20199134 2019-2020-2 《网络攻防实践》实践作业

20199134 赵兴波 《网络攻防实践》实践作业

1会议介绍

1.1 NDSS会议

网络和分布式系统安全研讨会(NDSS)促进了网络和分布式系统安全的研究人员和从业人员之间的信息交换。目标受众包括对网络和分布式系统安全性的实际方面感兴趣的人员,并着重于实际的系统设计和实现。一个主要目标是鼓励和使Internet社区能够应用,部署和提高可用安全技术的状态。

1.2 S&P会议

IEEE Symposium on Security and Privacy(简称 S&P)创办于1980年,是信息安全领域四大顶级学术会议之一,也是信息安全领域历史最悠久的国际顶级学术会议。IEEE安全与隐私研讨会一直是介绍计算机安全和电子隐私发展的主要论坛,并将该领域的研究人员和从业者聚集在一起。该会议一直在美国加州Oakland举办,因此也被同行称为Oakland会议。

1.3USENIX会议

USENIX协会是非营利组织,致力于支持先进的计算机系统社区并扩大创新研究的范围。自1975年以来,USENIX聚集了致力于计算世界前沿的工程师,系统管理员,SRE,研究人员和技术人员社区。USENIX会议已成为展示和讨论有关计算系统各个方面的最先进信息的重要会议场地。

1.4 CSS会议

CCS(Computer and Communications Security),是国际计算机学会SIGSAC小组的年度会议。该会议在每年的10-11月份进行,地点大部分集中在美国,近年来也有在加拿大和英国举办。CCS会议主要集结了来自世界各地的信息安全研究人员、实践者、开发人员和用户,探索和交流计算机安全领域前沿的想法和最新成果。

2.论文方向与内容

2.1 NDSS 2020会议

该会议2020年2月23日至26日在加利福尼亚州圣地亚哥的双体船度假酒店及水疗中心举行。共录取了88份论文,录用论文如下:

1 Web

FUSE: Finding File Upload Bugs via Penetration Testing
该论文设计并实现了FUSE,这是第一个用于发现服务器端PHP web应用程序中的无限制文件上传(UFU)和不受限制的可执行文件上传(UEFU)漏洞的渗透测试工具。FUSE的目标是生成上传请求;每个请求都会成为一个可触发UFU或UEFU漏洞的攻击载荷。作者们通过精心设计对标准上传请求进行了改变,解决了一些技术挑战,使得改变后的操作可以绕过内容过滤检查,并且不会篡改上传文件的执行。FUSE发现了30个以前未报告的UEFU漏洞,包括33个真实web应用程序中的15个cve,从而证明了它通过文件上传查找代码执行错误的有效性。
Melting Pot of Origins: Compromising the Intermediary Web Services that Rehost Websites

Deceptive Previews: A Study of the Link Preview Trustworthiness in Social Platforms
Cross-Origin State Inference (COSI) Attacks: Leaking Web Site States through XS-Leaks
Carnus: Exploring the Privacy Threats of Browser Extension Fingerprinting

2 Fuzzing

HYPER-CUBE: High-Dimensional Hypervisor Fuzzing
该论文设计了一种多维的、与平台无关的模糊化方法,可以有效地测试不同接口及其交互。作者描述了一种针对hypervisor执行模糊测试的高效方法。该方法独立于应该测试的hypervisor。作者在一个名为“超立方体”的自定义操作系统中实现了该技术。其实验表明,HYPER-CUBE能够在许多实际的hypervisor中发现安全漏洞。
HFL: Hybrid Fuzzing on the Linux Kernel
HotFuzz: Discovering Algorithmic Denial-of-Service Vulnerabilities Through Guided Micro-Fuzzing
Not All Coverage Measurements Are Equal: Fuzzing by Coverage Accounting for Input Prioritization

3 Censorship

Detecting Probe-resistant Proxies
该论文找出了一些方法,该方法使得审查员可以将抗探针的代理与互联网上其他无害的主机区分开来,哪怕这些代理是设计好的。作者发现了流行的规避软件中使用的五种抗探针协议的独特的TCP行为,这些协议允许审查者以最小的误报率有效地确认可疑的代理。作者评估和分析了其对数十万台服务器的攻击,并使用ZMap进行了主动扫描。结果发现,他们的攻击能够有效地识别代理服务器,只需少量的探测连接,而误报可以忽略不计。使用其数据集,作者还建议对这些攻击进行防御。
Decentralized Control: A Case Study of Russia

Measuring the Deployment of Network Censorship Filters at Global Scale

SymTCP: Eluding Stateful Deep Packet Inspection with Automated Discrepancy Discovery

MassBrowser: Unblocking the Censored Web for the Masses, by the Masses

4 “Smart” Home

Et Tu Alexa? When Commodity WiFi Devices Turn into Adversarial Motion Sensor
该论文利用一个带有单个天线的嗅探器识别出一种低成本、不可检测的人类感知攻击,并设计了一种新的运动检测多径信号方差模型。作者在商品智能手机上建立了攻击者系统的原型,并在真实环境中验证了攻击。作者还提出了一种使用基于AP的信号模糊处理的实用有效的防御方案,并进行了评估。
Metamorph: Injecting Inaudible Commands into Over-the-air Voice Controlled Systems
SurfingAttack: Interactive Hidden Attack on Voice Assistants Using Ultrasonic Guided Waves
Packet-Level Signatures for Smart Home Devices

5 Mobile & Smartphone Security

Learning-based Practical Smartphone Eavesdropping with Built-in Accelerometer

Automated Cross-Platform Reverse Engineering of CAN Bus Commands From Mobile Apps
Are You Going to Answer That? Measuring User Responses to Anti-Robocall Application Indicators
TKPERM: Cross-platform Permission Knowledge Transfer to Detect Overprivileged Third-party Applications
FlowPrint: Semi-Supervised Mobile-App Fingerprinting on Encrypted Network Traffic

6 Blockchains and MPC

Bobtail: Improved Blockchain Security with Low-Variance Mining
该论文证明了高方差是PoW区块链基本攻击的根源。作者提出了一种基于PoW的块发现的替代过程,该过程可以产生显著更低方差的块间时间。该算法称为Bobtail,它将当前使用单个PoW样本的算法推广到包含k个样本的算法。作者证明了块间时间的方差随着k的增加而减小。Bobtail极大地挫败了双重支出和自私的采矿攻击
Snappy: Fast On-chain Payments with Practical Collaterals

The Attack of the Clones Against Proof-of-Authority
Broken Metre: Attacking Resource Metering in EVM

Finding Safety in Numbers with Secure Allegation Escrows

7 Future Networks

When Match Fields Do Not Need to Match: Buffered Packets Hijacking in SDN
该论文利用一个现有的防御系统中的恶意数据包对系统中的漏洞进行识别。这种开发了四种有效的攻击,利用已识别的漏洞攻击所有三个SDN层。这种在真实的SDN试验台上进行了实验,以证明所识别攻击的劫持概率和有效性。这种设计并实现了一个名为ConCheck的轻量级对策,以防止劫持缓冲包。
Automated Discovery of Cross-Plane Event-Based Vulnerabilities in Software-Defined Networking

SVLAN: Secure & Scalable Network Virtualization

8 Software Defenses

µRAI: Securing Embedded Systems with Return Address Integrity
该论文建议将返回地址完整性(RAI)属性作为一个基本不变项,以保护mcu不受针对后缘的控制流劫持攻击。作者通过将SFI与SR编码机制耦合,甚至在没有特殊硬件扩展的情况下,对中断的特权和异步执行强制RAI属性。作者设计并实现了一个在mcu上强制RAI属性的原型。作者评估了CoreMark、代表性mcu应用程序和概念验证攻击的实现。结果表明,µRAI以0.1%的运行时开销强制RAI属性。
NoJITsu: Locking Down JavaScript Engines
SODA: A Generic Online Detection Framework for Smart Contracts

