arXiv subjects

Feng Luo

Publications and source records attributed to Feng Luo.

At least 91 records · Page 5Linked to original sources

arXiv2021

A systematic mapping study on security countermeasures of in-vehicle communication systems

The innovations of vehicle connectivity have been increasing dramatically to enhance the safety and user experience of driving, while the rising numbers of interfaces to the external world also bring security threats to vehicles. Many security countermeasures have been proposed and discussed to protect the systems and services against attacks. To provide an overview of the current states in this research field, we conducted a systematic mapping study on the topic area "security countermeasures of in-vehicle communication systems". 279 papers are identified based on the defined study identification strategy and criteria. We discussed four research questions related to the security countermeasures, validation methods, publication patterns, and research trends and gaps based on the extracted and classified data. Finally, we evaluated the validity threats, the study identification results, and the whole mapping process. We found that the studies in this topic area are increasing rapidly in recent years. However, there are still gaps in various subtopics like automotive Ethernet security, anomaly reaction, and so on. This study reviews the target field not only related to research findings but also research activities, which can help identify research gaps at a high level and inspire new ideas for future work.

cs.CR↗

arXiv2021

Double cycle-consistent generative adversarial network for unsupervised conditional generation

Conditional generative models have achieved considerable success in the past few years, but usually require a lot of labeled data. Recently, ClusterGAN combines GAN with an encoder to achieve remarkable clustering performance via unsupervised conditional generation. However, it ignores the real conditional distribution of data, which leads to generating less diverse samples for each class and makes the encoder only achieve sub-optimal clustering performance. Here, we propose a new unsupervised conditional generation framework, Double Cycle-Consistent Conditional GAN (DC3-GAN), which can generate diverse class-conditioned samples. We enforce the encoder and the generator of GAN to form an encoder-generator pair in addition to the generator-encoder pair, which enables us to avoid the low-diversity generation and the triviality of latent features. We train the encoder-generator pair using real data, which can indirectly estimate the real conditional distribution. Meanwhile, this framework enforces the outputs of the encoder to match the inputs of GAN and the prior noise distribution, which disentangles latent space into two parts: one-hot discrete and continuous latent variables. The former can be directly expressed as clusters and the latter represents remaining unspecified factors. This work demonstrates that enhancing the diversity of unsupervised conditional generated samples can improve the clustering performance. Experiments on different benchmark datasets show that the proposed method outperforms existing generative model-based clustering methods, and also achieves the optimal disentanglement performance.

cs.LG↗

arXiv2020

Final state Sommerfeld effect on dark matter relic abundance

If the annihilation products of dark matter (DM) are non-relativistic and if there is some long-range force between them, there can be Sommerfeld effect for the final state particles. We study this effect on DM relic abundance in the thermal freeze-out scenario. As a proof of concept, we consider the case of a DM pair annihilation into a final state pair, assuming that the mutual interactions between the two final state particles give rise to a Coulomb-like potential, and that the masses of the initial and final state particles are similar, so that both the initial and final state particles are non-relativistic. The size of the final state Sommerfeld (FSS) effect depends on the strength of the potential, as well as on the mass ratio of the final and initial state particles. We find that the impact of the FSS effect on DM relic abundance can be significant, and an electroweak sized long-range interaction is large enough to make a correction well beyond the observational accuracy. Another feature of the FSS effect is that it could be suppressed when its time scale is longer than the lifetime of the final state particles. As a corollary, we also study in the DM coannihilation scenario where the initial state Sommerfeld effect between two coannihilators could be reduced due to their instability, which may need to be taken into account for an accurate calculation of the DM relic abundance.

hep-ph↗

arXiv2020

Discrete Conformal Geometry of Polyhedral Surfaces and Its Convergence

The paper proves a result on the convergence of discrete conformal maps to the Riemann mappings for Jordan domains. It is a counterpart of Rodin-Sullivan's theorem on convergence of circle packing mappings to the Riemann mapping in the new setting of discrete conformality. The proof follows the same strategy that Rodin-Sullivan used by establishing a rigidity result for regular hexagonal triangulations of the plane and estimating the quasiconformal constants associated to the discrete conformal maps.

