Статьи журнала - International Journal of Information Technology and Computer Science
Все статьи: 1195
Performance Evaluation of the Loop Buffer Switch Under Prioritized Traffic and Optical Regeneration
Статья научная
In this paper, an all-optical regenerator based, photonic packet switch architecture, which consists of the fiber loop for the storage of the contending packets, is considered. In the loop buffer, the available buffer space may not be fully utilized due to the limited re-circulation count of the data placed on buffer. This limit can be counteracted by placing a pool of regenerators inside the buffer. As optical regenerators are costly devices, hence they should be placed optimally in the buffer. The simulations results are presented by consider Prioritized and non – prioritized traffic. It is shown in the results that regeneration of data is essential if prioritized traffic has to be considered.
Бесплатно
Performance Evaluation of the Loop Buffer Switch under Prioritized Traffic
Статья научная
In this paper, optical loop buffer based architecture is discussed; with its advantages over other architectures. In general the performance of the switch is measured at physical and network layer. The physical layer analysis deals with power budget analysis, however, at the network layer; the performance is measured in terms of packet loss probability and average delay. To obtain more realistic performance at the network layer the QoS parameters need to be included. In this paper, a QoS parameter which is known as priority of the incoming packets is included and corresponding results are presented, and it has been found that even at the load of 0.7, packet loss probabilities on the order of 10-5 can be achieved.
Бесплатно
Performance Optimization in WBAN Using Hybrid BDT and SVM Classifier
Статья научная
Wireless Body Area Network has attracted significant research interest in various applications due to its self-automaton and advanced sensor technology. The most severe issue in WBAN is to sustain its Quality of Service (QoS) under the dynamic changing environment like healthcare, and patient monitoring system. Another critical issue in WBAN is heterogeneous packet handling in such resource-constrained network. In this paper, a new classifier having hybrid Binary Decision Tree and Support Vector Machine classifier is proposed to tackle these important challenges. The proposed hybrid classifier decomposes the N-class classification problem into N-1 sub-problems, each separating a pair of sub-classes. This protocol dynamically updates the priority of packet and node, adjusts data rate, packet transmission order and time, and resource distribution for the nodes based on node priority. The proposed protocol is implemented and simulated using NS-2 network simulator. The result generated for proposed approach shows that new protocol can outperform in a dynamic environment, and yields better performance by leveraging advantages of both the Binary Decision Tree in terms of efficient computation and Support Vector Machine for high classification accuracy. This hybrid classifier significantly reduces loss ratio and delay and increase packet delivery ratio and throughput.
Бесплатно
Performance analysis of LT codec architecture using different processor templates
Статья научная
Luby Transform (LT) code plays a vital role in binary erasure channel. This paper portrays the design techniques for implementation of LT codec using application specific instruction set processor (ASIP) design tools. In ASIP design, a common approach to increase the performance of processors is to boost the number of concurrent operations. Therefore, optimizations like strategy of input design, processor and compiler architecture are very useful phenomenon to enhance the performance of the application specific processor. Using Tensilica and OpenRISC processor design tools, this paper shows the response of LT codec architectures in terms of cycle counts and simulating time. Result shows that, the simulation speed of Tensilica is very high compared to the OpenRisc tool. Among different configurations of Tensilica tool, proposed ConnXD2 design took 1 M cycles per second and 135.66 ms to execute LT codec processor and XRC_D2MR configuration consumed only 9 iterations for successful decoding of LT encoded signal. Besides this, OpenRisc tool took 142K cycles and 6ms for executing LT encoder.
Бесплатно
Performance based Ranking Model for Cloud SaaS Services
Статья научная
Cloud computing systems provide virtualized resources that can be provisioned on demand basis. Enormous number of cloud providers are offering diverse number of services. The performance of these services is a critical factor for clients to determine the cloud provider that they will choose. However, determining a provider with efficient and effective services is a challenging task. There is a need for an efficient model that help clients to select the best provider based on the performance attributes and measurements. Cloud service ranking is a standard method used to perform this task. It is the process of arranging and classifying several cloud services within the cloud, then compute the relative ranking values of them based on the quality of service required by clients and the features of the cloud services. The objective of this study is to propose an enhanced performance based ranking model to help users choose the best service they need. The proposed model combines the attributes and measurements from cloud computing field and the well-defined and established software engineering field. SMICloud Toolkit has been used to test the applicability of the proposed model. The experimentation results of the proposed model were promising.
