Md. Rafiqul Islam

Collaborative swarm intelligence based Trusted Computing

Security is one of major barriers for cloud computing popularity. For a secured system, along with protection of data storage it is also important to take rapid action to prevent evolving attacks and maintain a self-healing infrastructure. In this paper a security framework has been proposed, where cognitive and cooperative swarm intelligences are used in order to implement reputation system for establishing self-defensive secured multi-layered cloud systems. To implement the reputation system we have devised algorithms to provide better security solution. The uniqueness of our proposed security framework is its cognitive and self-healing nature with the scope of handling data leakage and data availability issue.

An approach to provide security to unstructured Big Data

Security of Big Data is a big concern. In broad sense Big Data contains two types of data such as structured and unstructured. To provide security to unstructured data is more difficult than that of structured. In this paper we have developed an approach to give adequate security to the unstructured data by considering the types of the data and their sensitivity levels. We have reviewed the different analytics methods of Big Data, which gives us the facility to build a data node of databases of different types of data. Each type of data has been further classified to provide adequate security and enhance the overhead of the security system. To provide security to data node a security suite has been designed by incorporating different security standards and algorithms. The proper security standards or algorithms can be activated using an algorithm, which has been interfaced with the data node. We have shown that data classification with respect to sensitivity levels enhance the performance of the system.

Data Intensive Dynamic Scheduling Model and Algorithm for Cloud Computing Security

As cloud is growing immensely, different types of data are getting more and more dynamic in terms of security. Ensuring high level of security for all data in storages is highly expensive and time consuming. Unsecured services on data are also becoming vulnerable for malicious threats and data leakage. The main reason for this is that, the traditional scheduling algorithms for executing different services on data stored in cloud usually sacrifice security privilege in order to achieve deadline. To provide adequate security without sacrificing cost and deadline for real time data- intensive cloud system, security aware scheduling algorithm has become an effective and important feature. Existing systems also merely provide efficient security aware scheduling and security for data. In order to ensure adequate security for different data storages in cloud, in this paper we have proposed a three tier security framework. We have analyzed mathematically the security overhead for different security services such as confidentiality, integrity as well as authenticity and shown that our framework can provide adequate level of security and enhance the processing speed of security services without taking additional overhead in time. We have also proposed a scheduling algorithm to ensure security for data intensive applications. The simulation results show that the proposed algorithm gives better success rate with adequate security in comparisons with existing algorithms. This algorithm ensures security of data and applications as well as confirms the job to be scheduled within deadline.

Dynamic scheduling approach for data-intensive cloud environment

The scheduling problem domain in cloud environment recently has been extended to deal with two new phenomenon such as data-intensive and security constraints in cloud environment. However traditional scheduling approaches have been failed to deal with these new addition. In this paper, the system architecture along with security constraint model for data-intensive cloud environment is designed. Moreover, a novel security constraints scheduling approach to schedule all jobs in cloud environment efficiently without compromising required security level for each job is presented in this paper. In the regard of cloud security, swarm intelligence is highly capable to provide better solutions for such potentially intractable problems. Therefore, an Ant Colony Optimization based scheduling algorithm is proposed in this paper. Several meta-heuristic mathematical models as well as explanations have been introduced to deal with effective security constraint scheduling strategy. Meanwhile, the experimental results shows that the overall performance of proposed scheduling algorithm is better than other existing scheduling algorithms on four basic measurements: the optimization rate of throughput, cost, CPU time and security constraints.

An Enhanced Short Text Compression Scheme for Smart Devices

Short Text Compression is a great concern for data engineering and management. The rapid use of small devices especially, mobile phones and wireless sensors have turned short text compression into a demand-of-thetime. In this paper, we propose an approach of compressing short English text for smart devices. The prime objective of this proposed technique is to establish a low-complexity lossless compression scheme suitable for smart devices like cellular phones and PDAs (Personal Digital Assistants) having small memory and relatively low processing speed. The main target is to compress short messages up to an optimal level, which requires optimal space, consumes less time and low overhead. Here a new static-statistical context model has been proposed to obtain the compression. We use character masking with space integration, syllable based dictionary matching and static coding in hierarchical steps to achieve low complexity lossless compression of short English text for low-powered electronic devices. We also propose an efficient probabilistic distribution based content-ranking scheme for training the statistical model. We analyze the performance of the proposed scheme as well as the other similar existing schemes with respect to compression ratio, computational complexity and compression-decompression time. The analysis shows that, the required number of operations for the proposed scheme is less than that of other existing systems. The experimental results of the implemented model give better compression for small text files using optimum resources. The obtained compression ratio indicates a satisfactory performance in terms of compression parameters including better compression ratio, lower compression and decompression time with reduced memory requirements and lower complexity. The compression time is also lower because of computational simplicity. In overall analysis, the simplicity of computational requirement encompasses the compression effective and efficient.

