Suchith Anand is currently doing PhD in "Metaheuristic Optimisation techniques in location based services" at the School of Computing, University of Glamorgan, Wales. After graduating in Civil Engineering, he worked for Department for International Development (DFID) research projects at Indian Institute of Technology, Madras and later at South Bank University, London. He completed his Masters degree from School of Informatics, City University, London. His research interests are in the field of Location Based Services, Mobile GIS, Map Generalization and Network Modelling. More details of his work are available at Google search keyword "Suchith"

Geographic Information System (GIS) in very simple terms can be defined as the technology to store, analyse, query and visualize spatial data. MobileGIS refers to the use of geographic data in the field on mobile devices such as networked Personal Digital Assistants (PDAs). The main components for Mobile GIS are a global positioning system (GPS), a handheld computer (e.g. a PDA) and a communication network, with GIS acting as the backbone .Location Based Services (LBS) refers to the infrastructure needed to provide various services to a user based on the user's location. LBS applications act according to a geographic trigger like input of a place name, postcode, position of a GPS user etc. MobileGIS is a relatively new technology but with the arrival of high bandwidth mobile networks its application potential has increased tremendously. There is a huge amount of available digital geographic information which can be re-purposed for mobile GIS application. This coupled with the ability to filter and personalize content by reference to a user's physical location will provide compelling business and research opportunities.

Brian Behlendorf founded CollabNet, with O'Reilly & Associates, in July 1999. The company provides tools and services based on open source methods. Before launching CollabNet, Behlendorf was co-founder and CTO of Organic Online, a Web design and engineering consultancy located in San Francisco. During his five years at Organic, Behlendorf helped create Internet strategies for dozens of Fortune 500 companies. During that time, he co-founded and contributed heavily to the Apache Web Server Project, co-founded and supported the VRML (Virtual Reality Modeling Language) effort, and assisted several IETF working groups, particularly the HTTP standardization effort. Before starting Organic, Behlendorf was the first Chief Engineer at Wired Magazine and later HotWired, one of the first large-scale publishing Web sites. Behlendorf is Director of the Apache Software Foundation. He also serves as a Technical Advisor to Critical Path (CPTH) and Topica.

The worldwide Open Source community thrives on some key functions: transparency of the source code within applications, the flexibility to change that code to adapt to new situations, and the radically inclusive development process that allows any individual to go from "ordinary user" to "core developer" based on their own skillset, interest, and ideas. The corpus of high quality open source software, and the archive of knowlege and process contained in the artifact databases and mail archives of every Open Source project, form a giant classroom; here developers can find whether the methodologies and theories taught formally work when applied to real-life situations. Brian will talk about his personal involvement in the launching of the open source Apache Software Foundation, how the ASF works and what drives its developers, and how Open Source answers any good engineer's basic desire to constantly learn more and learn from the best. Finally he'll talk about how his own company, CollabNet, brings this and other facets of Open Source development ideas into large organizations and corporations.

Mr. Prasanna Desikan Ph.D student Department of Computer Science, University of Minnesota

Prasanna Desikan is a Ph.D student at the Department of Computer Science, University of Minnesota. He is pursuing the doctoral program under the guidance of Prof. Jaideep Srivastava. His research interests are in Data Mining, Web Mining and Link Analysis. His primary research work is on Link Analysis techniques pertaining to Web Mining research. He also works as apart of the MINDS intrusion detection system team at the University of Minnesota. More details about his work can be found at: http://www-users.cs.umn.edu/~desikan/

Link Analysis on the Web is used for a wide variety of purposes, ranging from ranking pages returned from a web search engine to identifying Web communities. Web data has been evolving over time, reflecting the ongoing trends. These changes in data in the temporal dimension reveal new kind of information. This information has not captured the attention of the Web mining research community to a large extent. In the first part of our work we identify the key dimensions that span the design space for link analysis on the Web, which makes the similarities and complementarities of various approaches clearer. In the second part of our work we examine another important dimension of Web Mining, namely temporal dimension. In the third part of the work, we present the application of Link analysis of evolving graphs in a different domain to detect e-mail spamming machines. . Our experiments verify the significance of such analysis and also point to future directions in this area. The approach we take is generic and can be applied to other domains, where data can be modeled as graph, such as network intrusion detection or social networks.