Бесплатно
Статья научная
Reusability is the quality of a piece of software, which enables it to be used again, be it partial, modified or complete. A wide range of modeling techniques have been proposed and applied for software quality predictions. Complexity and size metrics have been used to predict the number of defects in software components. Estimation of cost is important, during the process of software development. There are two main types of cost estimation approaches: algorithmic methods and non-algorithmic methods. In this work, using genetic programming which is a branch of evolutionary algorithms, a new algorithmic method is presented for software development cost estimation, using the implementation of this method; new formulas were obtained for software development cost estimation in which reusability of components is given priority. After evaluation of these formulas, the mean and standard deviation of the magnitude of relative error is better than related algorithmic methods such as COCOMO formulas.
Бесплатно
Performance of Frequency Domain DFE Using Subcarrier Allocation
Статья научная
Due to the usage of single carrier, the performance of SC-FDMA systems degrades in deep frequency selective fading channels. In this paper, we propose a structure of equalizer based on frequency domain decision feedback which could be used for multi-user SC-FDMA systems. This algorithm is applicable to various carrier allocations in multi-user systems such as localized allocation, distributed allocation, and frequency-hopping (FH) allocation. To reduce the complexity, it is not necessary to derive the inversion of matrix, which is required in the traditional decision feedback equalizer for single carrier frequency domain equalization (SC-FDE-DFE). Simulation results show that equalization has been achieved for FD-LE and FD-DFE with distribution mapping. This structure can be used in the broadcasting uplink channels with SC-FDMA scheme.
Бесплатно
Статья научная
In this paper, the performance of cellular communications based on the ambitious technology of High-Altitude Platforms (HAPs) is discussed when using tapered concentric circular arrays. The coverage cell will be described and designed with an efficient beamforming technique where the Hamming window is proposed as a tapering function and applied to the uniform concentric circular arrays (UCCA) for sidelobe reduction. Based on establishing some mapping curves, this novel tapering window is optimized in its parameters to have the lowest possible sidelobe level that can be 45 dB below the main lobe level. The optimum weights of Hamming window are found to be function of the number of elements of the innermost ring and the number of rings in the array. The cell power pattern is also discussed where the out-of-cell radiation is greatly reduced which in turns reduces the co-channel interference and improves the Carrier-to-Interference Ratio (CIR).
Бесплатно
Performance of Linear Block Coding for Multipath Fading Channel
Статья научная
This paper deals with the performance of linear block codes which provide a new paradigm for transmission over multipath fading channels. Multi path channel fading is the main enemy for any wireless communications system. Therefore, for any novel approach applied at any wireless communication system, its efficiency is measured according to its ability of mitigating the distortion caused by fading. It causes time dispersion to the transmitted symbols resulting in inter symbol interference (ISI). ISI inter symbol interference problem is a major impairment of the wireless communication channel. To mitigate the ISI problem and to have reliable communications in wireless channel, channel equalizer and channel coding technique is often employed. In this paper the BER (Bit Error Rate) performance is shown from analytically and by means of simulation for multipath dispersive channels. We have designed a channel equalizer using MLSE (Viterbi algorithm) in this paper for such a multipath channel (introducing inter symbol interferences) with BPSK modulation based on the assumption that the channel can be perfectly estimated at the receiver. Obviously the performance of channel coding in terms of BER is better than uncoded channel.
Бесплатно
Performance of Machine Learning Algorithms with Different K Values in K-fold Cross-Validation
Статья научная
The numerical value of k in a k-fold cross-validation training technique of machine learning predictive models is an essential element that impacts the model’s performance. A right choice of k results in better accuracy, while a poorly chosen value for k might affect the model’s performance. In literature, the most commonly used values of k are five (5) or ten (10), as these two values are believed to give test error rate estimates that suffer neither from extremely high bias nor very high variance. However, there is no formal rule. To the best of our knowledge, few experimental studies attempted to investigate the effect of diverse k values in training different machine learning models. This paper empirically analyses the prevalence and effect of distinct k values (3, 5, 7, 10, 15 and 20) on the validation performance of four well-known machine learning algorithms (Gradient Boosting Machine (GBM), Logistic Regression (LR), Decision Tree (DT) and K-Nearest Neighbours (KNN)). It was observed that the value of k and model validation performance differ from one machine-learning algorithm to another for the same classification task. However, our empirical suggest that k = 7 offers a slight increase in validations accuracy and area under the curve measure with lesser computational complexity than k = 10 across most MLA. We discuss in detail the study outcomes and outline some guidelines for beginners in the machine learning field in selecting the best k value and machine learning algorithm for a given task.
