Publications Search Results

Now showing 1 - 10 of 53
  • Publication
    Not all Conflicts are Created Equal: Automated Error Resolution in RPKI Deployments
    We explore one of the central obstacles hindering Internet-wide adoption of RPKI: erroneous ROAs. The errors cause the ROV-filtering networks to drop legitimate traffic while leaving them exposed to hijack attacks. The fear of disconnection demotivates enforcement of ROV obviating the security benefits of RPKI.In this work we devise metrics for differentiating errors from traffic hijack attacks and evaluate them experimentally. We develop an extended ROV based on our metrics and integrate it into the ROV implementation of RIPE NCC, we call our extended validator ROV++. We evaluate the effectiveness of ROV++ in classifying conflicting BGP announcements via Internet experiments and simulations on empirically derived datasets.
  • Publication
  • Publication
    Machine Learning Analysis of IP ID Applications
    The IP identifier (IP ID) in the IP header has become enormously popular as a side channel leaking valuable information on destinations. In the recent decades, the researchers have exploited the IP ID in a variety of different applications, from estimating outgoing server traffic, to covert communication and to remotely understanding firewall rules and port status. However, the complexity of inferring IP ID due to high fluctuating traffic rates from multiple sources leaves it an open question how practical the applications that leverage IP ID are.We perform the first Internet wide study of IP ID behaviour in the Internet and evaluate how practical it is to build applications on top of IP ID. We analyse experimentally the applications on the dataset of IP ID values that we collected. We show that our SVM classifier can achieve the accuracy of the IP ID prediction of more than 99%.
  • Publication
    Negotiating PQC for DNSSEC
    ( 2021)
    Shrishak, Kris
    ;
    Domain Name System Security Extensions (DNSSEC) provides authentication and integrity to Domain Name System (DNS) through the use of digital signatures based on public-key cryptography. Quantum computers threaten public key cryptography and DNSSEC is unprepared. As the process to change algorithms in DNSSEC involves a lot of overhead, requires significant investment and takes many years, we advocate for deployment of long term cryptography for DNSSEC. In this work we explore the challenges and obstacles towards deployment of post-quantum signatures and explain that smooth adoption towards quantum-safe ciphers can be achieved with cipher-suite negotiation for DNSSEC.Cipher-suite negotiation, which DNSSEC currently does not support, ensures that the best cryptographic algorithms supported by the server and the resolver are used. Servers usually do not deprecate old algorithms because they are unaware whether resolvers support new algorithms. The signals in cipher-suite negotiation inform the servers and the resolvers of algorithm support that creates a feedback loop that could accelerate adoption of post-quantum signatures and the deprecation of old algorithms while preventing packet fragmentation. As a consequence, cipher-suite negotiation can contribute towards a greater adoption of DNSSEC.
  • Publication
    SSLChecker
    In this work we devise a SSLChecker tool, for testing server side vulnerabilities in SSL/TLS implementations. We integrate into our tool central vulnerabilities exposing to attacks and evaluate SSLChecker over them. The goal of SSLChecker is to help: (1) the web server operators to identify vulnerabilities and mitigate them, and (2) to warn users of accessing potentially vulnerable servers. We set SSLChecker as public service with open source code on github.
  • Publication
    DNS-over-TCP considered vulnerable
    The research and operational communities believe that TCP provides protection against IP fragmentation attacks and recommend that servers avoid sending DNS responses over UDP but use TCP instead. In this work we show that IP fragmentation attacks also apply to servers that communicate over TCP. Our measurements indicate that in the 100K-top Alexa domains there are 393 additional domains whose nameservers can be forced to (source) fragment IP packets that contain TCP segments. In contrast, responses from these domains cannot be forced to fragment when sent over UDP. Our study not only shows that the recommendation to use TCP instead of UDP in order to avoid attacks that exploit fragmentation is risky, but it also unveils that the attack surface due to fragmentation is larger than was previously believed. We evaluate IP fragmentation-based DNS cache poisoning attacks against DNS responses over TCP.