math.GT↗

arXiv2020

Second-order Neural Network Training Using Complex-step Directional Derivative

While the superior performance of second-order optimization methods such as Newton's method is well known, they are hardly used in practice for deep learning because neither assembling the Hessian matrix nor calculating its inverse is feasible for large-scale problems. Existing second-order methods resort to various diagonal or low-rank approximations of the Hessian, which often fail to capture necessary curvature information to generate a substantial improvement. On the other hand, when training becomes batch-based (i.e., stochastic), noisy second-order information easily contaminates the training procedure unless expensive safeguard is employed. In this paper, we adopt a numerical algorithm for second-order neural network training. We tackle the practical obstacle of Hessian calculation by using the complex-step finite difference (CSFD) -- a numerical procedure adding an imaginary perturbation to the function for derivative computation. CSFD is highly robust, efficient, and accurate (as accurate as the analytic result). This method allows us to literally apply any known second-order optimization methods for deep learning training. Based on it, we design an effective Newton Krylov procedure. The key mechanism is to terminate the stochastic Krylov iteration as soon as a disturbing direction is found so that unnecessary computation can be avoided. During the optimization, we monitor the approximation error in the Taylor expansion to adjust the step size. This strategy combines advantages of line search and trust region methods making our method preserves good local and global convergency at the same time. We have tested our methods in various deep learning tasks. The experiments show that our method outperforms exiting methods, and it often converges one-order faster. We believe our method will inspire a wide-range of new algorithms for deep learning and numerical optimization.

cs.LG↗

arXiv2020

SilkViser:A Visual Explorer of Blockchain-based Cryptocurrency Transaction Data

Many blockchain-based cryptocurrencies provide users with online blockchain explorers for viewing online transaction data. However, traditional blockchain explorers mostly present transaction information in textual and tabular forms. Such forms make understanding cryptocurrency transaction mechanisms difficult for novice users (NUsers). They are also insufficiently informative for experienced users (EUsers) to recognize advanced transaction information. This study introduces a new online cryptocurrency transaction data viewing tool called SilkViser. Guided by detailed scenario and requirement analyses, we create a series of appreciating visualization designs, such as paper ledger-inspired block and blockchain visualizations and ancient copper coin-inspired transaction visualizations, to help users understand cryptocurrency transaction mechanisms and recognize advanced transaction information. We also provide a set of lightweight interactions to facilitate easy and free data exploration. Moreover, a controlled user study is conducted to quantitatively evaluate the usability and effectiveness of SilkViser. Results indicate that SilkViser can satisfy the requirements of NUsers and EUsers. Our visualization designs can compensate for the inexperience of NUsers in data viewing and attract potential users to participate in cryptocurrency transactions.

cs.CY↗

arXiv2020

An STPA-based Approach for Systematic Security Analysis of In-vehicle Diagnostic and Software Update Systems

The in-vehicle diagnostic and software update system, which supports remote diagnostic and Over-The-Air (OTA) software updates, is a critical attack goal in automobiles. Adversaries can inject malicious software into vehicles or steal sensitive information through communication channels. Therefore, security analysis, which identifies potential security issues, needs to be conducted in system design. However, existing security analyses of in-vehicle systems are threat-oriented, which start with threat identification and assess risks by brainstorming. In this paper, a system-oriented approach is proposed on the basis of the System-Theoretic Process Analysis (STPA). The proposed approach extends the original STPA from the perspective of data flows and is applicable for information-flow-based systems. Besides, we propose a general model for in-vehicle diagnostic and software update systems and use it to establish a security analysis guideline. In comparison with threat-oriented approaches, the proposed approach shifts from focusing on threats to system vulnerabilities and seems to be efficient to prevent the system from known or even unknown threats. Furthermore, as an extension of the STPA, which has been proven to be applicable to high level designs, the proposed approach can be well integrated into high-level analyses and perform co-design in different disciplines within a unified STPA framework.

cs.CR↗

arXiv2020

Data-Flow-Based Extension of the System-Theoretic Process Analysis for Security (STPA-Sec)

Security analysis is an essential activity in security engineering to identify potential system vulnerabilities and achieve security requirements in the early design phases. Due to the increasing complexity of modern systems, traditional approaches, which only consider component failures and simple cause-and-effect linkages, lack the power to identify insecure incidents caused by complex interactions among physical systems, human and social entities. By contrast, a top-down System-Theoretic Process Analysis for Security (STPA-Sec) approach views losses as resulting from interactions, focuses on controlling system vulnerabilities instead of external threats and is applicable for complex socio-technical systems. In this paper, we proposed an extension of STPA-Sec based on data flow structures to overcome STPA-Sec's limitations and achieve security constraints of information-critical systems systematically. We analyzed a Bluetooth digital key system of a vehicle by using both the proposed and the original approach to investigate the relationship and differences between both approaches as well as their applicability and highlights. To conclude, the proposed approach can identify more information-related problems with technical details and be used with other STPA-based approaches to co-design systems in multi-disciplines under the unified STPA process framework.

cs.CR↗

arXiv2020

How Heavy can Neutralino Dark Matter be?

