Venue : Coral Deira - Dubai, Deira, Dubai, UAE.  &  Date : May 18~19, 2013

Accepted papers

  • TCP-WPAL: Congestion Control Mechanisms of TCP using Cross Layer Interaction in Multihop Wireless Networks
    G. Sankara Malliga1 and Dharmishtan K Varughese2, 1Anna Univcersity, India and 2Karpagam College of Engineering, India.
    The objective of the Transmission Control Protocol (TCP) is to provide reliable end-to-end delivery of data over unreliable networks. Over the past few years, the problem of congestion control has received wide-spread attention in multihop wireless networks. In these networks, however packet losses occur more often due to unreliable links than to congestion. But each packet loss on the wireless link results in invoking congestion control measures at the source, which leads to severe performance degradation of the TCP. Many authors have reported that the TCP interacts with the lower layers, but still it cannot predict route failures and network congestion. The proposed work describes the efficient techniques in various layers to improve the TCP performance over multihop wireless networks. In the data link layer, the link RED (Random Early Discard), LRED technique seeks to react earlier to link overload, and it solves the hidden terminal problem. In the link layer, the adaptive pacing (AP) technique seeks to improve spatial reuse, and it solves the exposed terminal problem. In the Physical and MAC layers, the WPAN physical layer (IEEE 802.15.4 PHY) and MAC enhancement (IEEE 802.15.4 MAC) has employed the TCP to interact with the network layer and application layer. The window size and PDR of the TCP-WPAL are increased, and the delay and jitter of the TCP-WPAL are better in all the topologies of multihop wireless networks. Both in analytical and in simulation, the similar results are achieved.
  • Hamming Distance and Data Compression of 1-D CA
    Raied Salman, Stratford University, United States.
    In this paper an application of von Neumann correction technique to the output string of some chaotic rules of 1-D Cellular Automata that are unsuitable for cryptographic pseudo random number generation due to their non uniform distribution of the binary elements is presented. The one dimensional (1-D) Cellular Automata (CA) Rule space will be classified by the time run of Hamming Distance (HD). This has the advantage of determining the rules that have short cycle lengths and therefore deemed to be unsuitable for cryptographic pseudo random number generation. The data collected from evolution of chaotic rules that have long cycles are subjected to the original von Neumann density correction scheme as well as a new generalized scheme presented in this paper and tested for statistical testing fitness using Diehard battery of tests. Results show that significant improvement in the statistical tests are obtained when the output of a balanced chaotic rule are mutually exclusive ORed with the output of unbalanced chaotic rule that have undergone von Neumann density correction.
  • Efficient Data Structure Based Smart Card Implementation
    Shalini Jain1, Anupam Shukla1, Bishwajeet Pandey2 and Mayank Kumar2, 1Indian Institute of Information Technology, Gwalior and 2VB Taneja Embedded System Design R&D Lab India
    To make smart card much faster, we need efficient data structure. Access time of on chip data depends on how and where we stored. Some Data Structure take maximum time and some take minimum time depending on the space and time complexity of data structure. In this work, we have taken some data structures and find that BST is the best suitable data structure for performing smart card operations in compare to other possible data structures.
  • A Novel Prior-Based Real-Time Click Through Rate Prediction Model
    The application of traditional machine learning algorithms on click data has met great challenges working with severe sparse and transient ID features, which tend to bloat data size and prolong training time considerably. On the other hand, due to the data amount requirement, training and publishing overhead, the bottleneck of minute-level incremental model updates has emerged. We propose a novel real-time Click Through Rate (CTR) prediction model based on empirical CTRs with a set of pre-learned priors, upon which a Minimum Variance Unbiased Estimator is constructed as the CTR prediction. The dimensions of the empirical CTRs are in the sparsest and nest ID levels, which can be strong indicators but are generally unsuited as machine learning features. Experiments on real-life click data show that our prior-based real-time estimator (PRE), combined with traditional machine learning model, gains signi cant improvement in both prediction accuracy and ranking capability, especially with latest data beyond the timeeffectiveness of the machine learning model.
  • Feedback Shift Registers as Cellular Automata Boundary Conditions
    K. Salman, Middle Tennessee State University Murfreesboro, USA.
    We present a new design for random number generation. The outputs of linear feedback shift registers (LFSRs) act as continuous inputs to the two boundaries of a one-dimensional (1-D) Elementary Cellular Automata (ECA). The results show superior randomness features and the output string has passed the Diehard statistical battery of tests. The design is good candidate for parallel random number generation, has strong correlation immunity and it is inherently amenable for VLSI implementation.