Бесплатно
Performance of a Slantlet Based OFDM Transceiver under Different Channel Conditions
Статья научная
A major goal of the next-generation wireless communication systems is the development of a reliable high-speed wireless communication system that supports high user mobility. Orthogonal Frequency Division Multiplexing (OFDM) system is one of the most promising technologies for current and future wireless communications that has drawn a lot of attention. OFDM usually achieved by Fast Fourier Transform (FFT). In this paper, Fast Fourier Transform (FFT) is replaced by SlantLet Transform (SLT) in order to reduce Inter-Carrier Interference (ICI), Inter-symbol Interference (ISI), and to improve the bandwidth efficiency by removing the Guard Interval (GI) needed in FFT-OFDM. The new structure was tested and compared with conventional FFT-based OFDM for Additive White Gaussian Noise (AWGN) channel, Flat Fading Channel (FFC), and multi-path Selective Fading Channel (SFC). Simulation tests were generated for different channels parameters values. The obtained results showed the proposed system has an improved Bit Error Rate (BER) performance compared with the reference system. For SFC the SLT-OFDM performs better than the FFT-OFDM on the lower SNR region, while the situation reverses with increasing SNR values.
Бесплатно
Phoenix: a framework to support transient overloads on cloud computing environments
Статья научная
This paper aims to present a computational framework capable of withstanding the effects produced by transient overloads on physical and virtual servers hosted on cloud computing environment. The proposed framework aims at automating management of virtual machines that are hosted in this environment, combining a proactive strategy, which performs load balancing when there is not overload of physical and/or virtual machines with a reactive strategy, which is triggered in the event of overload in these machines. On both strategies, it is observed the service level agreement (SLA) established for each hosted service according to the infrastructure as a service (IaaS) model. The main contribution of this paper is the implementation of a computational framework called Phoenix, capable of handling momentary overloads, considering the CPU, memory and network resources of physical and virtual machines and guaranteeing SLAs. The results demonstrate that Phoenix framework is effective, and it has outstanding performance in handling overloads virtual machine network, which has achieved the isolation of momentary overload on the physical machine preventing the propagation of their effects on the cloud.
Бесплатно
Phoneme concatenation method for myanmar speech synthesis system
Статья научная
This paper discusses the approach used to develop a Text-to-Speech (TTS) synthesis system for the Myanmar language. Concatenative method has been used to develop this TTS system using phoneme as the basic units for concatenation. In this proposed system, phoneme plays an important role so that Myanmar phoneme inventory is presented in detail. In Myanmar language, schwa is the only vowel that is allowed in a minor syllable or consonant that has half-sound of the original one. If these half sound can be handled, the TTS quality will be high. After analyzing the number of phoneme and half-sound consonant to be recorded, create the Myanmar phoneme speech database which contains total 157 phoneme speech sounds that can speech out for all Myanmar texts. These phonemes are fetched according to the result from the phonetic analysis modules and concatenated them by using proposed new phoneme concatenation algorithm. According to the experimental results, the system achieved the high level of intelligibility and acceptable level of naturalness. As the application area, it is intended for the resource limited device to use as language learning app and so on.
Бесплатно
Plagiarism detection system for the kurdish language
Статья научная
One of the serious issues is plagiarism, especially in the education field. Detecting the plagiarism became a challenging task, particularly in natural language texts. In the past years, some plagiarism detection tools have been developed for diverse natural languages, mainly English. Language-independent tools exist as well but are considered as too restrictive as they usually do not consider specific language features. The problem is there is no plagiarism Detection system for the Kurdish language. In this paper, we introduce a new system for plagiarism detection for Kurdish Language, based on n-gram algorithm, our system can detect the word, phrases, and paragraphs. Moreover, our system effectiveness for detect plagiarist texts in localhost and online especially in Google search engine. This system is more useful for the academic organizations such as schools, institutes, and universities for finding copied texts from another document.
Бесплатно
Plant disease detection system using bag of visual words
Статья научная
Plants are important to human life since plants provide the food, shelter, rain, building material, medicine, fuel such as coal, wood, etc. Therefore, planting, growing, and protecting the plants is essential for sustainable development of any nation. The plant disease can affect the growth of the plats that is caused by pathogens, living microorganisms, bacteria, fungi, nematodes, viruses, and living agents. Hence, identifying the plant disease is very essential to protect the plants in the early stage. Moreover, the plant diseases are identified from the symptoms that appear in stem, fruit, leaf, flower, root, etc. The common symptom of the plant disease can be predicted from the appearance of leaf since the appearance of leaves highly depends on the healthiness of the plant. Therefore, this paper presents a system to identify the lesion leaf from the plants in order to detect the disease occurred in the plant. This system is developed using the bag of visual words model. Moreover, the real time images are collected for various plants and tested with this system and the system produces better results for the given set of images.
