Este trabalho faz uma análise crítica dos elementos jurídicos-tecnológicos de uma assinatura digital certificada digitalmente. O primeiro aspecto a ser abordado advém da verificação da competência para o desenvolvimento da atividade de certificação, em decorrência da natureza jurídica do certificado digital. Consoante se verificou, o certificado digital é o instrumento hábil a assegurar a autenticidade dos documentos eletrônicos por meio de uma assinatura digital. Dessa forma, equipara-se ao ato de reconhecimento de firma, atividade notarial desenvolvida pelos Cartórios Notariais, de acordo com a competência fixada no artigo 236 da Constituição da República Federativa do Brasil. Todavia, segundo regra presente na Medida Provisória 2.200-2/01, desde 2001 essa atividade vem sendo desenvolvida sob a competência do Governo Federal, através do Instituto Nacional de Tecnologia da Informação - ITI (Autoridade Certificadora Raiz da Infraestrutura de Chaves Públicas do Brasil. Como decorrência tem-se que a Medida Provisória 2.200-2/01 é inconstitucional, uma vez que não respeita regra de competência material fixada pela Constituição da República Federativa do Brasil para o desenvolvimento da atividade notarial. Sob um prisma tecnológico, têm-se que a ICP-Brasil, por meio de seu Comitê Gestor, fixa expressamente qual a tecnologia que deve ser empregada para a produção das assinaturas digitais. Neste caminho, até maio de 2006, entre outros, foi indicado o algoritmo criptográfico de função hash MD5 para a geração das assinaturas digitais com autenticidade e integridade garantidas por lei. Todavia, o MD5 perdeu sua utilidade em 2004, quando foi quebrado, ocasionando a possibilidade de fraudes, inclusive a geração de documentos eletrônicos forjados. Sem dúvida, a legislação brasileira vinha assegurando validade jurídica e força probante a documentos eletrônicos assinados com algoritmo criptográfico de função hash MD5 que poderiam ter sido forjados. Para que o documento eletrônico assinado digitalmente possa ser amplamente utilizado em relações sociais é preciso que regras jurídicas e tecnológicas sejam respeitadas, sob pena de se criar uma enorme insegurança social. / This work presents a critical analysis of the technology and law aspects of certified digital signatures, and their implementation in Brazil. We discuss and verify the competency rules that apply to the certification activity according to the legal nature of the digital certificate. A digital certificate is the instrument that secures the authenticity of an electronic document by means of a digital signature. According to the article 236 of the Brazilian Constitution, authenticity certifications are of exclusive competence of public notaries. Nevertheless, based on an under constitutional statute, digital certification has being conducted by the Federal Government thru its National Institute of Information Technology (Instituto Nacional de Tecnologia da Informação - ITI), who is responsible for the Brazilian public key root certification authority. We found that the statute that supports those activities (Medida Provisória 2.200-2/01) is unconstitutional, and therefore invalid and unenforceable, since it does not satisfy constitutional rules of material competency. Under a technology view, we find that the Managing Committee of the Brazilian Public Key Infrastructure explicitly defines the technology to be used in digital signatures. According to that ruling, until may 2006, among others, the MD5 hashing algorithm was used to generate digital signatures with statutory presumption of authenticity and integrity. Nevertheless, MD5 lost its technical usefulness in 2004, when it was broken, and became prone to fraud such as the generation of forged electronic documents. There is no doubt that Brazilian legislation gave legal value and probatory force to electronic documents signed using the already broken MD5 hashing algorithm that could very well had been forged. Digitally signed electronic documents can only be successfully used if legal rules and the technological aspects be fully understood and respected. Otherwise, the result will be high levels of uncertainty in law relations.

Les nouvelles technologies ont profondément modifié nos usages mais ne sont pas que synonymes d’avantages pour leurs utilisateurs. Elles ont en effet de lourdes conséquences sur notre vie privée, ce qui est bien souvent sous-estimé. Les utilisateurs de moyens de paiement électronique ne réalisent par exemple pas toujours que leurs transactions peuvent révéler des informations particulièrement intimes à leur sujet, telles que leur localisation, leur état de santé ou mêmes leurs croyances.Nous nous intéressons dans ce mémoire aux techniques cryptographiques permettant de concilier les exigences de sécurité traditionnelles et le respect de la vie privée. Dans une première partie nous étudions deux cas particuliers, celui du paiement anonyme et celui de l’authentification anonyme. Nous proposons de nouvelles constructions qui offrent une meilleure efficacité que les solutions existantes, ouvrant ainsi la voie à de réelles applications pratiques. Chacun de ces systèmes fait l’objet d’une étude de sécurité montrant qu’ils offrent de solides garanties sous des hypothèses raisonnables.Cependant, afin de satisfaire des contraintes techniques souvent très fortes dans ces contextes, il peut être nécessaire d’optimiser ces constructions qui nécessitent souvent un nombre significatif de calculs. Dans une deuxième partie nous proposons donc des moyens pour améliorer l’efficacité des opérations et algorithmes les plus fréquemment utilisés. Chacune de ces contributions peut présenter un intérêt au-delà du contexte de l’anonymat. / New technologies offer greater convenience for end-users but usually at the cost of a loss in terms of privacy, which is often underestimated by the latter. For example, knowledge by a third party of the information related to a transaction is far from insignificant since it may reveal intimate details such as whereabouts, religious beliefs or health status.In this thesis, we are interested in cryptographic technics allowing to reconcile both security requirements and user’s privacy. In a first part, we will focus on two specific cases: anonymous payment and anonymous authentication. We propose new constructions, improving the efficiency of state-of-the-art solutions, which make all the features of these primitives more accessible for practical applications. We provide a detailed security analysis for each scheme, proving that they achieve the expected properties under reasonable assumptions.However, to fulfill the strong technical constraints of these use cases, it may be necessary to optimize these constructions which are usually rather complex. To this end, we propose in a second part, new solutions to improve the efficiency of most common operations and algorithms. Each of these contributions is not restricted to anonymous systems and thus may be of independent interest.

