Malicious Cluster Identification
in Cognitive Radio Spectrum Sensing Using Fuzzy-Based Classifier
--Sanvi Khandelwal, Preeti Trivedi and S V Charhate
The cooperative communication system has gained lots of interest due to effective utilization of the allotted spectrum. Spectrum sensing techniques are a key factor for the cooperative communication system. A number of sensors are deployed to sense the spectrum hole. The group or clusters of the sensor are used to make out the decision about the spectrum hole. The presence of malicious sensor affects the performance of spectrum sensing. In this paper, the trust on the group or cluster has been identified based on the fuzzy classification. The fuzzy classifier takes the input in terms of reporting cluster trust value, neighborhood trust value and primary weight index. The membership values of these three linguistic variables are used by fuzzy interference system to decide the trust value of the group or cluster. The number of malicious user has been introduced during the simulation and performance has been identified. The Eigenvalue-based spectrum sensing is also simulated in this paper and the effect of different SNRs on the decision accuracy is shown.
Enhanced Opportunistic Network Parameters
--Anushree S Dhore, R D Kanphade and G S Ambadkar
In the last few years, there has been a drastic increase in the use of mobile devices. Along with the increase of these devices, the data generated is massive. These data are also shared between them. Many protocols are used for data sharing. System parameters assume a critical part in the productive working of it. Additionally, these parameters are firmly identified with each other. Opportunistic sharing means that these devices have an opportunity to share data. When the data is shared, all the nodes are active during transmission. This causes network congestion or traffic and influences the execution of the system. This paper includes two methods for selecting only one node for transmission, instead of all nodes being active. After implementing these methods, network parameters are improved. Detection of network is done depending on the number of the nodes in network. The first method is to detect if the network is sparse or dense network. The second method is to count the hop limit and lifetime of sharing data between two mobile nodes. Parcel delivery proportion, throughput, vitality and overhead are the parameters considered in this paper.
© 2017 IUP. All Rights Reserved.
The Performance of Wireless Sensor Network Simulators Using Advanced On-Demand Vector (AODV) Protocol
--M Khaleel Ullah Khan and K S Ramesh
Simulation in wireless networks is very important before the real-time implementation of any project. For this purpose, design engineers and research community first simulate the networks design and then go for implementation of any project. These will help to build new theories and hypotheses. Since the invention of wireless networks, a number of simulators have been available, and upgradation and enhancement in characteristics are being done for the network simulators. A few examples of these simulators are NS-2, NS-3, SWAN, JIST, OMNET, GloMoSim, etc. The crucial decision for selecting the simulator depends upon its evaluating characteristics such as speed of evaluation, memory usage for running the simulator, utilization of the CPU cycles and scalability by simulating a routing protocol. Therefore it becomes important to choose the simulator accordingly depending upon the project. In this paper, different network simulators like NS-3, NS-2, GloMoSim and OMNet++ are used to compare network performance characteristics like memory usage, CPU utilization and computational time using AODV protocol.
© 2017 IUP. All Rights Reserved.
Thorough Investigation of Artificial Neural Network with Applied Back Propagation Algorithm in Aperture Coupled Microstrip Patch Antenna
--Satish K Jain and Jaya Chakrawarty
The paper presents application of Artificial Neural Network (ANN)-based technique for the analysis of aperture coupled feed microstrip patch antenna. Thorough discussion of ANN along with back propagation algorithm has been done. A neural network-based model using back prorogation algorithm was developed with the help of few antennae variable geometrical parameters as inputs and their responses. This trained network is able to locate the resonance frequency for microstrip patch antenna having aperture coupled feed operating within the L (1-2 GHz), S (2-4 GHz), C (4-8 GHz) and partial X-band X (8-10 GHz) bands. These frequency bands are useful for various indoor Wireless Local Area Network (WLAN) applications and little satellite communication respectively. Developed ANN model takes design (geometrical) parameters of antenna like square patch dimension(length = width), length of the slot on the ground plane, width of the slot as inputs and deliver the respective resonance frequency as output within almost no time. The validity of the network is tested with the simulation results obtained from the CST software and experimental result.
© 2017 IUP. All Rights Reserved.
|