Current Issue : July - September Volume : 2012 Issue Number : 3 Articles : 5 Articles
Optical wireless data transmission systems for indoor\r\napplication are usually affected by optical interference induced\r\nby sun light and artificial ambient lights. This paper presents a\r\ncharacterization of the optical interference produced in visible\r\nlight communication (VLC) systems and proposes an effective\r\nscheme to solve it. Regarding the sun light noise and some\r\nartificial light noises reduction, the common method is to adapt\r\nthe optical bandpass filter which can distinguish the wavelength\r\nbetween interference lights and information lights. However, for\r\nsome photo-electric systems, the visible lights from the\r\ntransmitters occupy the same wavelength range, in this case, the\r\noptical bandpass filter would not reduce the interference noise\r\nfrom the other user, for example, the optical interference caused\r\nby multi-user access of the optical medium. Therefore, we\r\nproposed a novel scheme, adaptive optical parallel interference\r\ncancellation (AOPIC) to reduce the multiple access interference\r\n(MAI) and multiple user interference (MUI) induced by multiuser\r\naccess of the optical medium, the conventional parallel\r\ninterference cancellation (PIC) is analyzed as the comparison.\r\nThrough the simulation results, we can conclude that the AOPIC\r\nscheme shows much better bit error rate (BER) performance\r\nthan the conventional PIC with the increasing number of user....
We analyze and discuss encryption schemes for JPEG2000 based on the wavelet packet transform with a keydependent\r\nsubband structure. These schemes have been assumed to reduce the runtime complexity of encryption\r\nand compression. In addition to this ââ?¬Å?lightweightââ?¬Â nature, other advantages like encrypted domain signal\r\nprocessing have been reported. We systematically analyze encryption approaches based on key-dependent\r\nsubband structures in terms of their impact on compression performance, their computational complexity and the\r\nlevel of security they provide as compared to more classical techniques. Furthermore, we analyze the prerequisites\r\nand settings in which the previously reported advantages actually hold and in which settings no advantages can\r\nbe observed. As a final outcome it has to be stated that the compression integrated encryption approach based\r\non the idea of secret wavelet packets can not be recommended....
To use the structure of networks for identifying the\r\nimportance of nodes in peer-to-peer networks, a distributed linkbased\r\nranking of nodes is presented. Its aim is to calculate\r\nthe nodes� PageRank by utilising the local-link knowledge of\r\nneighborhood nodes rather than the entire network structure.\r\nThereby, an algorithm to determine the extended PageRank,\r\nwhich is called NodeRank of nodes by distributed random walks\r\nthat supports dynamic P2P networks is presented here. It takes\r\ninto account not only the probabilities of nodes to be visited\r\nby a set of random walkers but also network parameters as\r\nthe available bandwidth. NodeRanks calculated this way are\r\nthen applied for content distribution purposes. The algorithm\r\nis validated by numerical simulations. The results show that the\r\nnodes suited best to place sharable contents in the community\r\non are the ones with high NodeRanks, which also offer highbandwidth\r\nconnectivity....
This article aims at providing (i) a general\r\npresentation of the techniques and types of the intrusion\r\ndetection and prevention systems, (ii) an in-depth description\r\nof the evaluation, comparison and classification features of\r\nthe IDS and the IPS and (iii) the implications of such study\r\non how to determinate the features of some more effective\r\nIDS and IPS in the commercial domains and open source....
Identity theft is one of the most profitable crimes committed by felons. In the cyber space, this is commonly\r\nachieved using phishing. We propose here robust server side methodology to detect phishing attacks, called\r\nphishGILLNET, which incorporates the power of natural language processing and machine learning techniques.\r\nphishGILLNET is a multi-layered approach to detect phishing attacks. The first layer (phishGILLNET1) employs\r\nProbabilistic Latent Semantic Analysis (PLSA) to build a topic model. The topic model handles synonym (multiple\r\nwords with similar meaning), polysemy (words with multiple meanings), and other linguistic variations found in\r\nphishing. Intentional misspelled words found in phishing are handled using Levenshtein editing and Google APIs\r\nfor correction. Based on term document frequency matrix as input PLSA finds phishing and non-phishing topics\r\nusing tempered expectation maximization. The performance of phishGILLNET1 is evaluated using PLSA fold in\r\ntechnique and the classification is achieved using Fisher similarity. The second layer of phishGILLNET\r\n(phishGILLNET2) employs AdaBoost to build a robust classifier. Using probability distributions of the best PLSA\r\ntopics as features the classifier is built using AdaBoost. The third layer (phishGILLNET3) further expands\r\nphishGILLNET2 by building a classifier from labeled and unlabeled examples by employing Co-Training.\r\nExperiments were conducted using one of the largest public corpus of email data containing 400,000 emails.\r\nResults show that phishGILLNET3 outperforms state of the art phishing detection methods and achieves F-measure\r\nof 100%. Moreover, phishGILLNET3 requires only a small percentage (10%) of data be annotated thus saving\r\nsignificant time, labor, and avoiding errors incurred in human annotation....
Loading....