What is the upper limit of the mass of the neutralino dark matter whose thermal relic is consistent with the observation? If the neutralino dark matter and colored sparticles are extremely degenerated in mass, with a mass difference less than the QCD scale, the dark matter annihilation is significantly increased and enjoys the "second freeze-out" after the QCD phase transition. In this case, the neutralino dark matter with a mass much greater than 100 TeV can realize the correct dark matter abundance. We study the dark matter abundance and its detection in the case of such highly degenerated mass spectrum of the neutralino dark matter and colored supersymmetric particles.

hep-ph↗

arXiv2019

An effective Lie--Kolchin theorem for quasi-unipotent matrices

We establish an effective version of the classical Lie--Kolchin Theorem. Namely, let $A,B\in\mathrm{GL}_m(\mathbb{C})$ be quasi--unipotent matrices such that the Jordan Canonical Form of $B$ consists of a single block, and suppose that for all $k\geq0$ the matrix $AB^k$ is also quasi--unipotent. Then $A$ and $B$ have a common eigenvector. In particular, $\langle A,B\rangle<\mathrm{GL}_m(\mathbb{C})$ is a solvable subgroup. We give applications of this result to the representation theory of mapping class groups of orientable surfaces.

math.GR↗

arXiv2019

Community Detection on Networks with Ricci Flow

Many complex networks in the real world have community structures -- groups of well-connected nodes with important functional roles. It has been well recognized that the identification of communities bears numerous practical applications. While existing approaches mainly apply statistical or graph theoretical/combinatorial methods for community detection, in this paper, we present a novel geometric approach which enables us to borrow powerful classical geometric methods and properties. By considering networks as geometric objects and communities in a network as a geometric decomposition, we apply curvature and discrete Ricci flow, which have been used to decompose smooth manifolds with astonishing successes in mathematics, to break down communities in networks. We tested our method on networks with ground-truth community structures, and experimentally confirmed the effectiveness of this geometric approach.

cs.SI↗

arXiv2019

Quadrilateral Mesh Generation II : Meromorphic Quartic Differentials and Abel-Jacobi Condition

This work discovers the equivalence relation between quadrilateral meshes and meromorphic quartic. Each quad-mesh induces a conformal structure of the surface, and a meromorphic differential, where the configuration of singular vertices correspond to the configurations the poles and zeros (divisor) of the meroromorphic differential. Due to Riemann surface theory, the configuration of singularities of a quad-mesh satisfies the Abel-Jacobi condition. Inversely, if a satisfies the Abel-Jacobi condition, then there exists a meromorphic quartic differential whose equals to the given one. Furthermore, if the meromorphic quadric differential is with finite, then it also induces a a quad-mesh, the poles and zeros of the meromorphic differential to the singular vertices of the quad-mesh. Besides the theoretic proofs, the computational algorithm for verification of Abel-Jacobi condition is explained in details. Furthermore, constructive algorithm of meromorphic quartic differential on zero surfaces is proposed, which is based on the global algebraic representation of meromorphic. Our experimental results demonstrate the efficiency and efficacy of the algorithm. This opens up a direction for quad-mesh generation using algebraic geometric approach.

cs.CG↗

arXiv2018

Stop Coannihilation in the CMSSM and SubGUT Models

Stop coannihilation may bring the relic density of heavy supersymmetric dark matter particles into the range allowed by cosmology. The efficiency of this process is enhanced by stop-antistop annihilations into the longitudinal (Goldstone) modes of the $W$ and $Z$ bosons, as well as by Sommerfeld enhancement of stop annihilations and the effects of bound states. Since the couplings of the stops to the Goldstone modes are proportional to the trilinear soft supersymmetry-breaking $A$-terms, these annihilations are enhanced when the $A$-terms are large. However, the Higgs mass may be reduced below the measured value if the $A$-terms are too large. Unfortunately, the interpretation of this constraint on the stop coannihilation strip is clouded by differences between the available Higgs mass calculators. For our study, we use as our default calculator FeynHiggs 2.13.0, the most recent publicly available version of this code. Exploring the CMSSM parameter space, we find that along the stop coannihilation strip the masses of the stops are severely split by the large $A$-terms. This suppresses the Higgs mass drastically for $μ$ and $A_0 > 0$, whilst the extent of the stop coannihilation strip is limited for $A_0 < 0$ and either sign of $μ$. However, in sub-GUT models, reduced renormalization-group running mitigates the effect of the large $A$-terms, allowing larger LSP masses to be consistent with the Higgs mass calculation. We give examples where the dark matter particle mass may reach $\gtrsim 8$ TeV.