  • IO Standard Based Green Multiplexer Design and Implementation on FPGA
    Bishwajeet Pandey, Deepa Singh and Anupam Shukla, Indian Institute of Information Technology, India
    In this work, we are using Stub Series Transistor Logic(SSTL) on the simplest VLSI circuit multiplexer and analyze the power dissipation with different class. Using SSTL15 in place of SSTL2_II_DCI, there is reduction of 304mW power i.e. 76.19% power reduction. Using HSTL_I_12 in place of HSTL_III_DCI_18, there is reduction of 157mW power i.e. 62.3% power reduction. HSTL and SSTL are IO standards taken under consideration. SSTL minimum power consumption is almost same as HSTL. But, the power dissipation of SSTL is 58.73% higher than HSTL, when we consider maximum power dissipation of both. Virtex-6 is an FPGA on which we implement this low power design. Xilinx ISE 14.1 is an ISE tool to design and synthesize multiplexer.
  • Intrusion Detection System based on Fuzzy C Means Clustering and Probabilistic Neural Network
    Ms. Rachna kulhare and Divakar Singh, BUIT, India.
    Security is always an important issue especially in the case of computer network which is used to transfer personal/confidential information's, ecommerce and media sharing. Since the network is closely related to operating its conditions hence a careful observation & analysis of network characteristics could describe the state of the network such as network is under specific attack or operating normally. This paper presents an intrusion detection system based on fuzzy C-means clustering and probabilistic neural network which not only reduces the training time but also increases the detection accuracy. The proposed system is tested using KDD99 dataset and the simulation results shows that by selecting effective characteristics and proper training the detection rate up to 99% is achievable.
  • Enhanced Discrete Particle Swarm Optimization Using Evolutionary Operators
    J. R. Ayala-Solares and H. Ponce, Solarium Research Labs, Mexico.
    It is known that algorithms for combinatorial optimization in continuous search space have to be adequated for discrete search space, as the continuous particle swarm optimization (PSO) to discrete particle swarm optimization (DPSO). However, DPSO has some drawbacks, especially in the calculation of velocities because particles tend to go far away of the global optimum. In this paper we introduce an enhancement to DPSO using inspiration in evolutionary operators, i.e. mutation. Results over the traveling salesman problem show that the proposed DPSO reach the global optimum rapidly even if the search space is large enough.
  • Proposed approach for Intrusion Detection based on Fuzzy C-Means Clustering and Probabilistic neural network
    Ms. Rachna kulhare and Divakar Singh, BUIT, India
    Network security has been one of the most important problems in Computer Network Management and Intrusion is the most publicized threats to security. Intrusion detection is a device or software application that monitors network or system has emerged as an important field for network security. Fuzzy C-means clustering & probabilistic neural network technique have been organized as a new approach for intrusion detection. The proposed approach combines the fuzzy C-means clustering with the probabilistic neural network technique.(FCM& PNN)This proposed approach analyze the KDD99 dataset and it is efficient in terms of accuracy, detection rate, failure analysis rate and false alarm.
  • An Energy-Aware Scheduling Algorithm for Precedence-constrained Applications in Virtualized Datacenters
    Vahid Ebrahimirad and Maziar Goudarzi, Sharif University of Technology Tehran, Iran.
    During the last decade, the main goal of designing and building data centers have been concentrated on reduction of execution time and enhancement of performance, so data centers were not energy-aware. Thus, the power consummation of computers has become a major concern for both industry and society in recent years. As a result, researchers have been developing some techniques at the hardware and software levels. At the software level, one of the important and effective techniques that have been used is task scheduling for saving energy in data centers. In this paper, we address the problem of scheduling precedence-constrained parallel Virtual Machines (VMs) on homogeneous distributed Physical Machines (PMs) and present energy-aware scheduling algorithm VMs consolidation and using dynamic voltage frequency scaling (DVFS). Our scheduling algorithm extends and optimizes a known algorithm that called ECS. The goal of our scheduling algorithm is enhancement of makespan and reduction the energy consumption by maximizing utilization in PMs. The extensive comparative evaluations conducted as a part of this work show that the performance of our algorithms is very compelling in terms of both application completion time and energy consumption.
  • Computational Performance of Quantum Phase Estimation Algorithm
    Zhuang Jiayu and Zhao Junsuo, Institute of software Chinese academy of sciences, China.
    A quantum computation problem is discussed in this paper. Many new features that make quantum computation superior to classical computation can be attributed to quantum coherence effect, which depends on the phase of quantum coherent state. Quantum Fourier transform algorithm, the most commonly used algorithm, is introduced. And one of its most important applications, phase estimation of quantum state based on quantum Fourier transform, is presented in details. The flow of phase estimation algorithm and the quantum circuit model are shown. And the error of the output phase value, as well as the probability of measurement, is analysed. The probability distribution of the measuring result of phase value is presented and the computational efficiency is discussed.