9 Network Crime and Privacy

A Practical Approach for Taking Down Avalanche Botnets Under Real-World Constraints
该论文评估了自动化方法可以在多大程度上帮助执法调查人员正确检测雪崩摧毁涉及的注册域与良性域的冲突,而不依赖批量恶意注册、持续的恶意软件活动或主动收集的流量。作者开发了一种技术,通过对信息最丰富和最困难的领域进行有针对性的手动标记来补充机器学习模型,以在多个执行迭代中保持性能,同时仍然大大减少所需的手动调查工作量。作者评估了这种方法在2017年和2018年的执行和转移情况:并获得了97.6%的准确率。
Designing a Better Browser for Tor with BLAST
Encrypted DNS --> Privacy? A Traffic Analysis Perspective

On Using Application-Layer Middlebox Protocols for Peeking Behind NAT Gateways

10 Side Channels

ABSynthe: Automatic Blackbox Side-channel Synthesis on Commodity Microarchitectures
该论文提出了一种以目标程序和微体系结构为输入,自动合成网络侧信道的系统ABsynthe。关键的见解是,通过限制作者自己使用(通常在核心上)基于竞争的侧通道,可以将目标CPU微体系结构视为一个黑匣子,从而实现自动化。为了使ABsynthe成为可能,作者为各种x86_64微体系结构自动生成了泄漏映射。为了利用优化后的指令序列来恢复秘密信息,ABsynthe依赖于递归神经网络来制造实用的侧信道攻击来恢复秘密比特流。评估表明,与目前针对单个组件的基于竞争的攻击相比,Absynth可以同时利用多个组件上的竞争来合成更好的攻击。此外,Absynth实现的自动化允许在本机和虚拟化环境中对四个不同的加密软件目标合成各种微体系结构的跨线程攻击。
PhantomCache: Obfuscating Cache Conflicts with Localized Randomization

Data-Driven Debugging for Functional Side Channels
Mind the Portability: A Warriors Guide through Realistic Profiled Side-channel Analysis

11 Network Defenses

Hold the Door! Fingerprinting Your Car Key to Prevent Keyless Entry Car Theft
该论文提出了一种RF指纹识别方法(HODOR,HOld the DOoR)来检测对无钥匙进入系统的攻击,这是在汽车领域首次尝试利用RF指纹技术。HODOR被设计为一种子认证方法,它支持现有的无钥匙进入系统的认证系统,并且不需要对主系统进行任何修改。通过一系列的实验,结果表明HODOR能够有效、可靠地检测到对无钥匙进入系统的攻击。HODOR的平均误报率(FPR)为0.27%,对模拟攻击的检测误报率(FNR)为0%。HODOR在保持方便性的同时,为无钥匙进入系统提供安全服务。
Poseidon: Mitigating Volumetric DDoS Attacks with Programmable Switches

EASI: Edge-Based Sender Identification on Resource-Constrained Platforms for Automotive Networks
BLAG: Improving the Accuracy of Blacklists
DefRec: Establishing Physical Function Virtualization to Disrupt Reconnaissance of Power Grids' Cyber-Physical Infrastructures

12 Oblivious Computation

Revisiting Leakage Abuse Attacks
该论文以多种方式重新审视了泄漏滥用攻击。作者首先强调了众所周知的IKK和Count攻击的一些实际限制和假设。然后,作者设计了四种新的泄漏滥用攻击,它们依赖于更弱的假设。其中三种攻击都是容量攻击,因为它们只利用与文档大小相关的泄漏。作者还介绍了两种容量注入攻击,它们使用对抗性文件添加来恢复查询,甚至可以从基于ORAM的解决方案中恢复查询。作者对所有攻击进行了经验评估,并考虑了许多实验设置,实验结果表明在实际假设下,导致合理恢复率的唯一设置是具有泄漏配置文件的高选择性查询,对于这一特定设置,作者提出了一些建议和对策。
Metal: A Metadata-Hiding File-Sharing System

MACAO: A Maliciously-Secure and Client-Efficient Active ORAM Framework

Heterogeneous Private Information Retrieval
Dynamic Searchable Encryption with Small Client Storage

13 Network Attacks

Withdrawing the BGP Re-Routing Curtain: Understanding the Security Impact of BGP Poisoning through Real-World Measurements
该论文进行了迄今为止最大规模的BGP中毒测量研究,包括1460例成功/1888例中毒病例。作者发布了他们的数据集、源代码和本文最终结果的数据分析。作者复制了最近的安全文件做的伪装或被动。作者还构建了统计模型,作为利用BGP中毒作为操作员的第一步,而不需要进行主动测试或说服高级IT管理员。作者从多个角度评估中毒路径过滤的程度和影响。
IMP4GT: IMPersonation Attacks in 4G NeTworks

Practical Traffic Analysis Attacks on Secure Messaging Applications
CDN Judo: Breaking the CDN DoS Protection with Itself

14 Program Analysis

DeepBinDiff: Learning Program-Wide Code Representations for Binary Diffing
该论文提出了一种无监督的程序代码表示学习方法。作者依赖代码语义信息和程序范围的控制流信息来生成块嵌入。此外,作者还提出了一个k跳贪婪匹配算法,利用生成的块嵌入来寻找最优的扩散结果。实现了一个名为DeepBinDiff的原型,并用大量二进制文件评估了它的有效性和效率。结果表明,无论是交叉版本还是交叉优化级别的扩散,该工具都能在很大程度上优于最新的二进制扩散工具。
Precisely Characterizing Security Impact in a Flood of Patches via Symbolic Rule Comparison

15 Malware

Unicorn: Runtime Provenance-Based Detector for Advanced Persistent Threats
该论文提出了一种基于异常的APT检测器UNICORN,它有效地利用了数据来源分析。从造型到检测,UNICORN专门针对APTs的独特特性对其进行剪裁设计。UNICORN通过广泛而高效的图分析,探索了提供丰富上下文和历史信息的来源图,以在没有预先定义的攻击特征的情况下识别隐蔽的异常活动。利用图形绘制技术,它总结了具有空间效率的长时间运行系统执行,以对抗长时间内发生的慢动作攻击。UNICORN通过一种新的建模方法来理解系统的长期行为,进一步提高了它的检测能力。评估表明,UNICORN的性能优于现有的最先进的APT检测系统,并且能够高精度地检测现实生活中的APT场景。
Custos: Practical Tamper-Evident Auditing of Operating Systems Using Trusted Execution

You Are What You Do: Hunting Stealthy Malware via Data Provenance Analysis
OmegaLog: High-Fidelity Attack Investigation via Transparent Multi-layer Log Analysis
Prevalence and Impact of Low-Entropy Packing Schemes in the Malware Ecosystem
When Malware is Packin' Heat; Limits of Machine Learning Classifiers Based on Static Analysis Features

UIScope: Accurate, Instrumentation-free, and Visible Attack Investigation for GUI Applications

16 Private Computation and Learning

Trident: Efficient 4PC Framework for Privacy Preserving Machine Learning
该论文提出了一个主动安全的四方协议(4PC)和一个PPML框架,展示了它在四种最广泛的机器学习算法上的应用:线性回归、Logistic回归、神经网络和卷积神经网络。该4PC协议最多只能容忍一个恶意损坏,实际上比Gordon等人的效率更高。此外,作者还提出了与隐私保护机器学习特别相关的转换。
Secure Sublinear Time Differentially Private Median Computation
CloudLeak: Large-Scale Deep Learning Models Stealing Through Adversarial Examples