  • Publication
    Off-path VoIP Interception Attacks
    The proliferation of Voice-over-IP (VoIP) technologies make them a lucrative target of attacks. While many attack vectors have been uncovered, one critical vector has not yet received attention: hijacking telephony via DNS cache poisoning. We demonstrate practical VoIP hijack attacks by manipulating DNS responses with a weak off-path attacker. We evaluate our attacks against popular telephony VoIP systems in the Internet and provide a live demo of the attack against Extensible Messaging and Presence Protocol at https://sit4.me/M4.
  • Publication
    Fragmentation Attacks on DNS over TCP
    The research and operational community believe that TCP provides protection against IP fragmentation based attacks and recommend that servers avoid sending responses over UDP and use TCP instead. In this work we show for the first time that IP fragmentation attacks may also apply to communication over TCP. We perform a study of the nameservers in the 100K-top Alexa domains and find that 454 domains are vulnerable to IP fragmentation attacks. Of these domains, we find 366 additional domains that are vulnerable only to IP fragmentation attacks on communication with TCP. We also find that the servers vulnerable to TCP fragmentation can be forced to fragment packets to much smaller sizes (of less than 292 bytes) than servers vulnerable to UDP fragmentation (not below 548 bytes). This makes the impact of the attacks against servers vulnerable to fragmentation of TCP segments much more detrimental. Our study not only shows that the recommendation to use TCP and avoid UDP is risky but it also shows that the attack surface due to fragmentation is larger than was previously believed. We evaluate known IP fragmentation-based DNS cache poisoning attacks against DNS responses over TCP.
  • Publication
    PrivInferVis: Towards Enhancing Transparency over Attribute Inference in Online Social Networks
    The European GDPR calls, besides other things, for innovative tools to empower online social networks (OSN) users with transparency over risks of attribute inferences. In this work, we propose a novel transparency-enhancing framework for OSN, PrivInferVis, to help people assess and visualize their individual risks of attribute inference derived from public details from their social graphs in different OSN domains. We propose a weighted Bayesian model as the underlying method for attribute inference. A preliminary evaluation shows that our proposal outperforms baseline algorithms on several evaluation metrics significantly. PrivInferVis provides visual interfaces that allow users to explore details about their (inferred and self-disclosed) data and to understand how inference estimates and related scores are derived.
  • Publication
    Let's Downgrade Let's Encrypt
    Following the recent off-path attacks against PKI, Let's Encrypt deployed in 2020 domain validation from multiple vantage points to ensure security even against the stronger on-path MitM adversaries. The idea behind such distributed domain validation is that even if the adversary can hijack traffic of some vantage points, it will not be able to intercept traffic of all the vantage points to all the nameservers in a domain. In this work we show that two central design issues of the distributed domain validation of Let's Encrypt make it vulnerable to downgrade attacks: (1) the vantage points are selected from a small fixed set of vantage points, and (2) the way the vantage points select the nameservers in target domains can be manipulated by a remote adversary. We develop off-path methodologies, based on these observations, to launch downgrade attacks against Let's Encrypt. The downgrade attacks reduce the validation with 'multiple vantage points to multiple nameservers', to validation with 'multiple vantage points to a single attacker-selected nameserver'. Through experimental evaluations with Let's Encrypt and the 1M-Let's Encrypt-certified domains, we find that our off-path attacker can successfully launch downgrade attacks against more than 24.53% of the domains, rendering Let's Encrypt to use a single nameserver for validation with them. We then develop an automated off-path attack against the 'single-server'-domain validation for these 24.53% domains, to obtain fraudulent certificates for more than 107K domains, which constitute 10% of the 1M domains in our dataset. We also evaluate our attacks against other major CAs and compare the security and efforts needed to launch the attacks, to those needed to launch the attacks against Let's Encrypt. We provide recommendations for mitigations against our attacks.