A secure communication suite for cluster-based Underwater Surveillance Networks

Surveilling the water territory of a country by installing an Underwater Surveillance Network (USN) is one of the prominent applications developed for underwater communication. The primary responsibility of a USN is to detect trespassers within the premises and to notify the activities to the responsible authority for further actions. Alike any surveillance network, a USN also possesses several security threats and may undergo several passive and/or active attacks. In this paper, we propose a noble secure communication suite which is resistible against most kind of known attacks. Our propose protocol aimed at gaining confidentiality, integrity and authenticity of the messages being exchanged within the cluster while taking into the consideration of the peculiarities of the underwater communication. Moreover, since our proposed scheme is designed for a cluster-based USN where routing layer activities are not eminent, it is embedded within a media access control (MAC) protocol.

Design and Analysis of an Effective Corpus for Evaluation of Bengali Text Compression Schemes

In this paper, we propose an effective platform for evaluation of Bengali text compression schemes. A novel scheme for construction of Bengali text compression corpus has also been incorporated in this paper. A methodical study on the formulation-approaches of text corpus for data compression and present an effective corpus named Ekushe-Khul for evaluating the Bengali text compression schemes has also been presented in this paper. To design the Bengali text compression corpus, Type to Token Ratio has been considered as the selection criteria with a number of secondary considerations. This paper also presents a mathematical analysis on data compression performance with structural aspects of corpora. A comprehensive analysis on the evolving criteria of text compression corpora with related issues in designing dictionary based compression are extensively incorporated here. The proposed corpus is effective for evaluating compression efficiency of small and middle sized Bengali text files.

Absorption rate analysis of cellular phone radiation on the human head

In this paper the presented research work analyzes the effects of electromagnetic (EM) radiation of mobile phone on human head. The EM radiation is measured based on Specific Absorption Rate (SAR). The human head is generally exposed to mobile phones operating at communication (GSM, CDMA etc) frequency bands. The radiation absorption analyzed through simulations by applying frequency domain using Comsole Multiphysics software. The specific absorption rate (SAR) was measured for different positions of mobile phone. SARs exhibited in much lower values as the mobile phone held in far position from human brain. This measurement demonstrates SAR is high at close distance; effectively the temperature increase is also high as the distance reduces. Comparative analysis also shows that FR4 substrate demonstrates higher SAR and temp raise compared to Rogers RO3006 and Rogers RO4003 substrates.

A unified architecture for utility service in cloud

Today our world is ready to make the maximum use of cloud computing facility as like other daily most common utilities such as electricity, gas and water. However, currently the cloud is showing its limited effect in case of utility service system. The main lacking in cloud is the absence of unified architecture for deploying service and use of application in an optimum way to make the maximum availability of service within the limited time frame duration and minimum configuration overhead. In this research a cloud based unified architecture has been proposed for utility service system. In addition, the working procedure of proposed architecture with performance description is also added in this paper.

Adaptive Instantaneous Traffic Signal Management Through Cascade Object Detection

Road traffic controlling is one of the prime issues of any busy town in the world. The situation is worst in the junction point of multiple roads. It is not always possible to predict upcoming traffic pressure intensity and direction with prior information. An adaptive approach is needed to solve this problem in real time. In this paper an efficient method of controlling traffic signal duration is proposed. Here cascade object detection technique is used to recognize and count the number of moving vehicles towards a junction from various directions through static video camera installed in each direction. Using the static camera, the proposed system senses the current traffic pressure of each connecting road towards the junction and thereby takes sensible decision of setting traffic signal duration in each road.