International Journal of Advanced Networking and Applications

4742-4745

Authors: ZHANG Xuan-yu, DENG Zuo-jie

In recent years, with the increasing demand for high-precision clothing image classification, people has attached great importance on the clothing image classification method based on CNN. In this paper, we discussed how to improve the accuracy of image classification by preprocessing the data set, combining the data set with CNN structure and improving the loss function. The comparative experiments between different structures of CNN and loss functions have been done. The experimental results show that the method we used can better improve the effect of classification and classification accuracy.

4750-4762

Authors: Yakubu Ajiji Makeri, Giuseppe T. Cirella, Francisco Javier Galas, Hamid Mohsin Jadah, Adetayo Olaniyi Adeniran

The aim of this paper was to enhance an enterprise network performance through VLAN implementation by simulating an existing LAN and rectify LAN using VLAN at, in two different scenarios. To rectify the drawbacks of existing enterprise network using virtual local area network. To simulate and compare the new (VLAN) and existing (NO_VLAN) enterprise network performance Virtual LAN technology can also be used to mitigate vulnerability surface for hackers by reducing the traffic request to servers, network visualization leads to ease of administration. There were more benefits of network virtualization through the implementation of VLAN. Thus, VLANs improved bandwidth utilization, power, speed and security. However, nowadays the high-speed network makes people to sense that network is a much more powerful tool to enhance in any business enterprise and make the company running more efficient. The main purpose of the network ability in the universities is to support learning, teaching, research and sharing of information. Unfortunately, even though the capacity of the bandwidth considered to be sufficient, the network can appear slow and inconsistent due to some factors like: the poor network design, the number of devices on the network, local network protocols (MAC Protocol) and misuse of the bandwidth by accessing to bandwidth hungry websites and applications (Dhurgham , 2018).

4776-4780

Authors: S. Rajarajeswari, T. Tamilarasi

Diseases are causing high rates of mortality in the modern world, chronic kidney disease (CKD) is one of the major causes of mortality, and it has a long-term disability. The predisposing factors for CKD include diabetes mellitus, hypertension, cardiovascular diseases, smoking, obesity, family history of kidney disease and congenital kidney problems. CKD is associated with many complications such as, proteinuria, anaemia of CKD, CKD-mineral and bone disorder, dyslipidemia and electrolytes imbalance. Renal replacement therapy (dialysis and kidney transplantation) is the treatment of choice for CKD. Data mining is an accurate technique helps to predict the disease using various methods includes logistic regression, naive bayes classification, k-nearest neighbours, and support vector machine. Apart from these previous techniques, it was necessary to use a classification method for data segmentation according to their diagnosis and regression method for finding risk factors. In this present study, data are classified using proposed Identification of Pattern Mining, Decision Tree methods and regression techniques are used to obtain the best levels and this can be taken as metrics that the proposed methods can help in diagnosing a patient with CKD.

4793-4799

Authors: Kayalvizhi E, Balamurugan Gopalakrishnan

In cognitive radio network, the spectrum sensing finds either the channel is occupied or idle, the problem is assigning the unused channels of the primary user (PU) to the secondary users in an efficient manner is most challenging issue. In this work, we investigates the above issue and proposed an adaptive resource allocation to the secondary users in terms of channel allocation and power allocation. The proposed work intelligently handles both frequency and space efficiently without affecting the quality of service (QoS) of the primary user. We considered both underlay and overlay spectrum access, based on that resource allocation is carried out in an efficient manner. The maximum transmitted data rate of the secondary user (SU) obtained is 225Kbps determined by using Shannon channel capacity theorem. The proposed work also shows the effectiveness of the simulation in terms of energy efficiency up to 8.25 x 10 5 bits/Joule.