BLAZE: Blazing Fast Privacy-Preserving Machine Learning

17 Authentication

OcuLock: Exploring Human Visual System for Authentication in Virtual Reality Head-mounted Display
该论文提出了一个基于EOG的虚拟现实认证框架,通过视觉刺激来触发HVS的响应,并收集EOG来表征HVS。作者设计了一个记录比较驱动的认证方案,在该方案中,提取了不同的行为和生理特征,并做出了准确的认证决策。作者还对提议的OcuLock系统进行了广泛的评估,包括身份验证的可靠性性能、针对多种攻击的安全分析以及VR HMD身份验证的用户研究。
On the Resilience of Biometric Authentication Systems against Random Inputs
Strong Authentication without Temper-Resistant Hardware and Application to Federated Identities

18 Case Studies & Human Factors

A View from the Cockpit: Exploring Pilot Reactions to Attacks on Avionic Systems
该论文在一个飞行模拟器中,针对三个系统:交通避碰系统(TCAS)、近地警告系统(GPWS)和仪表着陆系统(ILS),对航空电子设备实施逼真的无线攻击。作者对30名空客A320飞行员进行了实验,以了解机组人员如何应对这些攻击。作者还在模拟器中分析并采访了实验结果。
Genotype Extraction and False Relative Attacks: Security Risks to Third-Party Genetic Genealogy Services Beyond Identity Inference
Complex Security Policy? A Longitudinal Analysis of Deployed Content Security Policies
Into the Deep Web: Understanding E-commerce Fraud from Autonomous Chat with Cybercriminals

Compliance Cautions: Investigating Security Issues Associated with U.S. Digital-Security Standards

19 Crypto

Let's Revoke: Scalable Global Certificate Revocation
该论文提出了一种可伸缩的全局吊销策略Let's Revoke,它解决了当前撤销检查的问题。Let's Revoke为每个证书引入一个新的唯一标识符,它作为包含吊销状态信息的动态大小的位向量的索引。位向量方法为客户端和证书颁发机构提供了更有效的吊销检查。作者将Let's Revoke与现有的吊销方案进行了比较,发现它比其他系统需要更少的存储和网络带宽。
Post-Quantum Authentication in TLS 1.3: A Performance Study
DISCO: Sidestepping RPKI's Deployment Barriers

Proof of Storage-Time: Efficiently Checking Continuous Data Availability
20 Hardware & Speculative Attacks
SPEECHMINER: A Framework for Investigating and Measuring Speculative Execution Vulnerabilities
该论文介绍了SPEECHMINER,一个自动探测和测量语音漏洞的软件框架。SPEECHMINER经验地建立了一个新的两相故障处理模型与语音漏洞的可利用性和推测窗口之间的联系。它允许在同一软件框架下测试一个完整的异常触发指令列表,该框架利用隐蔽通道技术和差异测试来获得微观体系结构状态变化的可见性。作者对SPEECHMINER的9种不同处理器类型进行了评估,检查了21个潜在的漏洞变体,确认了各种已知的攻击,并确定了几个新的变体。
ProtectIOn: Root-of-Trust for IO in Compromised Platforms
ConTExT: A Generic Approach for Mitigating Spectre

21 Privacy

Towards Plausible Graph Anonymization
该论文在一个飞行模拟器中,针对三个系统:交通避碰系统(TCAS)、近地警告系统(GPWS)和仪表着陆系统(ILS),对航空电子设备实施逼真的无线攻击。作者对30名空客A320飞行员进行了实验,以了解机组人员如何应对这些攻击。作者还在模拟器中分析并采访了实验结果。

Adversarial Classification Under Differential Privacy
Locally Differentially Private Frequency Estimation with Consistency

DESENSITIZATION: Privacy-Aware and Attack-Preserving Crash Report

2.2 S&P 2020录取情况

IEEE Symposium on Security and Privacy(简称 S&P)是信息安全领域四大顶级学术会议之一,始于1980年,今年是第41届。S&P 2020共录取104篇论文。

1 Machine Learning

AdGraph: A Graph-Based Approach to Ad and Tracker Blocking
本文介绍了一种新颖的基于图的机器学习方法AdGraph,用于检测广告和跟踪网络上的资源。它构建HTML结构,网络请求和网页的JavaScript行为的图形表示形式,并使用此唯一表示形式来训练分类器以识别广告和跟踪资源。AdGraph考虑了发生网络请求。AdGraph在在线使用方面既准确又有效。
Throwing Darts in the Dark? Detecting Bots with Limited Data Using Neural Data Augmentation

2 Blockchains

OHIE: Blockchain Scaling Made Simple
该论文提出了一种新颖的无许可区块链协议OHIE,OHIE根据需要构成尽可能多的比特币原始(简单)主干协议并行实例,以实现出色的吞吐量。OHIE的目标就是简单。因为最近提出的许多区块链共识协议变得越来越复杂,使得越来越难以提供其安全保证的证明。
FlyClient: Super-Light Clients for Cryptocurrencies

Replicated State Machines without Replicated Execution

Flash Boys 2.0: Frontrunning in Decentralized Exchanges, Miner Extractable Value, and Consensus Instability

VerX: Safety Verification of Smart Contracts

VeriSmart: A Highly Precise Safety Verifier for Ethereum Smart Contracts

3 Privacy Protect

SoK: Differential Privacy as a Causal Property
该论文提出了差异性隐私的关联和因果观点。在关联视图下,数据点之间依赖的可能性排除了对差异性隐私保证的简单陈述,因为它以单个更改的数据点为条件。但是,该论文表明,在因果视图下确实存在对差异性隐私的简单表征,以限制单个数据点的影响,而没有关于数据点的独立性假设。
Privacy Risks of General-Purpose Language Models
Are Anonymity-Seekers Just like Everybody Else? An Analysis of Contributions to Wikipedia from To

The Value of Collaboration in Convex Machine Learning with Differential Privacy
Private Resource Allocators and Their Applications
Enabling Rack-scale Confidential Computing Using Heterogeneous Trusted Execution Environment

Breaking and (Partially) Fixing Provably Secure Onion Routing
xMP: Selective Memory Protection for Kernel and User Space
Ask the Experts: What Should Be on an IoT Privacy and Security Label?
Automated Reverse Engineering and Privacy Analysis of Modern Cars
Towards Effective Differential Privacy Communication for Users’ Data Sharing Decision and Comprehension
Automatically Detecting Bystanders in Photos to Reduce Privacy Risks

4 User Behavior Analysis

Can Voters Detect Malicious Manipulation of Ballot Marking Devices?
该论文的研究是通过探讨几种警告方式(口头和视觉)对人类检测恶意选票修改的影响。作者进行了实证研究,研究了使用选票标记设备(BMD)的选民在打印选票上发现错误的程度。此外,作者还发现,程序性干预可以提高错误检测和报告率,有可能提高BMD提供的安全性。
Influencing Photo Sharing Decisions on Social Media: A Case of Paradoxical Findings
How Not to Prove Your Election Outcome

5 Security Analysis

Message Time of Arrival Codes: A Fundamental Primitive for Secure Distance Measurement
该论文着手研究到达代码的消息时间(MTAC)及其安全性。该论文首次正式定义物理层措施的安全要求,以保护ToA测量系统免受攻击。该论文能够提供现有MTAC的统一表示,并提出保护ToA测量系统免受现有机制无法解决的攻击的基本原理。且还系统地探索适用于所有启用ToA测量的信号调制技术的安全性和性能之间的权衡。
Security Update Labels: Establishing Economic Incentives for Security Patching of IoT Consumer Products

Cornucopia: Temporal Safety for CHERI Heaps
Are We Susceptible to Rowhammer? An End-to-End Methodology for Cloud Providers
SEVurity: No Security Without Integrity - Breaking Integrity-Free Memory Encryption with Minimal Assumptions
Tactical Provenance Analysis for Endpoint Detection and Response Systems

SPECTECTOR: Principled Detection of Speculative Information Flows
Towards a Natural Perspective of Smart Homes for Practical Security and Safety Analyses

