Wednesday, September 19, 2018
   
Text Size
Login

Journal of Information and Communication Technology (JICT) Vol. 17, No. 1, January 2018

 
 
AN IMPROVED ARTIFICIAL DENDRITE CELL ALGORITHM FOR ABNORMAL SIGNAL DETECTION
Mohamad Farhan Mohamad Mohsin, Azuraliza Abu Bakar, Abdul Razak Hamdan & Mohd Helmy Abdul Wahab
 
 
 
 
 

 
TOWARDS A MATHEMATICAL LIBRARY FOR UNUMs, AN ALTERNATIVE TO IEEE 754 FLOATING POINT NUMBERS
Thomas Risse
Institute of Informatics and Automation, Faculty EEE & CS City
University of Applied Sciences
Bremen, Germany
risse@hs-bremen.de

ABSTRACT | FULL TEXT
The 1985 IEEE 754 standard for the representation of and the arithmetic with floating point numbers has been reconsidered. On the one hand today, its technological assumptions are by no means longer valid. On the other hand, the irritating numerical phenomena which have been collected cast a doubt as to whether this much uncertainty in numerical results is fate. Fortunately, around 2015, Gustafson proposed UNUMs, a modification of the IEEE 754 standard with the potential to heal the said shortcomings. Till now, there are some attempts to implement his ideas, both in software and in hardware. With these activities well under way, the other necessity is development of a mathematical library for UNUMs when one wants UNUMs to become the new floating point standard. This paper presented the ideas leading to UNUMs, gave some hints on floating point units for UNUMs and illustrated the difficulties in developing the said mathematical library by the example of approximating zeroes of analytic functions.

Keywords: Floating point numbers, IEEE 754, UNUMs, arithmetic, mathematical library.
 
Received: 25 June 2017, Accepted 3 August 2017
 

 
IMPLEMENTATION ANALYSIS OF CUCKOO SEARCH FOR THE BENCHMARK ROSENBROCK AND LEVY TEST FUNCTIONS
Julius Beneoluchi Odili
Faculty of Computer Systems and Software Engineering
Universiti Malaysia Pahang, Malaysia
odili_julest@yahoo.com

ABSTRACT | FULL TEXT
This paper presents the implementation analysis of the benchmark Rosenbrock and Levy test functions using the Cuckoo Search with emphasis on the effect of the search population and iterations count in the algorithm’s search processes. After many experimental procedures, this study revealed that deploying a population of 10 nests is sufficient to obtain acceptable solutions to the Rosenbrock and Levy test functions (or any similar problem to these test landscapes). In fact, increasing the search population to 25 or more nests was a demerit to the Cuckoo Search as it resulted in increased processing overhead without any improvement in processing outcomes. In terms of the iteration count, it was discovered that the Cuckoo Search could obtain satisfactory results in as little as 100 iterations. The outcome of this study is beneficial to the research community as it helps in facilitating the choice of parameters whenever one is confronted with similar problems.

Keywords: Cuckoo search, iteration, Levy function, population, Rosenbrock function.
 
Received: 6 June 2017, Accepted:24 July 2017
 

 
AN IMPROVED ARTIFICIAL DENDRITE CELL ALGORITHM FOR ABNORMAL SIGNAL DETECTION
1Mohamad Farhan Mohamad Mohsin, 2Azuraliza Abu Bakar, 3Abdul Razak Hamdan & 4Mohd Helmy Abdul Wahab
1School of Computing, Universiti Utara Malaysia, Malaysia
2,3Faculty of Science & Information Technology
Universiti Kebangsaan Malaysia, Malaysia,
4Faculty of Electrical and Electronic Engineering
Universiti Tun Hussein Onn Malaysia, Malaysia
farhan@uum.edu.my; aab@ukm.my; arh@ftsm.ukm.my; helmy@uthm.edu.my
 