Graduation date: 2003

This paper discusses the hardware foundations of the cryptosystem employed by the Xbox(TM) video game console from Microsoft. A secret boot block overlay is buried within a system ASIC. This secret boot block decrypts and verifies portions of an external FLASH-type ROM. The presence of the secret boot block is camouflaged by a decoy boot block in the external ROM. The code contained within the secret boot block is transferred to the CPU in the clear over a set of high-speed busses where it can be extracted using simple custom hardware. The paper concludes with recommendations for improving the Xbox security system. One lesson of this study is that the use of a high-performance bus alone is not a sufficient security measure, given the advent of inexpensive, fast rapid prototyping services and high-performance FPGAs.

Thesis (M.S.)--Rochester Institute of Technology, 2007. / Typescript. Includes bibliographical references (leaf 246).

In 2001, Boneh, Lynn, and Shacham presented a pairing-based signature scheme known as the BLS signature scheme. In 2003, Boneh, Gentry, Lynn, and Shacham presented the first aggregate signature scheme called the BGLS aggregate signature scheme. The BGLS scheme allows for N users with N signatures to combine their signatures into a single signature. The size of the resulting signature is independent of N. The BGLS signature scheme enjoys roughly the same level of security as the BLS scheme. In 2005, Waters presented a pairing-based signature scheme which does not assume the existence of random oracles. In 2007, Lu, Ostrovsky, Sahai, Shacham, and Waters presented the LOSSW aggregate signature scheme which does not assume the existence of random oracles. The BLS, BGLS, Waters, and LOSSW authors each chose to work with a restricted class of pairings. In each scheme, it is clear that the scheme extend to arbitrary pairings. We present the schemes in their full generality, explore variations of the schemes, and discuss optimizations that can be made when using specific pairings. Each of the schemes we discuss is secure assuming that the computational Diffie-Hellman (CDH) assumption holds. We improve on the security reduction for a variation of the BGLS signature scheme which allows for some restrictions of the BGLS signature scheme can be dropped and provides a stronger guarantee of security. We show that the BGLS scheme can be modified to reduce public-key size in presence of a certifying authority, when a certain type of pairing is used. We show that patient-free bit-compression can be applied to each of the scheme with a few modifications.

Internet commerce continues to grow rapidly. Over 60% of US households use the internet to shop online. A secure payment protocol is required to support this rapid growth. A new payment protocol was recently invented at IBM. We refer to the protocol as SPP or Secure Payment Protocol. This thesis presents a protocol analysis of SPP. It is essential that a thorough security analysis be done on any new payment protocol so that we can better understand its security properties. We first develop a method for analyzing payment protocols. This method includes a list of desirable security features and a list of proofs that should be satisfied. We then present the results of the analysis. These results validate that the protocol does contain many security features and properties. They also help understand the security properties and identify areas where the protocol can be further secured. This led us to extend the design of the protocol to enhance its security. This thesis also presents a prototype implementation of SPP. Three software components were implemented. They are the Electronic Wallet component, the merchant software component and the Trusted Third Party component. The architecture and technologies that are required for implementation are discussed. The prototype is then used in performance measurement experiments. Results on system performance as a function of key size are presented. Finally, this thesis presents an extension of SPP to support a two buyer scenario. In this scenario one buyer makes an order while another buyer makes the payment. This scenario enables additional commerce services.

In 2001, Boneh, Lynn, and Shacham presented a pairing-based signature scheme known as the BLS signature scheme. In 2003, Boneh, Gentry, Lynn, and Shacham presented the first aggregate signature scheme called the BGLS aggregate signature scheme. The BGLS scheme allows for N users with N signatures to combine their signatures into a single signature. The size of the resulting signature is independent of N. The BGLS signature scheme enjoys roughly the same level of security as the BLS scheme. In 2005, Waters presented a pairing-based signature scheme which does not assume the existence of random oracles. In 2007, Lu, Ostrovsky, Sahai, Shacham, and Waters presented the LOSSW aggregate signature scheme which does not assume the existence of random oracles. The BLS, BGLS, Waters, and LOSSW authors each chose to work with a restricted class of pairings. In each scheme, it is clear that the scheme extend to arbitrary pairings. We present the schemes in their full generality, explore variations of the schemes, and discuss optimizations that can be made when using specific pairings. Each of the schemes we discuss is secure assuming that the computational Diffie-Hellman (CDH) assumption holds. We improve on the security reduction for a variation of the BGLS signature scheme which allows for some restrictions of the BGLS signature scheme can be dropped and provides a stronger guarantee of security. We show that the BGLS scheme can be modified to reduce public-key size in presence of a certifying authority, when a certain type of pairing is used. We show that patient-free bit-compression can be applied to each of the scheme with a few modifications.