Бесплатно
Platform-Independent Courseware Sharing
Статья научная
Courseware distribution between different platforms is the major issue of current e-Learning. SCORM (Sharable Content Object Reference Model) is one of the solutions for courseware sharing. However, to make SCORM-conformable courseware, some knowledge about HTML and JavaScript is required. This paper presents a SWF (Sharable Web Fragment)-based e-Learning system, where courseware is created with sharable Web fragments such as Web pages, images and other resources, and the courseware can be distributed to another platform by export and import facilities. It also demonstrates how to export a subject that contains assignments and problems and how to import the whole subject, only the assignments, or only the problems. The exported meta-information is architecture-independent and provides a model of courseware distribution.
Бесплатно
Polynomial Differential-Based Information-Theoretically Secure Verifiable Secret Sharing
Статья научная
In Pedersen’s VSS scheme the secret is embedded in commitments. And the polynomial used is of degree at most (t-1). In strong – (t, n) VSS which based on Pedersen’s scheme that polynomial in verification purpose is public polynomial. The public polynomial in their scheme which acts in verification purpose is not secure. And the secret is secure if the dealer cannot solve the discrete logarithm problem. In our propose scheme we will satisfy the security requirements in strong t-consistency and consider the security on verification polynomial used. We will show in shares verification algorithm the participants can verify that their shares are consistent and the dealer is honest (i.e. the dealer cannot success in distributing incorrect shares even the dealer can solve the discrete logarithm problem.) before start secret reconstruction algorithm. The security strength of the proposed scheme lies in the fact that the shares and all the broadcasted information convey no information about the secret.
Бесплатно
Статья научная
The recommendation is playing an essential part in our lives. Precise recommendations facilitate users to swiftly locate desirable items without being inundated by irrelevant information. In the last few years, the amount of customers, products and online information has raised speedily and results out into the huge data analysis problem for recommender systems. While handling and evaluating such large-scale data, usual service recommender systems regularly undergo scalability and inefficiency problems. Nowadays, in multimedia platform such as movie, music, games, the use of Recommender System is increased. Collaborative Filtering is a dominant filtering technique used by many RSs. CF utilizes the rating history of the user to find out "like minded" users and this set of like-minded user is then used to recommend the movies which are liked by these like-minded users but did not watch by the active user. Thus, in CF, to find out the "neighborhood" the rating history of a user is used, but the reason behind the rating is not considered at all. This will lead to inaccuracy in finding a neighborhood set and subsequently in recommendation also. To cope with these scalability and accuracy challenges, this paper proposes an innovative solution, Clustering and Review based Approach for Collaborative Filtering based Recommendation. This innovative approach is enacted with the two stages; in the first stage the clustering of the available movies for recommendation is clustered into the subclasses for further computation. In the succeeding stage, the methodology based on reviews is utilized for finding neighborhood set in User Based Collaborative Filtering.
Бесплатно
Prediction Model of the Stock Market Index Using Twitter Sentiment Analysis
Статья научная
Stock market prediction has been an interesting research topic for many years. Finding an efficient and effective means of predicting the stock market found its way in different social networking platforms such as Twitter. Studies have shown that public moods and sentiments can affect one's opinion. This study explored the tweets of the Filipino public and its possible effects on the movement of the closing Index of the Philippine Stock Exchange. Sentiment Analysis was used in processing individual tweets and determining its polarity - either positive or negative. Tweets were given a positive and negative probability scores depending on the features that matched the trained classifier. Granger causality testing identified whether or not the past values of the Twitter time series were useful in predicting the future price of the PSE Index. Two prediction models were created based on the p-values and regression algorithms. The results suggested that the tweets collected using geo location and local news sources proved to be causative of the future values of the Philippine Stock Exchange closing Index.
Бесплатно
Prediction Models for Diabetes Mellitus Incidence
Статья научная
Diabetes mellitus is an incurable disease with global prevalence and exponentially increasing incidence. It is one of the greatest health hazards of the twenty-first century which poses a great economic threat on many nations. The premise behind effective disease management in healthcare system is to ensure coordinated intervention targeted towards reducing the incidence of such disease. This paper presents an approach to reducing the incidence of diabetes by predicting the risk of diabetes in patients. Diabetes mellitus risk prediction model was developed using supervised machine learning algorithms of Naïve Bayes, Support Vector Machine and J48 Decision Tree. The decision tree was able to give a prediction accuracy of 95.09% using rules of prediction that give acceptable results, that is, the model was approximately 95% accurate. The easy-to-understand rules of prediction got from J48 decision tree make it excellent in developing predictive models.
Бесплатно