A Tale of Sea and Sky: On the Security of Maritime VSAT Communications
Transys: Leveraging Common Security Properties across Hardware Designs
A Security Analysis of the Facebook Ad Library

6 Program Analysis

Semantic Understanding of Smart Contracts: Executable Operational Semantics of Solidity

Rigorous Engineering for Hardware Security: Formal Modelling and Proof in the CHERI Design and Implementation Process

MarkUs: Drop-in Use-After-Free Prevention for Low-Level Languages

A Programming Framework for Differential Privacy with Accuracy Concentration Bounds

JIT Leaks: Inducing Timing Side Channels through Just-In-Time Compilation

Burglars' IoT Paradise: Understanding and Mitigating Security Risks of General Messaging Protocols on IoT Clouds
TRRespass: Exploiting the Many Sides of Target Row Refresh

7 Network Security

A Stealthier Partitioning Attack against Bitcoin Peer-to-Peer Network
该论文提出了一种更隐蔽的比特币攻击,称之为EREBUS攻击,它允许网络对手控制目标比特币节点的对等连接,而无需任何路由操作,从而不留下任何攻击的控制平面证据。由于攻击仅使用数据平面攻击消息,因此任何控制平面监视系统都无法检测到该攻击;即使捕获到攻击的数据平面痕迹,攻击者也很容易拒绝执行攻击。
Do Cookie Banners Respect My Choice? Measuring Legal Compliance of Banners from IAB Europe's Transparency and Consent Framework
NetCAT: Practical Cache Attacks from the Network

ICLab: A Global, Longitudinal Internet Censorship Measurement Platform
Tactical Provenance Analysis for Endpoint Detection and Response Systems
Combating Dependence Explosion in Forensic Analysis Using Alternative Tag Propagation Semantics

SoK: Cyber Insurance - Technical Challenges and a System Security Roadmap

8 Web

Browsing Unicity: On the Limits of Anonymizing Web Tracking Data
该论文的作者与欧洲最大的受众测量协会合作,通过使用全面的跟踪数据集来评估可识别性和令人信服的匿名浏览数据的可能性。结果表明,通过泛化实现匿名化并不能充分保护匿名性。将浏览数据的单一性降低到可以忽略的级别将需要删除所有客户端和web域信息以及单击计时。该论文的结论是,虽然可以匿名地存储单一的粗化点击,但是任何复杂度更高的集合都将包含大量的假名数据
Unexpected Data Dependency Creation and Chaining: A New Attack to SDN

SoK: Cyber Insurance - Technical Challenges and a System Security Roadmap
Meddling Middlemen: Empirical Analysis of the Risks of Data-Saving Mobile Browsers
TARDIS: Rolling Back the Clock on CMS-Targeting Cyber Attacks

9 Crypto

EverCrypt: A Fast, Verified, Cross-Platform Cryptographic Provider
该论文展示了EverCrypt,是一个全面的、可验证的密码提供者,提供了灵活性、复用性和速度,该API,可供已验证和未验证的客户使用,它的多路复用实现,加上经过验证的零成本通用编程技术,在不牺牲性能的情况下简化了密码算法的验证,提出了一种基于依赖类型泛型编程的新方法,证明了C和程序集之间细粒度互操作的安全性和正确性。
Automatic Uncovering of Hidden Behaviors from Input Validation in Mobile Apps
Gesture Authentication for Smartphones: Evaluation of Gesture Password Selection Policies
Is FIDO2 the Kingslayer of User Authentication? A Comparative Usability Study of FIDO2 Passwordless Authentication
The Last Mile: High-Assurance and High-Speed Cryptographic Implementations
ZEXE: Enabling Decentralized Private Computation
Binsec/Rel: Efficient Relational Symbolic Execution for Constant-Time at Binary-Level
HydRand: Efficient Continuous Distributed Randomness
Path Oblivious Heap: Optimal and Practical Oblivious Priority Queue

Efficient and Secure Multiparty Computation from Fixed-Key Block Ciphers

10 wireless network

Even Black Cats Cannot Stay Hidden in the Dark: Full-band De-anonymization of Bluetooth Classic Devices
该论文分析了蓝牙经典(BT)的设计,并设计了一种强大的去匿名技术,该技术依赖于从帧编码泄漏的明显良性信息,并开发了一个基于软件定义的无线电(SDR)的嗅探器来验证这种攻击,该嗅探器可以进行完整的BT频谱分析,并实现建议的去匿名技术。该论文还证明了其开发的全频段SDR系统捕获整个BT频谱的好处。
BIAS: Bluetooth Impersonation AttackS

****11 Side Channel
Leveraging EM Side-Channel Information to Detect Rowhammer Attacks
该论文研究了某些电磁发射与rowchammer攻击之间的相关性,在此基础上提出了一种系统的rowchammer攻击检测方法RADAR(Rowchammer Attack Detection Via Radio)。该论文提出了一种方法来逆转扩频时钟对计算设备中高频时钟发出的EM侧信道信息的散射效应。并已经使用一个价值299美元的软件无线电设备实现了一个雷达原型,评估了不同场景下基于EM的rowhammer攻击检测的有效性和鲁棒性。

WaveSpy: Remote and Through-wall Screen Attack via mmWave Sensing
Pseudorandom Black Swans: Cache Attacks on CTR_DRBG

12 Sensor System

Detection of Electromagnetic Interference Attacks on Sensor Systems
本文针对一些低水平电磁干扰攻击,提出了一种新型检测系统。该系统基于以下思想:如果微控制器与物理世界进行交互时的传感器关闭,则微控制器读取的信号应为0V(或其他一些已知值)。使用此想法以一种对手无法预测的方式来调制传感器的输出。如果微控制器检测到传感器输出的波动,则可以检测到攻击信号。该检测方法的检测机制既有效又可靠。
SoK: A Minimalist Approach to Formalizing Analog Sensor Security

13 Fuzzing

Fuzzing JavaScript Engines with Aspect-Preserving Mutation
该论文提倡了一种新的方面保持突变方法,旨在保持种子输入的期望属性和前提条件。还开发了一个成熟的JavaScript fuzzer DIE,通过使用轻量级静态和动态类型分析实现了两种新的变异策略,即结构和类型保留。并且报告了48个新的bug,38个在负责任的公开过程中被修复。
IJON: Exploring Deep State Spaces via Fuzzing

SAVIOR: Towards Bug-Driven Hybrid Testing
RetroWrite: Statically Instrumenting COTS Binaries for Fuzzing and Sanitization
Pangolin: Incremental Hybrid Fuzzing with Polyhedral Path Abstraction

Krace: Data Race Fuzzing for Kernel File Systems
High Precision Open-World Website Fingerprinting
RAMBleed: Reading Bits in Memory without Accessing Them
****14 Protocol Analysis
TextExerciser: Feedback-Driven Text Input Exercising for Android Applications
该论文提出了第一个反馈驱动的输入练习器,它使用基于目标应用程序提示的约束求解器迭代生成文本输入,实现了一个文本输入练习器的原型,评估了TextExerciser在流行的Google Play应用程序上的性能。评估结果表明,textextexerciser比现有的动态分析工具实现了更高的代码覆盖率,并发现了更多的隐私泄露和漏洞。
SEIMI: Efficient and Secure SMAP-Enabled Intra-process Memory Isolation

Ex-vivo Dynamic Analysis Framework for Android Device Drivers
Kobold: Evaluating Decentralized Access Control for Remote NSXPC Methods on iOS
KARONTE: Detecting Insecure Multi-binary Interactions in Embedded Firmware
CrypTFlow : Secure TensorFlow Inference

15 cryptographic protocol

Transparent Polynomial Delegation and Its Applications to Zero Knowledge Proof
该论文提出了一种基于GKR的透明零知识论证(ZKP)协议,在构造表示语句的算术电路时,具有简洁的证明大小和验证时间。该方案的验证时间特别有效,至少比现有的ZKP系统快一个数量级,验证时间仅为几十毫秒。
Sync HotStuff: Simple and Practical Synchronous State Machine Replication
Towards Scalable Threshold Cryptosystems

