements have been introduced. Its character of parallel and non-deterministic lookups, however, so far has prevented any concise formal analysis. This paper introduc…
▽ More
The family of Kademlia-type systems represents the most efficient and most widely deployed class of internet-scale distributed systems. Its success has caused plenty of large scale measurements and simulation studies, and several improvements have been introduced. Its character of parallel and non-deterministic lookups, however, so far has prevented any concise formal analysis. This paper introduces the first comprehensive formal model of the routing of the entire family of systems that is validated against previous measurements. It sheds light on the overall hop distribution and lookup delays of the different variations of the original protocol. It additionally shows that several of the recent improvements to the protocol in fact have been counter-productive and identifies preferable designs with regard to routing overhead and resilience.
△ Less
Submitted 26 July, 2013;
originally announced July 2013.
-
The intermediate-redshift galaxy cluster CL 0048-2942. Stellar populations
Authors:
M. Serote Roos,
C. Lobo,
F. Durret,
A. Iovino,
I. Marquez
Abstract:
We present a detailed study of the cluster CL 0048-2942, located at z~0.64, based on a photometric and spectroscopic catalogue of 54 galaxies in a 5 x 5 square arcmin region centred in that cluster. Of these, 23 galaxies were found to belong to the cluster. Based on this sample, the line-of-sight velocity dispersion of the cluster is approximately 680 +- 140 km/s. We have performed stellar popul…
▽ More
We present a detailed study of the cluster CL 0048-2942, located at z~0.64, based on a photometric and spectroscopic catalogue of 54 galaxies in a 5 x 5 square arcmin region centred in that cluster. Of these, 23 galaxies were found to belong to the cluster. Based on this sample, the line-of-sight velocity dispersion of the cluster is approximately 680 +- 140 km/s. We have performed stellar population synthesis in the cluster members as well as in the field galaxies of the sample and found that there are population gradients in the cluster with central galaxies hosting mainly intermediate/old populations whereas galaxies in the cluster outskirts show clearly an increase of younger populations, meaning that star formation is predominantly taking place in the outer regions of the cluster. In a general way, field galaxies seem to host less evolved stellar populations than cluster members. In fact, in terms of ages, young supergiant stars dominate the spectra of field galaxies whereas cluster galaxies display a dominant number of old and intermediate age stars. Following the work of other authors (e.g. Dressler et al. 1999) we have estimated the percentage of K+A galaxies in our sample and found around 13% in the cluster and 10% in the field. These values were estimated through means of a new method, based on stellar population synthesis results, that takes into account all possible absorption features in the spectrum and thus makes optimal use of the data.
△ Less
Submitted 6 September, 2004;
originally announced September 2004.
-
The Nuclear Region of Low Luminosity Flat Radio Spectrum Sources. II. Emission-Line Spectra
Authors:
A. C. Goncalves,
M. Serote Roos
Abstract:
We report on the spectroscopic study of 19 low luminosity Flat Radio Spectrum (LL FRS) sources selected from Marcha's et al. (1996) 200 mJy sample. In the optical, these objects are mainly dominated by the host galaxy starlight. After correcting the data for this effect, we obtain a new set of spectra clearly displaying weak emission lines; such features carry valuable information concerning the…
▽ More
We report on the spectroscopic study of 19 low luminosity Flat Radio Spectrum (LL FRS) sources selected from Marcha's et al. (1996) 200 mJy sample. In the optical, these objects are mainly dominated by the host galaxy starlight. After correcting the data for this effect, we obtain a new set of spectra clearly displaying weak emission lines; such features carry valuable information concerning the excitation mechanisms at work in the nuclear regions of LL FRS sources. We have used a special routine to model the spectra and assess the intensities and velocities of the emission lines; we have analyzed the results in terms of diagnostic diagrams. Our analysis shows that 79% of the studied objects harbour a Low Ionization Nuclear Emission-line Region (or LINER) whose contribution was swamped by the host galaxy starlight. The remaining objects display a higher ionization spectrum, more typical of Seyferts; due to the poor quality of the spectra, it was not possible to identify any possible large Balmer components. The fact that we observe a LINER-type spectrum in LL FRS sources supports the idea that some of these objects could be undergoing an ADAF phase; in addition, such a low ionization emission-line spectrum is in agreement with the black hole mass values and sub-Eddington accretion rates published for some FRS sources.
△ Less
Submitted 23 September, 2003;
originally announced September 2003.
-
The Nuclear Region of Low Luminosity Flat Radio Spectrum Sources. I. Stellar Content
Authors:
M. Serote Roos,
A. C. Goncalves
Abstract:
We have examined the spectroscopic properties of a sample of 19 optically bright, low luminosity Flat Radio Spectrum (LL FRS) sources. Our study focuses on the properties of their host galaxies, namely the nuclear stellar populations and dust content. In the optical, the objects in the sample are mainly dominated by the host galaxy starlight, which strongly dilutes the non-thermal continuum as w…
▽ More
We have examined the spectroscopic properties of a sample of 19 optically bright, low luminosity Flat Radio Spectrum (LL FRS) sources. Our study focuses on the properties of their host galaxies, namely the nuclear stellar populations and dust content. In the optical, the objects in the sample are mainly dominated by the host galaxy starlight, which strongly dilutes the non-thermal continuum as well as possible emission-line features related to the active nucleus. We have computed the nuclear stellar populations contributing to the spectra of our objects. The stellar population synthesis has been performed by using a very reliable mathematical method, which yields a Global Principal Geometrical solution. Our results show that, for most of the objects in the sample, the populations are composed of old stars of solar metallicity, or lower; the populations are mainly composed of late-type stars, i.e. G, K and M spectral types, the young component coming thus from supergiant stars; the dust content is weak. Both the stellar populations and the dust content are in agreement with what is usually observed in "normal" elliptical galaxies. Similar stellar content has equally been found in the nuclear regions of galaxies hosting a Low Ionization Nuclear Emission Line Region, or LINER. The present work is important in illustrating the different applications of stellar population synthesis in the study of low luminosity radio sources. In fact, the synthesis allows us not only to obtain valuable information about the stellar populations and dust content of the host galaxies, therefore providing material for further studies on the connection between host galaxy and active nucleus, but also to reveal the so-far unstudied optical emission-line features present in the spectrum of our objects.
△ Less
Submitted 19 September, 2003;
originally announced September 2003.
-
Stellar population in Active Galactic Nuclei. I. The Observations
Authors:
M. Serote Roos,
C. Boisson,
M. Joly,
M. J. Ward
Abstract:
Recent observations supported by theoretical models have lead to the view that giant and supergiant stars are over abundant, and/or a high metallicity component may be present, in the stellar populations at the centres of active galaxies. Here we attempt to quantify these effects by observing the strengths of the stellar absorption lines of Mg~b, NaI, CaII triplet as well as molecular bands such…
▽ More
Recent observations supported by theoretical models have lead to the view that giant and supergiant stars are over abundant, and/or a high metallicity component may be present, in the stellar populations at the centres of active galaxies. Here we attempt to quantify these effects by observing the strengths of the stellar absorption lines of Mg~b, NaI, CaII triplet as well as molecular bands such as CN and TiO. Using long-slit spectroscopic data we are able to separate the stellar populations in and around the nucleus, for a sample including, normal, LINER, starburst and Seyfert galaxies.
In this paper we present the data, namely spectra of the nucleus and of a number of circum-nuclear regions. Comparisons reveal gradients in both the reddening and the stellar population within the central regions of most galaxies. Detailed stellar population synthesis is presented in a companion paper.
△ Less
Submitted 3 April, 1998;
originally announced April 1998.