ABSTRACT | FULL TEXT
In dendrite cell algorithm (DCA), the abnormality of a data point is determined by comparing the multi-context antigen value (MCAV) with anomaly threshold. The limitation of the existing threshold is that the value needs to be determined before mining based on previous information and the existing MCAV is inefficient when exposed to extreme values. This causes the DCA fails to detect new data points if the pattern has distinct behavior from previous information and affects detection accuracy. This paper proposed an improved anomaly threshold solution for DCA using the statistical cumulative sum (CUSUM) with the aim to improve its detection capability. In the proposed approach, the MCAV were normalized with upper CUSUM and the new anomaly threshold was calculated during run time by considering the acceptance value and min MCAV. From the experiments towards 12 benchmark and two outbreak datasets, the improved DCA is proven to have a better detection result than its previous version in terms of sensitivity, specificity, false detection rate and accuracy.

Keywords: Anomaly threshold, dendrite cell algorithm, multi-context antigen value.
 
Received: 20 February 2017, Accepted: 30 July 2017
 

 
ENHANCED HEURISTIC ALGORITHMS WITH A VEHICLE TRAVEL SPEED MODEL FOR TIME-DEPENDENT VEHICLE ROUTING: A WASTE COLLECTION PROBLEM
Nur Azriati Mat, Aida Mauziah Benjamin & Syariza Abdul-Rahman
Institute of Strategic Industrial Decision Modelling, School of Quantitative Sciences, Universiti Utara Malaysia, Malaysia
nur_azriati@ahsgs.uum.edu.my; mauziah@uum.edu.my; syariza@uum.edu.my

ABSTRACT | FULL TEXT
This paper proposes a vehicle travel speed model to enhance two heuristic algoritihms from previous studies, namely current initial solution (CIS) and different initial customer (DIC). Both algorithms are used to solve a real-life waste collection vehicle routing benchmark problem with dynamic travel speeds. This problem is referred to as Time-Dependent Vehicle Routing Problem (TD-VRP) in previous literature. The benchmark problem consisted of ten sub problems, involving up to 2092 customers. Previous studies solved the benchmark problem using DIC and CIS algorithms with the assumption that the vehicles are travelling with a static speed when collecting the waste. However, in this paper the static speed that was considered in both algorithms were improved by introducing dynamic travel speeds to construct vehicle routes for the waste collection drivers. Compared to previous studies the enhanced CIS and DIC with dynamic travel speeds affected the waste collection problem in terms of the number of vehicles used, the total distance travelled and the total travel time. However, different settings of speed may give different impacts to the solution. The study reveals that with a setting of dynamic speed between 40 mph and 55 mph, DIC is able to reduce two vehicles (from 98 to 96 number of vehicles used), 7.85% of total distance travelled, and 19.10% of total travel time.

Keywords: Time-dependent vehicle routing, dynamic travel speed, waste collection, heuristics.
 
Received: 23 July 2017 Accepted: 19 November 2017
 

 
FRAMEWORK FOR MODELLING MOBILE NETWORK QUALITY OF EXPERIENCE THROUGH BIG DATA ANALYTICS APPROACH
1Ayisat Wuraola Yusuf-Asaju, 2Zulkhairi Md Dahalin & 2Azman Ta’a
1,2Department of Computer Science, University of Ilorin, Nigeria
2School of Computing, Universiti Utara Malaysia, Malaysia
ayisatwuraola@gmail.com; zul@uum.edu.my; azman@uum.edu.my

