*12:17*[Pub][ePrint]

Get an update on changes of the IACR web-page here. For questions, contact *newsletter (at) iacr.org*.
You can also receive updates via:

To receive your credentials via mail again, please click here.

You can also access the full news archive.

Further sources to find out about changes are CryptoDB, ePrint RSS, ePrint Web, Event calender (iCal).

Name: Enrico Thomae

Topic: About the Security of Multivariate Quadratic Public Key Schemes

Category: public-key cryptography

Description: The primary goal of this thesis is to evaluate the security of multivariate quadratic public key schemes. We investigate three main topics related to the security of MQ-schemes, namely the MQ-Problem, the IP-Problem and the MinRank-Problem.

\r\nSection 2 discusses the MQ-Problem, which relates to direct pre-image attacks using the\r\npublic key, i.e. finding x for a given y and P(x) = y, which is known to be difficult in\r\ngeneral. In section 2.1 we provide a brief survey on algorithms to solve such systems, like F4, F5, XL and MutantXL. We recap the complexity analysis of the first three algorithms and provide a detailed complexity analysis of the latter. Our contribution is a proof of theorem 2.7 which is hopefully simpler than that in [CKPS, Section 8]. Further we derived theorem 2.29 and thus confirmed results from Yang and Chen [YC04a] in a different way.

\r\nIn section 2.2 we present a new direct attack on the Unbalanced Oil and Vinegar signature scheme, which forces to raise parameters in order to obtain the same\r\nsecurity level again. More generally we present an algorithm to solve underdetermined\r\nsystems of MQ-equations faster than before.

\r\nSection 3 presents the main part of this work and is dedicated to algebraic key recovery\r\nattacks on MQ-schemes.\r\nUnfortunately naive algebraic attacks are usually far from being efficient due to the large number of variables. So we first formalize the underlying class of problems and introduce the Isomorphism of Polynomials with partial Knowledge (IPpK) Problem in section 3.3. We relate this new problem to known problems, like the Isomorphism of Polynomials Problem with one and two secrets. Our main contribution is to provide a general algebraic\r\nframework to tackle the IPpK-Problem. Therefore we generalize the notion of equivalent keys to so-called good keys. In a nutshell equivalent keys allow to reduce the number of variables of an algebraic attack. Good keys further reduce the number of vari[...]

2013-06-19

Hello, In case you want to follow up this forum - but not to poll it every other day, there are several options: * RSS feed: http://eprint.iacr.org/forum/rss.php * IACR News system (eMail, Twitter) via "Forum": http://www.iacr.org/news/ Best, Christopher From: 2013-19-06 04:56:53 (UTC)

2013-06-18

The problem with radical redesign is that it is hard to understand what change has caused which effect. I suggest that we as a community focus on one problem at a time. If we want to focus on multiple problems, maybe each conference should attack one at a time, so at least each variable can be tested separately. Let\'s start with the problem of low quality reviews. Here is a modest initial proposal based on an economic model: Each review should have two components: (1) technical summary and feedback, and (2) subjective evaluation wholly supported by technical evaluation in (1) The technical summary should be presented to the authors before decisions are made, and the authors will rate reviews based on understanding. So will other PC members (anonymously). The results will be used to rate PC members and reviewers and provide them with tokens. PC members and reviewers will need to spend these tokens to get their papers published at top conferences in the future. The monetary system will need to be worked out, but we can let junior researchers borrow tokens from the central bank at the start of their careers so as not to harm their initial careers. But eventually everyone has to pay in quality reviews for papers that they want to publish. These are initial thoughts and the proposal should certainly be refined to address potential abuses. For example, technical parts of the review should be devoid of all subjective opinions and hidden praise, so that the temptation to flatter the authors for earning tokens can be avoided. Also, probably feedback from authors of papers in the bottom 33% should not be counted towards awarding tokens. Amit From: 2013-18-06 09:28:41 (UTC)

In computer forensics, log files are indispensable resources that support auditors in identifying and understanding system threats and security breaches. If such logs are recorded locally, i.e., stored on the monitored machine itself, the problem of log authentication arises: if a system intrusion takes place, the intruder might be able to manipulate the log entries and cover her traces. Mechanisms that cryptographically protect collected log messages from manipulation should ideally have two properties: they should be *forward-secure* (the adversary gets no advantage from learning current keys when aiming at forging past log entries), and they should be *seekable* (the auditor can verify the integrity of log entries in any order or access pattern, at virtually no computational cost).

We propose a new cryptographic primitive, a *seekable sequential key generator* (SSKG), that combines these two properties and has direct application in secure logging. We rigorously formalize the required security properties and give a provably-secure construction based on the integer factorization problem. We further optimize the scheme in various ways, preparing it for real-world deployment. As a byproduct, we develop the notion of a *shortcut one-way permutation* (SCP), which might be of independent interest.

Our work is highly relevant in practice. Indeed, our SSKG implementation has become part of the logging service of the systemd system manager, a core component of many modern commercial Linux-based operating systems.

Most security models for authenticated key exchange (AKE) do not explicitly model the associated certification system, which includes the certification authority (CA) and its behaviour. However, there are several well-known and realistic attacks on AKE protocols which exploit various forms of malicious key registration and which therefore lie outside the scope of these models. We provide the first systematic analysis of AKE security incorporating certification systems (ASICS). We define a family of security models that, in addition to allowing different sets of standard AKE adversary queries, also permit the adversary to register arbitrary bitstrings as keys. For this model family we prove generic results that enable the design and verification of protocols that achieve security even if some keys have been produced maliciously. Our approach is applicable to a wide range of models and protocols; as a concrete illustration of its power, we apply it to the CMQV protocol in the natural strengthening of the eCK model to the ASICS setting.