hep-ph↗

arXiv2018

CrescendoNet: A Simple Deep Convolutional Neural Network with Ensemble Behavior

We introduce a new deep convolutional neural network, CrescendoNet, by stacking simple building blocks without residual connections. Each Crescendo block contains independent convolution paths with increased depths. The numbers of convolution layers and parameters are only increased linearly in Crescendo blocks. In experiments, CrescendoNet with only 15 layers outperforms almost all networks without residual connections on benchmark datasets, CIFAR10, CIFAR100, and SVHN. Given sufficient amount of data as in SVHN dataset, CrescendoNet with 15 layers and 4.1M parameters can match the performance of DenseNet-BC with 250 layers and 15.3M parameters. CrescendoNet provides a new way to construct high performance deep convolutional neural networks without residual connections. Moreover, through investigating the behavior and performance of subnetworks in CrescendoNet, we note that the high performance of CrescendoNet may come from its implicit ensemble behavior, which differs from the FractalNet that is also a deep convolutional neural network without residual connections. Furthermore, the independence between paths in CrescendoNet allows us to introduce a new path-wise training procedure, which can reduce the memory needed for training.

cs.LG↗

arXiv2017

Effects of QCD bound states on dark matter relic abundance

We study scenarios where there exists an exotic massive particle charged under QCD in the early Universe. We calculate the formation and dissociation rates of bound states formed by pairs of these particles, and apply the results in dark matter (DM) coannihilation scenarios, including also the Sommerfeld effect. We find that on top of the Sommerfeld enhancement, bound-state effects can further significantly increase the largest possible DM masses which can give the observed DM relic abundance, by $\sim 30 - 100\%$ with respect to values obtained by considering the Sommerfeld effect only, for the color triplet or octet exotic particles we consider. In particular, it indicates that the Bino DM mass in the right-handed stop-Bino coannihilation scenario in the Minimal Supersymmetric extension of the Standard Model (MSSM) can reach $\sim 2.5$ TeV, even though the potential between the stop and antistop prior to forming a bound state is repulsive. We also apply the bound-state effects in the calculations of relic abundance of long-lived or metastable massive colored particles, and discuss the implications on the BBN constraints and the abundance of a super-weakly interacting DM. The corrections for the bound-state effect when the exotic massive colored particles also carry electric charges, and the collider bounds are also discussed.

hep-ph↗

arXiv2016

DynIMS: A Dynamic Memory Controller for In-memory Storage on HPC Systems

In order to boost the performance of data-intensive computing on HPC systems, in-memory computing frameworks, such as Apache Spark and Flink, use local DRAM for data storage. Optimizing the memory allocation to data storage is critical to delivering performance to traditional HPC compute jobs and throughput to data-intensive applications sharing the HPC resources. Current practices that statically configure in-memory storage may leave inadequate space for compute jobs or lose the opportunity to utilize more available space for data-intensive applications. In this paper, we explore techniques to dynamically adjust in-memory storage and make the right amount of space for compute jobs. We have developed a dynamic memory controller, DynIMS, which infers memory demands of compute tasks online and employs a feedback-based control model to adapt the capacity of in-memory storage. We test DynIMS using mixed HPCC and Spark workloads on a HPC cluster. Experimental results show that DynIMS can achieve up to 5X performance improvement compared to systems with static memory allocations.

cs.PF↗

arXiv2016

Pseudo-developing maps for ideal triangulations II: Positively oriented ideal triangulations of cone-manifolds

We generalise work of Young-Eun Choi to the setting of ideal triangulations with vertex links of arbitrary genus, showing that the set of all (possibly incomplete) hyperbolic cone-manifold structures realised by positively oriented hyperbolic ideal tetrahedra on a given topological ideal triangulation and with prescribed cone angles at all edges is (if non-empty) a smooth complex manifold of dimension the sum of the genera of the vertex links. Moreover, we show that the complex lengths of a collection of peripheral elements give a local holomorphic parameterisation of this manifold.

math.GT↗

arXiv2015

A new combinatorial class of 3-manifold triangulations

We define a new combinatorial class of triangulations of closed 3-manifolds, satisfying a weak version of 0-efficiency combined with a weak version of minimality, and study them using twisted squares. As an application, we obtain strong restrictions on the topology of a 3-manifold from the existence of non-smooth maxima of the volume function on the space of circle-valued angle structures.

math.GT↗