16 Mobile and Malware

PMP: Cost-Effective Forced Execution with Probabilistic Memory Pre-planning
该论文提出了一种实用的强制执行技术。它不需要跟踪单个内存或算术指令。它也不需要按需分配内存。还提出了一种新的内存预规划方案,它提供了避免崩溃和伪程序依赖的概率保证。该方案下的执行与本机执行非常相似。一旦在开始时预先计划并初始化了内存,执行就可以正常进行,而不需要任何跟踪或动态分析。并且实现了一个名为PMP的原型,并在SPEC2000程序(包括gcc)和400个最新的真实恶意软件样本上对其进行了评估。结果表明,PMP是一种高效的强制执行技术。
An Analysis of Pre-installed Android Software
SPIDER: Enabling Fast Patch Propagation in Related Software Repositories
This PIN Can Be Easily Guessed: Analyzing the Security of Smartphone Unlock PINs
Intriguing Properties of Adversarial ML Attacks in the Problem Space

ICAS: An Extensible Framework for Estimating the Susceptibility of IC Layouts to Additive Trojans
Combating Dependence Explosion in Forensic Analysis Using Alternative Tag Propagation Semantics
The Many Kinds of Creepware Used for Interpersonal Attacks

15 BUG

Flaw Label: Exploiting IPv6 Flow Label
该论文介绍了Windows和Linux/Android环境下IPv6流标签生成算法。此外,通过利用Windows和Linux/Android设备使用的键控哈希算法生成设备ID来生成流标签,证明了跟踪Windows和Linux/Android设备的可行性。该论文还创建了一个跟踪服务器,并在19个不同的网络上测试了该技术。
Dragonblood: Analyzing the Dragonfly Handshake of WPA3 and EAP-pwd
Humpty Dumpty: Controlling Word Meanings via Corpus Modifications
SPECCFI: Mitigating Spectre Attacks Using CFI Informed Speculation
C3APSULe: Cross-FPGA Covert-Channel Attacks through Power Supply Unit Leakage
LVI: Hijacking Transient Execution through Microarchitectural Load Value Injection

Plundervolt: Software-Based Fault Injection Attacks against Intel SGX
The State of the Uniform: Attacks on Encrypted Databases beyond the Uniform Query Distribution
SoK: Understanding the Prevailing Security Vulnerabilities in TrustZone-Assisted TEE Systems

16 Neural Network

HopSkipJumpAttack: A Query-Efficient Decision-Based Attack
该论文提出了一种新的梯度方向无偏估计,在决策边界上完全基于对模型决策的访问,并提出了控制偏离边界误差的方法,并基于所提出的估计和作者的分析,设计了一系列算法HopSkipJumpAttack,该算法无超参数,查询效率高,并具有收敛性分析,还通过大量的实验,证明了该算法比几种最先进的基于决策的攻击具有更高的效率。在攻击几种广泛使用的防御机制时,它也具有竞争优势。
Neutaint: Efficient Dynamic Taint Analysis with Neural Networks

OAT: Attesting Operation Integrity of Embedded Devices

2.3 USENIX会议

USENIX Security是信息安全领域四大顶级学术会议之一,始于上世纪90年代初,USENIX Security 2020 按照春夏秋冬4期论文提交截止日期收录论文,共录取156篇论文。

1 Wireless Security (5篇)

A Formal Analysis of IEEE 802.11's WPA2: Countering the Kracks Caused by Cracking the Counters
该论文指出,IEEE 802.11 WPA2协议涉及各种以微妙的方式彼此交互的机制,也许正因为如此,不存在正式或加密的论据来表明核心协议的补丁确实阻止了相应的攻击。作者解决了这种情况,并提出了WPA2协议设计的广泛形式分析。该模型是第一个足够详细以检测KRACK攻击的模型;它包括诸如四次握手,组密钥握手,WNM睡眠模式,数据机密协议及其复杂交互之类的机制。作者分析提供了任何形式上的第一个安全性论证,即面对复杂的现代攻击,修补的WPA2协议可以满足其所声称的安全性保证。
Frankenstein: Advanced Wireless Fuzzing to Exploit New Bluetooth Escalation Targets
Breaking Secure Pairing of Bluetooth Low Energy Using Downgrade Attacks
You Are What You Broadcast: Identification of Mobile and IoT Devices from (Public) WiFi
Call Me Maybe: Eavesdropping Encrypted LTE Calls With ReVoLTE

2 Human Factors (5篇)

A Comprehensive Quality Evaluation of Security and Privacy Advice on the Web
为了确定在线文章中提供的安全建议的质量,该论文首先进行了一项以用户为导向的大规模测量研究,以识别264种在线安全和隐私建议文档中包含的374种独特的推荐行为。其次,作者开发和验证了用于评估质量的测量方法。第三,作者在1586位用户和41位专业安全专家的用户研究中部署了这些评估方法,以评估374条独特的安全建议。结果表明建议优先级的危机。大多数用户认为大多数建议至少是可行的,也是可以理解的。但是,用户和专家都难以确定该建议的优先级。
Understanding security mistakes developers make: Qualitative analysis from Build It, Break It, Fix It
Empirical Measurement of Systemic 2FA Usability
What Twitter Knows: Characterizing Ad Targeting Practices, User Perceptions, and Ad Explanations Through Users' Own Twitter Data
The Impact of Ad-Blockers on Product Search and Purchase Behavior: A Lab Experiment

3 Software Security and Verification(11篇)

Symbolic execution with SymCC: Don't interpret, compile!
该论文提出了一种基于编译的符号执行方法,其性能比最新的实现要好几个数量级。作者介绍了SymCC,这是一种基于LLVM的C和C ++编译器,可以在二进制文件中构建concolic执行。软件开发人员可以将其用作clang和的替代产品。 与KLEE相比,SymCC的速度提高了三个数量级,平均因子为12。它也优于Qsym,该系统最近在性能上比其他实施方案有了很大提高,提高了两个数量级和平均因子。
Sys: A Static/Symbolic Tool for Finding Good Bugs in Good (Browser) Code
Everything Old is New Again: Binary Security of WebAssembly
AURORA: Statistical Crash Analysis for Automated Root Cause Explanation
SmartVerif: Push the Limit of Automation Capability of Verifying Security Protocols by Dynamic Strategies
Datalog Disassembly
KOOBE: Towards Facilitating Exploit Generation of Kernel Out-Of-Bounds Write Vulnerabilities
Automatic Techniques to Systematically Discover New Heap Exploitation Primitives
The Industrial Age of Hacking
BScout: Direct Whole Patch Presence Test for Java Executables
MVP: Detecting Vulnerabilities using Patch-Enhanced Vulnerability Signatures

4 Mobile and Malware(10篇)

BigMAC: Fine-Grained Policy Analysis of Android Firmware
该论文开发了一个新的框架BIGMAC,从静态固件图片中提取、绘制和查询Android安全策略,而不需要根设备,并通过实例化进程、文件和IPC对象以98%的准确率恢复运行时的安全状态。作者结合了MAC、DAC、功能和标记的外部输入源,创建了一个支持数百万边缘的实例化、细粒度的整个系统攻击图。然后作者使用Prolog引擎提供了一个交互式用户界面来查询攻击图。作者评估了BIGMAC与三星S8+和LG G7 Firmware的对比。
From Needs to Actions to Secure Apps? The Effect of Requirements and Developer Practices on App Security
FANS: Fuzzing Android Native System Services via Automated Interface Analysis
Chaperone: Real-time Locking and Loss Prevention for Smartphones