Prof. TRN Rao, University of Louisiana at Lafayette

TRN Rao received his B.Sc from Andhra University, DIISc from Indian Institute of Science and M.S. and Ph.D. degrees in Electrical Engineering from University of Michigan, Arbor.¨T He served in the faculty of University of Maryland, CollegePark, Md (1996-1975) and Southern Methodist University,¨T Dallas,TX (1975-80). He is presently with Center for Advanced Computer Studies in University of Louisiana at Lafayette, and holds the Z. L. Loflin Chair Professorship and is the director of CRYPTO Laboratory.¨T He graduated twenty five Ph.D.s and authored /edited over a dozen volumes (texts and monographs) in the areas of Error Correcting Codes, Fault Tolerant Computing and Cryptography.
Dr. Rao received several awards, which include IEEE Fellow (1984), ACM Fellow (1994), IEEE Information Theory Best paper of the year (1993) and ACM lecture of the year (1992). He served as Fulbright Professor to India and Japan during 1986-87 and 1993-94.

Public-key crypto-algorithms are widely employed for authentication, signatures, secret-key generation and access control. The new range of public-key sizes for RSA and DSA has gone up to 1024 bits and beyond. Elliptic-curve key range is from 162 bits to 256 bits. Many varied software and hardware algorithms are being developed for implementation for smart-card crypto-coprocessors and for public-key infrastructure. We begin with an algorithm from Aryabhatiya, for solving the indeterminate equation a∙x + c = b∙y of degree one (also known as Diophantine equation) and its extension to solve the system of two residues X mod mi = Xi¨T (for i =1, 2).¨T This contribution known as Aryabhatiya Algorithm (AA) is very profound in the sense that the problem of two congruences was solved with just one modular inverse operation and a modular reduction to a smaller modulus than the compound modulus. We extend AA to any set of t residues and is stated as Aryabhata Remainder Theorem (ART) and an iterative algorithm is given to solve for t moduli mi (i=1, 2,^§Ö, t).¨T The ART, which has much in common with Extended Euclidean Algorithm (EEA), Chinese Remainder Theorem (CRT) and Garner^§âs algorithm (GA), is shown to have a complexity comparable or better than CRT and GA.

Bruhadeshwar Bezawada, Software Engineering and Network Systems Laboratory, Department of Computer Science and Engineering, Michigan State University, East Lansing, USA.

Many applications such as conferencing, distributed interactive simulations, and news dissemination are group oriented. In these applications, it is necessary to secure the group communication as the data is sensitive or it requires the users to pay for it. We focus on the problem of secure group communication in dynamic groups. In this problem, a group of users communicate using a shared group key. Due to the dynamic nature of these groups, to preserve secrecy, it is necessary to change and distribute the new group key whenever the group membership changes. While the group key is being distributed, the group communication needs to be interrupted until the rekeying is complete. This interruption is especially necessary if the rekeying is done because a user has been revoked (or voluntarily left) from the group. In our work, we propose adaptive solutions for distributing the new group key when users are revoked from the group. Our solutions reduce two important cost parameters: the duration of interruption of the group communication and the encryption cost for distributing the new group key. In our solutions, the group controller can dynamically change the algorithm for key distribution to adapt to changing application requirements. Moreover, we show that our solutions allow the group controller to effectively manage heterogeneous groups where users have different requirements/capabilities.

Soma Paul Researcher LTRC

Soma Paul is doing PhD in Applied Linguistics at University of Hyderabad, awaiting her defense. She has published in both 'Linguists Conference' and Naltural Language Processing conferences.

