A. Siva Sangari ¹, D. Saraswady ², G. Sasikumar ³

Affiliations
1 Department of ECE, GMR Institute of Technology, Visakapattinam - 532127, Andhra Pradesh, IN
2 Department of ECE, Pondicherry Engineering College, Pillaichavadi - 605014, Pondicherry, IN
3 Department of Mechanical Engineering, GMR Institute of Technology, Visakapattinam - 532127, Andhra Pradesh, IN

Abstract

In gift agricultural subject, pest identity causes enormous discount in both fine and quantity of crop cultivation. If you want to growth the production price of crop, the presence of tiny pests such as aphids, whiteflies, and spider mites which purpose leaf deformation is the main hassle. Consequently early pest detection is a primary project in agricultural field. This research specializes in photograph processing, sensible set of rules and computer technology to expand a new pest detection device that’s vital and imperious to crop cultivation. In existing system, the pest detection using fuzzy c-manner clustering address the problem of overlapping of gadgets can purpose for hiding the pest. As cluster length will increase the rims are brittle and compactness of the clusters gets altered. Therefore the prevailing FCM segmentation now not able to cope with the constituent parts of the picture implicitly. The proposed technique known as ant colony optimization method resolves this hassle, and offers the implicit pest segmentation. The segmentation approach uses the swarm intelligence approach based totally at the behavior of the ant colonies. Ant colony optimization became used to extract the area of the insect pest and also to attain the most suitable constant parameters of recognized insect pest. In the beginning the captured pix are processed for pre-processing. Then photo segmentation is done based totally on ACO to get pest target location. Later constant parameters are measured for pest image segmentation the use of ACO which include structural content, top signal to noise ratio, normalized correlation coefficient, and average difference and normalized absolute mistakes. The matlab simulation experiments demonstrate that this proposed ACO technique is extra powerful than fuzzy c-way clustering consequently it can phase the pest photograph higher.

Keywords

Ant Colony Optimization, Consistency Measures, Early Pest detection, Fuzzy C-means.

Full Text

   

P. V. Sabiq ¹, D. Saraswady ¹

Affiliations
1 Department of ECE, Pondicherry Engineering College, Puducherry, IN

Abstract

Background/Objectives: The spectrum scarcity problem arising due to reckless development in wireless communication can be addressed using cognitive radio technology. An important step in cognitive radio technology is spectrum sensing and associated with that is an attack called primary user emulation attack. Methods/Statistical Analysis: A two channel Quadrature Mirror Filter (QMF) bank is employed for a single user spectrum sensing. The energy detection method is used as reference for comparing the performance of the filter bank method in the presence of attackers. Existing approaches like Neyman-Pearson Criterion, Improved Detection Scheme with Double threshold and Location Verification Method for detecting PUE attacks were examined. A new method which combines improved detection scheme and location verification method is proposed for detecting PUE attacks. Findings: The simulation result shows that the filter bank method shows better performance than the energy detection method in the presence of attackers. The probability of detection of PUE attacks using improved detection scheme and location verification method for the filter bank method and energy detection method was analyzed and found out that the filter bank based spectrum sensing method outperforms energy detection method. By incorporating AND rule logic, OR rule logic and Alternate logic approach into the filter bank spectrum sensing technique, the probability of detection of PUE attack was investigated and it is seen that the probability of detection has increased to 0.98 in FB method at SNR = 1 dB when employing an alternate approach. Application/Improvements: The feasibility of deploying the filter bank method integrated with the alternate logic is to be tested using Universal Software Radio Peripheral (USRP) module.

Keywords

Double Threshold, Filter Bank, Fusion Logic, Primary User Emulation Attack, Spectrum Sensing

Full Text

   

D. Jagadiswary ¹, D. Saraswady ¹

Affiliations
1 Department of Electronics and Communication Engineering, Pondicherry Engineering College, Pillaichavadi - 605014, Puducherry, IN

Abstract

Generally, security, expediency and outlay are the three crucial factors influencing the espousal of biometrics. Seeing that, a newly scheme is proposed from the fused biometrics specifically fingerprint, finger vein and retina using Modified RSA (MDRSA), a public key cryptosystem used for both numeric and images. But the existing Hyper Image Encryption Algorithm (HIEA) is used only for images. Using MATLAB 2014, the performance is measured through GAR and FAR. Consequently, comparison of simulation results shows that MDRSA gives higher genuine rate of 95.3% and false rate of 0.01% which is reduced in case of HIEA, as it provides genuine rate of 92% and false rate of 0%.

Keywords

Biometric Authentication, Fused Biometrics, MDRSA, HIEA.

Full Text