Trapdoor functions, introduced in the seminal paper of Diffie and Hellman, are a fundamental notion in modern cryptography. Informally, trapdoor functions are injective functions that are easy to evaluate but hard to invert unless given an additional input called the trapdoor. Specifically, the classical security notion considered for trapdoor functions is one-wayness, which asks that it be hard to invert (except with very small probability) a uniformly random point in the range without the trapdoor. Motivated by the demands of emerging applications of cryptography as well as stronger security properties desired from higher-level cryptographic primitives constructed out of trapdoor functions, this thesis studies new strengthenings to the classical notion of one-way trapdoor functions and their applications. Our results are organized along two separate threads, wherein we introduce two new cryptographic primitives that strengthen the notion of one-wayness for trapdoor functions in different ways: Deterministic Encryption: Our notion of deterministic (public-key) encryption addresses the weaknesses of using trapdoor functions directly for encryption articulated by Goldwasser and Micali, to the extent possible without randomizing the encryption function (whereas Goldwasser and Micali address them using randomized encryption). Specifically, deterministic encryption ensures no partial information is leaked about a high-entropy plaintext or even multiple correlated such plaintexts. Deterministic encryption has applications to fast search on encrypted data, securing legacy protocols, and ``hedging' randomized encryption against bad randomness. We design a conceptually appealing semantic-security style definition of security for deterministic encryption as well as an easier-to-work-with but equivalent indistinguishability style definition. In the random oracle model of Bellare and Rogaway, we show a secure construction of deterministic encryption for an unbounded number of arbitrarily correlated high-entropy plaintexts based on any randomized encryption scheme, as well as length-preserving such construction based on RSA. In the standard model, we develop a general framework for constructing deterministic encryption schemes based on a new notion of ``robust' hardcore functions. We show a secure construction of deterministic for a single high-entropy plaintext based on exponentially-hard one-way trapdoor functions; single-message security is equivalent to security for an unbounded number of messages drawn from a block-source (where each subsequent message has high entropy conditioned on the previous). We also show a secure construction of deterministic encryption for a bounded number of arbitrarily correlated high-entropy plaintexts based on the notion of lossy trapdoor functions introduced by Peikert and Waters. paragraph*{Adaptive Trapdoor Functions:} Our notion of adaptive trapdoor functions asks that one-wayness be preserved in the presence of an inversion oracle that can be queried on some range points. The main application we give is the construction of black-box chosen-ciphertext secure public-key encryption from weaker general assumptions. (``Black-box' means that the specific code implementing the trapdoor function is not used in the construction, which typically incurs a huge efficiency cost.) Namely, we show such a construction of chosen-ciphertext secure public-key encryption from adaptive trapdoor functions. We then show that adaptive trapdoor functions can be realized from the recently introduced notions of lossy trapdoor functions by Peikert and Waters and correlated-product secure trapdoor functions by Rosen and Segev. In fact, by extending a recent result of Vahlis, we show adaptivity is strictly weaker than the latter notions (in a black-box sense). As a consequence, adaptivity is the weakest security property of trapdoor functions known to imply black-box chosen-ciphertext security. Additionally, by slightly extending our framework and considering ``tag-based' adaptive trapdoor functions, we obtain exactly the chosen-ciphertext secure encryption schemes proposed in prior work, thereby unifying them, although the schemes we obtain via adaptive trapdoor functions are actually more efficient. Finally, we show that adaptive trapdoor functions can be realized from a (non-standard) computational assumption on RSA inversion, leading to a very efficient RSA-based chosen-ciphertext secure encryption scheme in the standard model.

<p>The recent increase in performance of embedded processors has enabled the use of computationally heavy asymmetric cryptography in small and power efficient embedded systems. The goal of this thesis is to analyze whether it is possible to use this type of cryptography to enhance the security in access systems.</p><p>This report contains a literature study of the complications related to access systems and their functionality. Also a basic introduction to cryptography is included.</p><p>Several cryptographic algorithms were implemented using the public library LibTomCrypt and benchmarked on an ARM7-processor platform. The asymmetric coding schemes were ECC and RSA. The tested symmetric algorithms included AES, 3DES and Twofish among others. The benchmark considered both codesize and speed of the algorithms.</p><p>The two asymmetric algorithms, ECC and RSA, are possible to be used in an ARM7 based access system. Although, both technologies can be configured to finish the calculations within a reasonable time-frame of 10 Sec, ECC archives a higher security level for the same execution time. Therefore, an implementation of ECC would be preferable since it is faster and requires less resources. Some further suggestions of improvements to the implementation is discussed in the final chapters.</p>