I will attempt to present a semantically-grounded constraint-based account of the composition of Compound Verbs (CV) of Indo-Aryan languages within a lexicalist framework of Head-Driven Phrase Structure Grammar (HPSG). Building a system that identifies CVs in a given text and outputs their structural representations is significant for sentential parsing, machine translation and speech analysis. A Bangla CV is a kind of multi-word expression in which the first member (V1) chooses between the usual conjunctive participial form ^æe while the second member (V2) bears the inflection. 1. mee-Ta: heS-e uTh-lo girl-cl laugh-cp rise-3 pt ^ÿThe girl burst into laughter^ð The boldfaced sequence of verbs in (1) is a CV. In Hindi-Urdu, the V1 is a bare verb form as illustrated in 2: 2. anju-ne khat kal likh lia: Anju-erg letter yesterday write take-pt ^ÿAnju wrote the letter yesterday^ð A given V1 does not combine with every V2. The combinatorial well-formedness between a V1 and a V2 depends on the semantic compatibility between the two verbs. I will propose to impose a constraint on the semantic component of verbal lexical signs, which are arranged in a system of multiple inheritance hierarchy. This constraint controls the unification of a V1 with a V2 to ensure that the grammar licenses only well-formed CV sequences. I will propose for a two-level representation of the semantic component of verbs: 1.A participant level contains ^ñgrammatically relevant^ò information that defines the relations among the participants involved in the situation denoted by the semantic type of the verb. 2.A supra-lexical level includes information related to temporality and aspect. The proposal is implemented in HPSG framework by introducing two attributes THEM (representing participant level) and GRAM (representing supra-lexical level) as the values of the semantic feature SEM. The postulation of the two levels for organizing the meaning of verbs is significant for constituting compound verb constructions. The semantic constraint that governs the unification of V1 with a particular V2 is declared on the value of THEM of the verb lexeme type, while the semantic compounding principle that builds the temporal and aspectual features of the resultant CV is operative on the value of GRAM. The GRAM value of V2 is inherited by the CV of which the V2 is a part. Just like a simple verb, CVs that are a kind of lexical variant of their V1 associate project the sentences they head. As far as constituting the phrasal structure of the CVs is concerned, I will argue in favor of taking the V1 to be the head of the phrasal structure of its CV variant even though the V2 bears the categorial information for the whole construction. The requirement that a CV will have one vector in its surface spell-out has been specified within the syntactic description of its V1 participant. Two current trends of thought on the lexicon will be incorporated in the talk. One is the body of research in lexical semantics that argues for the numerous fine-grained semantically characterized classes of predicators. The other is the technique of arranging linguistic objects using a monotonic multiple inheritance network system. In a monotonic hierarchy, constraints on supertypes affect all instances of subtypes without exception. This reduces redundancy in specifying information by cross-classifying the linguistic objects. Finally I will propose to extend the mechanism for constituting CVs of Indo-Aryan languages to account for the composition of phrasal verbs in English.

Dr. Sumit Kundu Lecturer Dept. of lectronics & Communication Engg, NIT-Durgapur, India

Dr. Kundu received his Ph.D. in 2004 from IIT Kharagpur. He received his M.Tech and B.E. from IIT-Kharagpur and REC-Durgapur in 1994 and 1991 respectively. He has been a lecturer in the Dept. of Electronics & Communication Engg. in NIT-Durgapur since 1995. His research interests are in Cellular CDMA, Radio resource management in wireless networks, Wireless LAN, QoS issues in wireless networks.