Towards HTTPS Everywhere on Android: We Are Not There Yet
On Training Robust PDF Malware Classifiers
Measuring and Modeling the Label Dynamics of Online Anti-Malware Engines
FIRMSCOPE: Automatic Uncovering of Privilege-Escalation Vulnerabilities in Pre-Installed Apps in Android Firmware
Automatic Hot Patch Generation for Android Kernels
iOS, Your OS, Everybody's OS: Vetting and Analyzing Network Services of iOS Applications

5 Phishing, Spam, and Threat Intelligence(5篇)

Sunrise to Sunset: Analyzing the End-to-end Life Cycle and Effectiveness of Phishing Attacks at Scale
该论文通过测量大规模网络钓鱼攻击的端到端生命周期来隔离和识别这些检测漏洞。作者开发了一个独特的框架Golden Hour,它允许被动地测量网络钓鱼页面的受害者流量,同时在此过程中主动保护数以万计的帐户。在一年的时间里,作者的网络监视器记录了480万访问钓鱼网页的受害者,不包括爬虫流量。作者使用这些事件和相关的数据源来剖析网络钓鱼活动:从它们第一次上线到电子邮件分发,到访客流量,再到生态系统检测,再到最终的账户泄露。作者发现从一开始到最后一个受害者平均只需要21个小时。
PhishTime: Continuous Longitudinal Measurement of the Effectiveness of Anti-phishing Blacklists
Who's Calling? Characterizing Robocalls through Audio and Metadata Analysis
See No Evil: Phishing for Permissions with False Transparency
A different cup of TI? The added value of commercial threat intelligence

6 Trusted Execution Environments(11篇)

HybCache: Hybrid Side-Channel-Resilient Caches for Trusted Execution Environments
该论文提出了一种用于集合关联缓存的灵活和软分区的通用机制,并提出了一种混合缓存架构,称为HybCache。HybCache可以配置为仅对隔离的执行域有选择地应用侧信道弹性缓存行为,同时为非隔离执行提供常规的缓存行为,容量和性能。隔离域可以包括一个或多个进程,代码的特定部分或受信任的执行环境作者证明了,通过最少的硬件修改和内核支持,HybCache可以仅为隔离执行提供侧通道弹性缓存,其性能开销为3.5–5%,而对其余执行工作负载则不产生性能开销。
CopyCat: Controlled Instruction-Level Attacks on Enclaves
An Off-Chip Attack on Hardware Enclaves via the Memory Bus
Civet: An Efficient Java Partitioning Framework for Hardware Enclaves
BesFS: A POSIX Filesystem for Enclaves with a Mechanized Safety Proof
SENG, the SGX-Enforcing Network Gateway: Authorizing Communication from Shielded Clients
APEX: A Verified Architecture for Proofs of Execution on Remote Devices under Full Software Compromise
PARTEMU: Enabling Dynamic Analysis of Real-World TrustZone Software Using Emulation
PHMon: A Programmable Hardware Monitor and Its Security Use Cases
Horizontal Privilege Escalation in Trusted Applications
TeeRex: Discovery and Exploitation of Memory Corruption Vulnerabilities in SGX Enclaves

7 Network Security-网络安全(6篇)

EPIC: Every Packet Is Checked in the Data Plane of a Path-Aware Internet
该论文提出了EPIC,这是一组数据平面协议,可提供越来越强大的安全性,可以满足三个要求:首先,网络运营商仍然需要能够实施自己的策略以排除不经济的路径并在数据平面上执行这些决策。其次,终端主机应该能够验证网络实际遵循了它们的转发决定。最后,中间路由器和接收者都应该能够验证数据包的来源。与同类系统相比,EPIC协议的通信开销要低得多。
ShadowMove: A Stealthy Lateral Movement Strategy
Poison Over Troubled Forwarders: A Cache Poisoning Attack Targeting DNS Forwarding Devices
Programmable In-Network Security for Context-aware BYOD Policies
A Longitudinal and Comprehensive Study of the DANE Ecosystem in Email
NXNSAttack: Recursive DNS Inefficiencies and Vulnerabilities

8 Web Security and Privacy(6篇)

Shim Shimmeny: Evaluating the Security and Privacy Contributions of Link Shimming in the Modern Web
该论文以Facebook的实际部署为例,对链接匀场的安全性和隐私贡献做出了大规模的实证评估。结果表明,即使在现代网络中,链接填充也可以为用户广泛地提供有意义的安全性和隐私性收益。对于作者所观察到的相当大数量的人群(尤其是在以移动设备为中心的发展中国家)来说,传统浏览器客户端的普及率很高,这些优势最为明显。作者讨论了这些收益与潜在成本之间的权衡。
Cached and Confused: Web Cache Deception in the Wild
A Tale of Two Headers: A Formal Analysis of Inconsistent Click-Jacking Protection on the Web
Retrofitting Fine Grain Isolation in the Firefox Renderer
Zero-delay Lightweight Defenses against Website Fingerprinting
Achieving Keyless CDNs with Conclaves

9 Automotive and Drone Security(6篇)

Stealthy Tracking of Autonomous Vehicles with Cache Side Channels
该论文首先表明了感知和控制算法的自适应行为可能会引入新的侧通道漏洞,从而揭示车辆或其环境的物理特性。第二,介绍了统计学习模型,这些模型可以根据缓存访问模式推断AMCL算法的状态,并根据AMCL状态的轨迹预测车辆的路线或位置。第三,使用记录在城市道路上的真实传感器数据,在现实的软件堆栈上实施和演示了攻击。发现结果表明,自动驾驶软件需要强大的时序通道保护来保护位置隐私。
Towards Robust LiDAR-based Perception in Autonomous Driving: General Black-box Adversarial Sensor Attack and Countermeasures
SAVIOR: Securing Autonomous Vehicles with Robust Physical Invariants
From Control Model to Program: Investigating Robotic Aerial Vehicle Accidents with MAYDAY
Drift with Devil: Security of Multi-Sensor Fusion based Localization in High-Level Autonomous Driving under GPS Spoofing
Plug-N-Pwned: Comprehensive Vulnerability Analysis of OBD-II Dongles as A New Over-the-Air Attack Surface in Automotive IoT

10 Privacy Enhancing Technologies(5篇)

PCKV: Locally Differentially Private Correlated Key-Value Data Collection with Optimized Utility
该论文提出了一个本地差分私有键值数据收集框架,该框架利用相关扰动来增强实用性。作者通过两种协议PCKV-UE(基于一元编码)和PCKV-GRR(基于广义随机响应)实例化了其框架,在其中设计了高级的填充和采样机制以及改进的非交互均值估计器。在合成数据集和实际数据集上的实验结果表明,与最新机制相比,在相同的本地差异隐私(LDP)保证下,该协议在频率和均值估计方面都具有更好的实用性。
Actions Speak Louder than Words: Entity-Sensitive Privacy Policy and Data Flow Analysis with PoliCheck
Walking Onions: Scaling Anonymity Networks while Protecting Users
Differentially-Private Control-Flow Node Coverage for Software Usage Analysis
Visor: Privacy-Preserving Video Analytics as a Cloud Service

11 Embedded/IoT Security(6篇)

Shattered Chain of Trust: Understanding Security Risks in Cross-Cloud IoT Access Delegation
在该论文中作者基于其自主开发的半自动验证工具,报告了有关现实世界物联网访问授权的系统研究。其研究揭示了这些委派机制中普遍存在的安全风险,从而使对手能够未经授权地访问受害者的设备或冒充这些设备来触发其他设备。作者通过对它们的端到端攻击来确认这些机制中是否存在关键安全漏洞,并进一步进行了度量研究。
HALucinator: Firmware Re-hosting Through Abstraction Layer Emulation
Silhouette: Efficient Protected Shadow Stacks for Embedded Systems
P2IM: Scalable and Hardware-independent Firmware Testing via Automatic Peripheral Interface Modeling
COUNTERFOIL: Verifying Provenance of Integrated Circuits using Intrinsic Package Fingerprints and Inexpensive Cameras
Hall Spoofing: A Non-Invasive DoS Attack on Grid-Tied Solar Inverter

