We\‘re looking into the fundamental problem of watermarking finite numeric sets. The wide area of applicability of the problem ranging from numeric database content to stock market analysis data, makes it especially intriguing when considering a generic solution and particularities of its various applications. Given a range of associated numeric constraints and assumptions we provide a generic solution and analyze associated attacks. We further present several problem applicability domains and relate it to some of our past and ongoing research in watermarking semistructures.

We present a Hierarchical Hybrid Network architecture for wireless networks.  In such a network, mobile nodes are hierarchically organized into groups.  Different groups can have different routing protocols.  Mobile nodes communicate with nodes outside their groups through the group agents.  The groups are highly autonomous.  This architecture is flexible and scalable.  We conduct experiments to compare the new architecture with Ad Hoc networks.  The new architecture has a more stable topology and higher throughput when the number of mobile nodes is large.  The objective of our research is to set up a survivable, secure mobile wireless network.

The development of inter-networked enterprises created a new computing environment in which information assurance is critical. The objective of this article is to investigate the information assurance needs of today

The design and operation of autonomous agents to assure information in ERP systems of inter-networked enterprises are investigated. A variable information assurance implementation model is proposed based on the AIMIS model, and a risk assessment procedure is applied. The protocols and models needed to support variable assurance are introduced and their performance is assessed. Experimentation shows the possibility to reduce the processing time of requests without decreasing the proportion of trusted requests, compared to a systematic total assurance approach.

Differential and linear cryptanalysis, two of the most important techniques in modern block cipher cryptanalysis, still lack a sound, generally-applicable analysis of their success probabilities. In this paper, we present an analytical calculation of the success probability of differential and linear cryptanalytic attacks. In addition to a formulation of the success probability, the analysis yields a formulation of the attacks’ data requirements as well, and it also finds the size of the aimed key information in differential cryptanalysis as one of the factors affecting the success probability. We also discuss the issue of key dependence in linear cryptanalysis, which can be a serious limitation for this technique of attack.

We construct several examples showing that two distinct pulses can have identical intensity autocorrelations and power spectra, from which we infer that retrieval methods based on these two datasets alone produce ambiguous solutions.

We construct several examples of distinct asymmetric-symmetric pulse pairs with identical or essentially identical intensity autocorrelations and power spectra. From these examples we infer that pulse retrieval methods based on these two data sets alone produce ambiguous solutions. Furthermore, we used the constructed pulse pairs as test cases to assess the degree of difference in the corresponding interferometric autocorrelations. In several cases we found that the differences in the interferometric autocorrelations are sufficiently small that they might be quite challenging to distinguish in a practical experimental context.

We study the limiting distribution of the height in a generalized trie in which external nodes are capable to store up to b items (the so called b-tries).  We assume that such a tree is build from n random strings (items) generated by an unbiased memoryless source.  In this paper, we discuss the case when b and n are both large.  We shall identify six natural regions of the height distribution that should be compared to three regions obtained for fixed b.  We prove that for most n, the limiting distribution is concentrated at the single point k1 = [log2 (n/b)] + 1 as n,b approach infinity.  We observe that this is quite different than the height distribution for fixed b, in which case the limiting distribution is of an extreme value type concentrated arount (1 + 1/b)log2 n.  We derive our results by analytic methods, namely generating functions and the saddle point method.  We also present some numerical verification of our results.

For a Markovian source, we analyze the Lempel-Ziv parsing scheme that partitions sequences into phrases such that a new phrase is the shortest phrase not seen in the past. We consider three models: In the Markov Independent model, several sequences are generated independently by Markovian sources, and the ith phrase is the shortest prefix of the ith sequence that was not seen before as a phrase (i.e., a prefix of previous (I - 1) sequences).  In the other two models, only a single sequence is generated by a Markovian source. In the second model, called the Gilbert-Kadota model, a fixed number of phrases is generated according to the Lempel-Ziv algorithm, thus producing a sequence of a variable (random) length. In the last model, known also as the Lempel-Ziv model, a string of fixed length is partitioned into a variable (random) number of phrases. These three models can be efficiently represented and analyzed by digital search trees that are of interest to other algorithms such as sorting, searching and pattern matching. In this paper, we concentrate on analyzing the average profile (i.e., the average number of phrases of a given length), the typical phrase length, and the length of the last phrase. We obtain asymptotic expansions for the mean and the variance of the phrase length, and we prove that appropriately normalized phrase length in all three models tends to the standard normal distribution, which leads to bounds on the average redundancy of the Lempel-Ziv code. For Markov Independent model, this finding is established by analytic methods (i.e., generating functions, Mellin transform and depoissonization), while for the other two models we use a combination of analytic and probabilistic analyses.

Two fundamental problems in combinatorics on words and string manipulation are string matching and sequence comparison. In string matching one searches for all occurrences of a given string, understood as a sequence of consecutive symbols, in a text. In sequence comparison a subsequence rather than a string is searched in a text.  The string-matching problem has been extensively studied in literature from algorithmic and probabilistic points of view. The sequence comparison problem, also known as hidden pattern problem, is harder and it has been much less investigated. In this paper we study the number of occurrences of a given pattern w of length m as a subsequence in a random text of length n generated by a memoryless source. In particular, we consider two versions of this problem, namely the unconstrained one in which the subsequence w can appear anywhere in the text, and the constrained one that puts bounds on the distances between symbols of the word w. We determine the mean and the variance of the number of occurrences, and establish a Gaussian limit law. These results are obtained via combinatorics on words, formal languages, and methods of analytic combinatorics based on generating functions and moment methods. The motivation to study this problem comes from an attempt at finding a reliable threshold for intrusion detections, from textual data processing applications, and from molecular biology.

Current approaches for authorization on Web servers are mostly based on a predefined set of users or domains.  They are not suitable for Internet Web sites where the user set is unbounded and authorized users can be non-predefined.  We propose an authorization approach that applies role-based access control (RBAC) to WWW.  Under this approach, system administrators predefine roles, role-permission relations, and the policies that assign roles to users (user-role assignment policy).  The system automatically collects trustworthy information (valid evidence) and assigns roles to Internet users according to user-role assignment policies.  Trustworthiness information plays an important role in user-role assignment.  The validity of evidence is assessed based on the trustworthiness information of the evidence provider.  In addition, system administrators can specify the trustworthiness constraints that users have to satisfy for holding roles.  In this paper, the schema of using RBAC on the Web and the procedure of user-role assignment are presented.  The classification and evaluation of trustworthiness are discussed.