Asian Journal of Engineering, Sciences & Technology
Abbreviated key title: AJEST
Published by: Iqra University
Months: March & September
Start year: 2011
Title: A Comprehensive Performance Study of Classification of Emotions using EEG Centroid Data Author(s): Muhammad Umair, Syed Sajjad Hussain Rizvi, Kamran Raza
Abstract: Classification of human emotion using EEG signal is an interesting research topic for many. In this paper a detailed study of classification efficiency of different machine learning algorithms using EEG signal was done. The study make use of DEAP data set. The Valance, Arousal and Dominance values reported in DEAP data set were ranked on Mehrabian 3D emotion classification model to produce 8 emotions. k-mean was used to reduce number of observations in original data set and 3, 10 and 40 centroids were calculated for EEG data. 11 different combinations of emotions were tested using 20 different variations of Decision Trees, SVM, kNN and Ensemble classifiers. This study is one of its kind which uses k-mean as number of observation reducing technique and rank top performing algorithm for each set of emotion..
Title: A Novel Approach to Classify Human Emotions Based on Electroencephalogram (EEG) Signals Author(s): Syed Sajjad Hussain Rizvi, Jawwad Ahmed, Muhammad Zubair, Salman Zaidi
Abstract: An extensive growth in EEG based systems have been observed in the recent decade. This includes treatment and support of mentally challenged persons, gaming, marketing, development of brain controlled devices etc. The performance of all these applications are primarily based on precise classification of human emotions. In this paper, an effective fractional least mean square based emotion classifier is proposed. The rigorous training and testing of the proposed classifier was performed on a benchmark video stimuli based EEG signal dataset. Moreover, the performance of the proposed classifier is also compared with the classical and most employed approach. Results prove that the proposed classifier for human emotions is far better than the most employed classification approach.
Title: Demonstration of Self Activating Failure Recovery in SDN Author(s): Muhammad Faizan, Maheen Iqbal, Waleed Aftab, Muhammad Irfan Anis
Abstract: In this paper, we propose and experimentally evaluate a novel automatic network failure recovery mechanism based on Software Defined Networking (SDN). We demonstrate the introduction of services in this network with a reserve controller. The implemented architecture is setup to enable an efficient failure recovery mechanism in a way, in which reserve controller executes only in case of a main controller failure. Upon failure, the main controller automatically shifts control to the reserve controller for performing the computations of server load balancing and automatic bandwidth allocation until the main controller failure gets resolved. Experimented results show the estimated failure recovery time up to 24 ms.
Title: Provision of Fault Tolerance Using Multiple Controllers in Software-Defined Networks Author(s): Abdul Khaliq, Fareeha Ghafoor, Aqsa Saleem, Junaid Jamshid, Kamran Raza
Abstract: Software defined networks decouples or separate control and data planes from each other while providing great amount of flexibility to the network administrators to program and manage the network and to make it functional as needed or required. High availability of controllers or distributed controllers is needed to provide fault tolerance in an environment where delays or losses in the delivery of packets cannot be tolerated. In this paper we conducted experiments of multiple controllers which provide fault tolerance in such a way that each controller has full state of information of the network topology. The throughput results for a data center network connected to single controller and then connected to multiple controllers were observed and we found that only 2.82 % more data transfer was observed in single controller environment as compared to multiple controller environment.
Title: Low Side Lobe Feed Network for Automotive Application Author(s): Junaid Jamshid, Omar Masood Khan, Kamran Raza
Abstract: This paper presents a simple feed network design approach for reducing the side lobe level. To prove the results, a micro-strip array antenna is formed with electrically-thin substrate over a single layer to give a cost-effective answer for automotive applications. The substrate is RT-Droid and have "r = 2:2 dielectric constant. Array feed network is design with an equal power divider for four element array and extend it to 8 elements array. Results demonstrated that this simple design lowers the side lobe level more then -22dB to -24.5dB depend upon the number of array element. The simulation is done on CST microwave studio software.
Title: Embodiment of Load Balancing and Path Loss IP Networks in Software Defined Networking Author(s): Muhammad Adeel, Muhammad Irfan Anis
Abstract: The excessive growth of internet traffic and broadband services has initiated the requirement for a bandwidth flexible network making services more capable of elasticity. The issue of network management can be brought to simplification with the emergence of Software Defined Networks (SDN) through the introduction of a programmable network. The major problems such as scalability and availability has been addressed which may lead to a cause of failure in load balancing. In order to ensure dynamic load balancing within a SDN, the technique of IP load balancing and provision of a congestion free network has been employed through comparison between random and round robin approach. The severity of path loss has also been addressed to convey a simple, essential and add-on network paradigm within the SDN architecture. Moreover, 100 and 500 packets have been transmitted in the described load balancing scenarios where packet throughput is the main concern, and 15% of the link losses have been observed in path loss. The results represent profound traffic steering capabilities bringing up an efficient network performance.
Title: Queuing Mechanisms to improve QoS for VOIP over IPv4 and MPLS networks Author(s): Sumera Saleem, Mansoor Ahmed Tahir
Abstract: In this paper, an implementation of different Queuing mechanisms over MPLS network for QoS of VOIP. Voice over Internet Protocol is a technology for delivery of telephone calls and other voice communication over the internet, involving the conversion of analog voice signals to digital form. To improve QoS of VOIP; different queuing mechanisms (FIFO, WFQ, CBWFQ, and LLQ) are used over MPLS network. Lan Traffic V2 (software testing tool) used to generate traffic over an MPLS network. The results gathered from the analysis on Wireshark shows that queuing mechanisms (FIFO, WFQ, CBWFQ, and LLQ) have a significant impact on QoS of VOIP and the impact varies according to the parameters (delay and packet loss). CBWFQ and LLQ queuing mechanisms are most appropriate to improve QoS of VOIP.
Title: Multi-Band Printed Antenna for GPS /LTE/GSM Mobile Handset Applications Author(s): Syed Sharjeel Tirmizi, Omar Masood Khan, Kamran Raza
Abstract: This paper presents printed antenna with multiband characteristics for mobile handset applications. The antenna operates for the frequency bands of 700 MHz, 850 MHz, 900 MHz, 1575 MHz, 1800 MHz, 1900 MHz, 2100 MHz, 2400 MHz, 2500 MHz and 2700 MHz for GPS, GSM, DCS, PCS and LTE applications with S11 = - 6 dB matching criteria (VSWR 3:1). The Antenna consists of driven monopole with branch lines. By properly controlling the dimensions of each slot and length of elements, desired frequencies are achieved with good radiation characteristics. The desired operating bands are achieved in a compact area of 15 × 60 mm. The results for the surface currents, s-parameters and 3D gain plots are illustrated for the proof of concept.
Freely Accesible Scientific Literature. Higher Impact. Global Reach.
Find out more
The submissions will be peerreviewed by two independent reviewers.
Faculty of Engineering, Sciences and Technology, IQRA University, Defence View, Shaheed-e-Millat Road, Karachi-75500
Phone: + 9221 111 264 264
Fax: + 9221 35894806