12 Machine Learning(11篇)

Updates-Leak: Data Set Inference and Reconstruction Attacks in Online Learning
该论文研究了黑盒ML(机器学习)模型更新前后的输出变化是否会泄漏用于执行更新的数据集的信息,即更新集。这构成了针对黑盒ML模型的新攻击面,并且这种信息泄漏可能会损害ML模型所有者的知识产权和数据隐私。作者提议按照编码器-解码器公式进行四次攻击,从而可以推断出更新集的各种信息。特别是,作者提出了一种基于生成对抗网络(GAN)的混合生成模型(CBM-GAN),但其中包含允许重建准确样本的重建损失。
Exploring Connections Between Active Learning and Model Extraction
Hybrid Batch Attacks: Finding Black-box Adversarial Examples with Limited Queries
High Accuracy and High Fidelity Extraction of Neural Networks
Adversarial Preprocessing: Understanding and Preventing Image-Scaling Attacks in Machine Learning
TextShield: Robust Text Classification Based on Multimodal Embedding and Neural Machine Translation
Fawkes: Protecting Privacy against Unauthorized Deep Learning Models
Stolen Memories: Leveraging Model Memorization for Calibrated White-Box Membership Inference
Local Model Poisoning Attacks to Byzantine-Robust Federated Learning
Justinian's GAAvernor: Robust Distributed Learning with Gradient Aggregation Agent
Interpretable Deep Learning under Fire

13 Microarchitectural Attacks(6篇)

Data Recovery from “Scrubbed” NAND Flash Storage: Need for Analog Sanitization
PKU Pitfalls: Attacks on PKU-based Memory Isolation Systems
Medusa: Microarchitectural Data Leakage via Automated Attack Synthesis

V0LTpwn: Attacking x86 Processor Integrity from Software

DeepHammer: Depleting the Intelligence of Deep Neural Networks through Targeted Chain of Bit Flips
SpecFuzz: Bringing Spectre-type vulnerabilities to the surface

14 Financial Tech and Voting(5篇)

Security Analysis of Unified Payments Interface and Payment Apps in India
该论文对尚未发布的UPI 1.0协议进行了第一次深入的安全分析,该协议为印度许多流行的移动支付应用程序提供了一个通用的支付接口,并允许不同应用程序的用户之间进行银行间转账。作者展示了如何从无法访问UPI服务器的对手的角度系统地反向工程这一复杂的应用层协议。作者还在UPI协议中发现了一些细微的设计缺陷,对手可以利用安卓设计中已知的缺陷,利用attacker控制的应用程序来构建可伸缩的远程攻击。作者展示了对手如何在不知道用户的情况下进行攻击。作者讨论了在设计此类协议时应考虑的经验教训和潜在缓解策略。
Cardpliance: PCI DSS Compliance of Android Applications
The Ballot is Busted Before the Blockchain: A Security Analysis of Voatz, the First Internet Voting Application Used in U.S. Federal Elections
VoteAgain: A scalable coercion-resistant voting system
Boxer: Preventing fraud by scanning credit cards

15 Systems Security(5篇)

(Mostly) Exitless VM Protection from Untrusted Hypervisor through Disaggregated Nested Virtualization
该论文介绍了CloudVisor-D,这是一种有效的嵌套虚拟机管理程序设计,同时包含强大的保护和高性能。CloudVisor-D的核心思想是通过将主要的保护逻辑与每个来宾VM隔离到受保护的Guardian-VM中来分解嵌套的虚拟机管理程序。Guardian-VM由嵌套的虚拟机管理程序安全地隔离和保护,并为来宾VM的大多数特权操作提供安全服务。通过利用最新的硬件功能,来宾VM的大多数特权操作都不需要退出嵌套虚拟机管理程序的VM,这是先前设计中性能下降的主要来源。实验评估表明,即使对于I / O密集型基准测试,CloudVisor-D所产生的性能开销也可以忽略不计。
Donky: Domain Keys – Efficient In-Process Isolation for RISC-V and x86
DECAF: Automatic, Adaptive De-bloating and Hardening of COTS Firmware
McTiny: Fast High-Confidence Post-Quantum Key Erasure for Tiny Network Servers
Temporal System Call Specialization for Attack Surface Reduction

16 Analysis of Crypto(6篇)

Big Numbers - Big Troubles: Systematically Analyzing Nonce Leakage in (EC)DSA Implementations
该论文对密码实现中的随机数泄漏进行了系统分析。特别是,作者扩展了DATA(一种开源的旁渠道分析框架)来检测随机数泄漏。其分析在涉及随机数的所有基本计算步骤中发现了多个未知的随机数泄漏漏洞。除其他外,作者还发现了Bignumber实现中的内在问题,如果机密接近单词边界,则这些固有问题会破坏(EC)DSA实现所要求的固定时间保证。作者发现,在OpenSSL和LibreSSL中延迟调整Bignumbers的大小会产生一个高度准确且易于利用的副通道。作者还在OpenSSL和BoringSSL的恒定时间标量乘法中发现了微小但富有表现力的泄漏。
Estonian Electronic Identity Card: Security Flaws in Key Management
The Unpatchable Silicon: A Full Break of the Bitstream Encryption of Xilinx 7-Series FPGAs
Automating the Development of Chosen Ciphertext Attacks

17 Specific User Populations(5篇)

An Observational Investigation of Reverse Engineers’ Processes
该论文的目的是为改进逆向工程工具的交互设计提供见解。作者介绍了反向工程师(N = 16)的半结构化观察访谈研究的结果。每个观察都调查了逆向工程师在探查程序时提出的问题,他们如何回答这些问题以及在整个逆向工程过程中做出的决策。从访谈回复中,作者提取了逆向工程过程的模型,分为三个阶段:概述,子组件扫描和重点实验。随着反向工程师的心理表述变得更加具体,每个分析阶段的结果将进入下一个阶段。作者发现,逆向工程师通常在前两个阶段使用静态方法,而在最后阶段使用动态方法,并在每个阶段都扮演着重要角色,但角色各不相同。基于这些结果,作者为逆向工程工具提供了五种交互设计指南。
The Tools and Tactics Used in Intimate Partner Surveillance: An Analysis of Online Infidelity Forums
DatashareNetwork: A Decentralized Privacy-Preserving Search Engine for Investigative Journalists
"I am uncomfortable sharing what I can't see": Privacy Concerns of the Visually Impaired with Camera Based Assistive Applications
'I have too much respect for my elders': Understanding South African Mobile Users' Perceptions of Privacy and Current Behaviors on Facebook and WhatsApp

18 Side Channel Attacks(6篇)

RELOAD+REFRESH: Abusing Cache Replacement Policies to Perform Stealthy Cache Attacks
该论文指出,如果不改变缓存状态并因此强制对所观察的进程进行驱逐,就不可能观察到共同驻留进程的缓存活动,这一事实有力地帮助了检测。我们证明了这种广泛持有的假设是不正确的。通过巧妙地使用高速缓存替换策略,可以跟踪受害者进程的高速缓存访问而无需强制逐出根据受害者的数据。因此,可以绕过依赖于这些驱逐的在线检测机制,因为它们不会检测到引入的RELOAD + REFRESH攻击,攻击需要对缓存替换策略有深刻的理解。作者提出了一种恢复替换策略的方法,并将其应用于最近五代英特尔处理器。通过进一步经验证明了,在加密实现上,RELOAD+REFRESH的性能与其他广泛使用的缓存攻击相媲美。
Timeless Timing Attacks: Exploiting Concurrency to Leak Secrets over Remote Connections