Future mobile communication systems will support wide use of multimedia services (voice, video and data) with different characteristics and requirements (BER, delay etc.). CDMA has been proposed as a promising candidate for supporting multimedia traffic in cellular system. In cellular CDMA, QoS (quality of service) is often determined by SIR (signal to interference ratio) of the radio link. The fall of SIR below a specified threshold (determined by the QoS) is termed as an outage. The cellular capacity can be estimated on the basis of outage probability. Analysis of cellular capacity has been an important issue and several researchers have studied the capacity estimation of cellular CDMA. The present work is based on my research carried out for Ph.D dissertation. The research analyzes the outage probability for assessing the over all capacity of a cellular CDMA system supporting integrated voice and data services under various conditions and studies call admission scheme for efficient utilization of radio resources. Outage probability has been analyzed considering correlation among signal and interferers in voice, data integrated system. Á^þ^üMinimum duration outageÁ^þ^û analysis has also been carried out with a two state voice model with correlated interferers in voice/data system. The analytical results have been supported by simulation. Power control, an effective way to combat the near-far problem is closely related to cellular capacity. An analytical model has been developed for evaluating the performance of a proposed power control scheme, Á^þ^ýpower limitingÁ^þ^ù which avoids the forceful termination of mobile as in Á^þ^ýpower truncation algorithmÁ^þ^ù. The outage performance of Á^þ^ýlimited power controlÁ^þ^ù scheme has been compared with that of Á^þ^ý truncation schemeÁ^þ^ù for a single cell. The performance of cellular system will depend on the efficient allocation of communication resources. Packet data transmission at high rate is becoming increasingly important in wide band CDMA. The resource allocation in DS-CDMA has been studied with imperfect power control and correlated interferers. The effects of correlation and power control error (pce) have been investigated on throughput and delay performance of data users for variable processing gain CDMA and MC (multi-code) CDMA. It has been observed that correlation amongst signal and interference increases the throughput and reduces the delay under certain range of processing gain since stronger correlation reduces the packet error rate.Call admission control (CAC) is a resource provisioning strategy to limit the number of call connections into the network to ensure satisfactory link quality of the allowed users (existing and newly admitted calls). A new channel assignment scheme has been proposed by us with different level of priorities among directed retry calls, handoff calls and new calls. A Markov model for the proposed prioritized channel assignment scheme has been developed and the performance has been evaluated in terms of blocking probability, hand-off failure and other related parameters in a micro-cellular environment such as highway cells. The above study on channel access scheme has also been extended to a general cellular environment to incorporate characteristics of the cellular scenario like distribution of base stations (BS), propagation model (path loss, shadowing), speed of mobile etc. One of the important features of CDMA is soft handoff, which reduces the interference and improves the reverse link capacity. A suitable analytical model of interference statistics has been considered with soft handoff for assessing the cellular capacity based on an outage criterion. Studies on resource allocation and admission control with soft hand off have also been carried out.

Mehul Shah HP Labs -- Palo Alto, USA

Mehul Shah graduated from MIT in 1996 with undergraduate degrees in Physics and Computer Science. He received his MEng degree from the MIT EECS department in 1997. He completed his PhD from U.C. Berkeley in 2004. At Berkeley, he was a member of the TelegraphCQ project. His dissertation work aims to provide high availability, fault tolerance, and load balancing for parallel continuous query (CQ) dataflows. His research interests include fault tolerance, CQ systems, adaptive query optimization, parallel data-intensive applications, distributed computing, indexing techniques, power-aware algorithms, and long-term digital preservation. Currently, he is a researcher at HP Labs in Palo Alto, CA.

We present techniques for robustly scaling high-throughput, 24x7, data-stream processing applications. Examples of such applications include intrusion or denial-of-service detection, click-stream processing, and online analysis of financial quote streams. As part of the TelegraphCQ project, we implement these applications using a general-purpose continuous query (CQ) engine that executes long-running dataflows. We can scale the performance of these dataflows by parallelizing them across a cluster of workstations. For these critical applications, high availability, fault tolerance, and scalability are important goals. These goals are challenging to achieve on a cluster because machines are bound to fail, and load imbalances are likely to arise. We describe the design of Flux, a reusable communication abstraction that enables long-running parallel dataflows to adapt on-the-fly to these problems. Flux encapsulates mechanisms that allow a dataflow to mask faults and to automatically recover from them as they occur during execution. These same mechanisms are also used to periodically rebalance a dataflow and keep it running efficiently. Thus, by simply constructing a parallel dataflow using Flux, an application developer can make the dataflow robust.