ABSTRACT | FULL TEXT
The increase in the usage of different mobile internet applications can cause deterioration in the mobile network performance. Such deterioration often declines the performance of the mobile network services that can influence the mobile Internet user’s experience, which can make the internet users switch between different mobile network operators to get good user experience. In this case, the success of mobile network operators primarily depends on the ability to ensure good quality of experience (QoE), which is a measure of users’ perceived quality of mobile Internet service. Traditionally, QoE is usually examined in laboratory experiments to enable a fixed contextual factor among the participants even though the results derived from these laboratory experiments presented an estimated mean opinion score representing perceived QoE. The use of user experience dataset involving time and location gathered from the mobile network traffic for modelling perceived QoE is still limited in the literature. The mobile Internet user experience dataset involving the time and location constituted in the mobile network can be used by the mobile network operators to make data-driven decisions to deal with disruptions observed in the network performance and provide an optimal solution based on the insights derived from the user experience data. Therefore, this paper proposed a framework for modelling mobile network QoE using the big data analytics approach. The proposed framework describes the process of estimating or predicting perceived QoE based on the datasets obtained or gathered from the mobile network to enable the mobile network operators effectively to manage the network performance and provide the users a satisfactory mobile Internet QoE.
 
Keywords: Big data analytics, mean opinion score; mobile network operators, telecommunication, users experience.
 
Received: 19 June 2017, Accepted: 19 November 2017
 

 
LEVEL OF RESILIENCE MEASURE FOR COMMUNICATION NETWORKS
Mariam Wajdi Ibrahim
School of Applied Technical Sciences
German Jordanian University, Jordan
mariam.wajdi@gju.edu.jo
 
ABSTRACT | FULL TEXT
Our daily life applications have come to depend on communication networks to deliver services in an efficient manner. This has made it possible for an attacker to sabotage its operation. Network resiliency is concerned with the degree the network is able to bounce back to a normal operation in the face of attacks. This paper introduced a new resiliency measure, called Levelof-Resilience (LoR) for communication networks, determined by examining: (a) the Level-of-Stability-Reduction (LoSR), as measured by percentage of “IP traffic dropped”, (b) the eventual Level-of-Performance-Reduction (LoPR), as captured by the percentage of reduction in the application Quality-of-Service (QoS), namely latency and (c) Recovery-Time (RT), which is the time the network takes to detect and recover from an attack or a fault, as measured by convergence duration. Previous resiliency measures may only consider one aspect of the above parameters, while this measure is a composite of them. This paper showed that network topology can affect the network resilience, as indicated by the LoR metric. This measure is illustrated by comparing the resiliency level of two communication networks that served the same traffic, but differed in their network topology, under three different attack scenarios.
 
Keywords: Level-of-resilience, stability, quality-of-service, convergence.
 
Received: 14 June 2017, Accepted: 7 September 2017
 

 
EXAMINING THE INFORMATION DISSEMINATION PROCESS ON SOCIAL MEDIA DURING THE MALAYSIA 2014 FLOODS USING SOCIAL NETWORK ANALYSIS (SNA)
Abdus-Samad Temitope Olanrewaju & Rahayu Ahmad
School of Computing, Universiti Utara Malaysia, Malaysia
samad.olanrewaju@gmail.com; rahayu@uum.edu.my
 
ABSTRACT | FULL TEXT
This article is based on a study which examined the information dissemination process on the social media during the Malaysia 2014 floods by employing the Social Network Analysis. Specifically, the study analyzed the type of network structure formed and its density, the influential people involved, and the kind of information shared during the flood. The data was collected from a non-governmental organization fan page (NGOFP) and a significant civilian fan page (ICFP) on Facebook using NodeXL. The two datasets contained 296 posts which generated different network structures based on the state of the flood, information available, and the needs of the information. Through content analysis, five common themes emerged from the information exchanges for both fan pages which helped in providing material and psychological support to the flood victims. However, only 5% of the networks’ population served as information providers, and this prompted the need for more active participation especially from organizations with certified information. Based on the findings presented and elaborated, this article concluded by stating the implications and recommendations of the study conducted.

Keywords: social network analysis, disaster, flood, Malaysia, information dissemination, influencers.
 
Received: 3 February 2017, Accepted: 3 August 2017

Universiti Utara Malaysia Press
 Universiti Utara Malaysia, 06010 UUM Sintok
Kedah Darul Aman, MALAYSIA
Phone: +604-928 4816, Fax : +604-928 4792

All Right Reserved. Copyright © 2010, Universiti Utara Malaysia Press