Cache Telepathy: Leveraging Shared Resource Attacks to Learn DNN Architectures
Certified Side Channels
NetWarden: Mitigating Network Covert Channels while Preserving Performance
TPM-FAIL: TPM meets Timing and Lattice Attacks

19 Implementations of Crypto(5篇)

Scaling Verifiable Computation Using Efficient Set Accumulators
可验证的外包系统将大量计算任务卸载到远程服务器,但要求远程服务器提供简洁的证明(称为SNARK),以证明服务器正确执行了计算。实际上,状态更新的可验证外包是通过更新Merkle树的叶子,重新计算所得的Merkle根以及使用SNARK证明状态更新正确完成来完成的。该论文将现有技术和新颖技术结合使用,在SNARK内部实现了RSA累加器,并将其用作Merkle树的替代品。作者专门优化了累加器,使其与SNARK兼容。实验表明,与使用Merkle树提交到当前状态的现有方法相比,生成的系统可以大大降低成本。
Pixel: Multi-signatures for Consensus
SANNS: Scaling Up Secure Approximate k-Nearest Neighbors Search
MIRAGE: Succinct Arguments for Randomized Algorithms with Applications to Universal zk-SNARKs

Secure Multi-party Computation of Differentially Private Median

20 Authentication(5篇)

That Was Then, This Is Now: A Security Evaluation of Password Generation, Storage, and Autofill in Browser-Based Password Managers
该论文指出,先前的研究已经发现了现有密码管理器中的重大漏洞,特别是在基于浏览器的密码管理器中。但从那时起,已经过去了五年,目前尚不清楚密码管理器是否仍然易受攻击或是否已解决已知的安全问题。为了回答这个问题,作者评估了十三种流行的密码管理器。该评估是对密码管理器中密码生成的首次分析,通过查找几种非随机字符分布,并确定实例,结果其中所生成的密码容易受到在线和离线猜测攻击。对于密码存储和自动填充,作者复制了以前的评估结果,并表明尽管自从先前的评估以来,密码管理器已有所改进,但仍然存在重大。根据结果,作者确定了需要避免的密码管理器,并提供了有关如何改进现有密码管理器的建议。
Composition Kills: A Case Study of Email Sender Authentication
Detecting Stuffing of a User’s Credentials at Her Own Accounts
Liveness is Not Enough: Enhancing Fingerprint Authentication with Behavioral Biometrics to Defeat Puppet Attacks
Human Distinguishable Visual Key Fingerprints
SHA-1 is a Shambles: First Chosen-Prefix Collision on SHA-1 and Application to the PGP Web of Trust
A Spectral Analysis of Noise: A Comprehensive, Automated, Formal Analysis of Diffie-Hellman Protocols

21 Fuzzing(11篇)

FuzzGuard: Filtering out Unreachable Inputs in Directed Grey-box Fuzzing through Deep Learning
该论文提出了一种基于深度学习的方法,以在执行目标程序之前预测输入的可到达性(即是否错过目标),帮助定向灰盒模糊测试(DGF)过滤掉不可达的输入以提高模糊测试的性能。为了使用DGF进行深度学习,作者设计了一套新技术来解决标记数据不平衡和训练过程中时间不足的问题。此外,作者还实现了称为FuzzGuard的建议方法,并为其配备了最新的DGF。对45个实际漏洞的评估表明,FuzzGuard将AFLGo的模糊处理效率提高了17.1倍。
FuzzGen: Automatic Fuzzer Generation

ParmeSan: Sanitizer-guided Greybox Fuzzing
EcoFuzz: Adaptive Energy-Saving Greybox Fuzzing as a Variant of the Adversarial Multi-Armed Bandit
MUZZ: Thread-aware Grey-box Fuzzing for Effective Bug Hunting in Multithreaded Programs
Analysis of DTLS Implementations Using Protocol State Fuzzing
Agamotto: Accelerating Kernel Driver Fuzzing with Lightweight Virtual Machine Checkpoints
USBFuzz: A Framework for Fuzzing USB Drivers by Device Emulation

GREYONE: Data Flow Sensitive Fuzzing

Fuzzing Error Handling Code using Context-Sensitive Software Fault Injection
Montage: A Neural Network Language Model-Guided JavaScript Engine Fuzzer

22 Data Security/ Secure Computation(5篇)

SEAL: Attack Mitigation for Encrypted Databases via Adjustable Leakage
为了有效防御基于可搜索加密(SE)的系统上的泄漏滥用攻击,该论文提出了SEAL,这是一系列具有可调泄漏功能的新型SE方案。在SEAL中,隐私丢失量以搜索或访问模式的泄漏位表示,可以在设置时定义。实验表明,当保护只有几位泄漏的访问,足以让现有的,甚至新的更具侵略性的攻击失败。因此,SEAL可以成为一种建立高效、健壮的加密数据库的有前途的方法。
Pancake: Frequency Smoothing for Encrypted Data Stores
Droplet: Decentralized Authorization and Access Control for Encrypted Data Streams
Secure parallel computation on national scale volumes of data
Delphi: A Cryptographic Inference Service for Neural Networks

23 Voice and Speech(5篇)

Light Commands: Laser-Based Audio Injection Attacks on Voice-Controllable Systems
该论文通过将光物理转换为声音,提出了对麦克风的新型信号注入攻击。作者展示了攻击者如何通过将调幅光对准麦克风的孔径,将任意音频信号注入目标麦克风。然后,继续说明了这种影响如何导致远程语音控制系统的语音命令注入攻击。作者研究了各种使用亚马逊Alexa、苹果Siri、Facebook门户和谷歌助手的产品,展示了如何利用光线在距离110米远的两栋大楼内控制这些设备。最后,总结了总结了可能的软件和硬件防御该攻击。
SkillExplorer: Understanding the Behavior of Skills in Large Scale
Devil’s Whisper: A General Approach for Physical Adversarial Attacks against Commercial Black-box Speech Recognition Devices
Void: A fast and light voice liveness detection system
Preech: A System for Privacy-Preserving Speech Transcription

24 Blockchains(5篇)

BlockSci: Design and applications of a blockchain analysis platform
该论文介绍BlockSci,一个用于区块链分析的开源软件平台。BlockSci在支持不同的区块链和分析任务方面具有多种功能。它合并了一个内存中的分析(而不是事务性)数据库,使其比使用通用图形数据库快了几个数量级。作者描述了BlockSci的设计,并提供四个分析说明了其功能,阐明了加密货币的安全性,隐私性和经济性。
Remote Side-Channel Attacks on Anonymous Transactions
ETHBMC: A Bounded Model Checker for Smart Contracts

TXSPECTOR: Uncovering Attacks in Ethereum from Transactions
An Ever-evolving Game: Evaluation of Real-world Attacks and Defenses in Ethereum Ecosystem

3 未来值得研究的方向

3.1 研究方向

通过对三个会议的对比与分析,可以看到有关Web、隐私保护、机器学习、侧信道、系统与应用安全、加密、认证、网络安全等方向的论文较多,比较热门,而最近新兴的一些如区块链和加密货币、移动互联安全、硬件安全、用户行为分析、物联网安全等也很热门,所以未来值得研究的方向有:区块链、物联网安全、用户行为分析、移动互联安全、web、隐私保护、机器学习。

4 总结

通过对四大顶会论文的查找与分析,开拓了自己的视野,增长了见识,了解到了许多知识。虽然对只是简单的了解了论文的研究内容,但也算是对目前的主流的方向有了认识。因为之前对一些知识不太了解,再加上自己的英语不好,Emmmm,有些翻译出来的词,真的是看上去有点懵,然后再单独查一下有没有这个词,才能知道是什么。尤其要说的是,一定要尽早完成自己的任务,555555555555,要是不拖着,也不至于多出来156篇论文,(7月1号我看到了USENIX会议的录取列表),大补特补。

posted @ 2020-07-05 21:22  zzxb  阅读(6611)  评论(1编辑  收藏  举报