Performance Comparison of Face Recognition using Transform Domain Techniques  Abstract: The biometrics is a powerful tool to authenticate a person for  multiple applications. The face recognition is better biometrics compared  to other biometric traits as the image can be captured without the  knowledge and cooperation of a person. In this paper, we propose  Performance Comparison of Face Recognition using Transform Domain  Techniques (PCFTD). The face databases L – Spacek, JAFFE and NIR are  considered. The features of face are generated using wavelet families such  as Haar, Symelt and DB1 by considering approximation band only. The face  features are also generated using magnitudes of FFTs. The test image  features are compared with database features using Euclidian Distance  (ED). The performance parameters such as FAR, FRR, TSR and EER  computed using wavelet families and FFT. It is observed that the  performance of FFT is better compared to wavelet families. The success  rate of recognition is 100% for L – Spacek and JAFFE face databases as  compared to 95% for NIR face databases Keywords : Face Recognition; DWT; FFT; ED; Biometrics. Leader Election Algorithm in 3D Torus Networks with the Presence of One Link Failure Abstract:  Leader election is the process of choosing a leader for symmetry  breaking where each node in the network eventually decides whether it is a  leader or not. This paper proposes a new leader election algorithm to solve  the problem of leader failure in three dimensional torus networks. The  proposed algorithm solves the election problem despite the existent of link  failure.  In a network of N nodes connected by three dimensional torus  network, the new algorithm needs O(N) messages to elect a new leader in    time steps. These results are valid for two cases: the simple case where the leader failure is detected by one node, and the worst case where the  failure is discovered by N-1 nodes. Keywords :  Concurrency; Leader Election; Link Failure; leader failure; 3D  Torus Networks. A Three Stages Segmentation Model for a Higher Accurate off- line Arabic Handwriting Recognition  Abstract:  Arabic handwriting recognition considers a one of the hardest  applications of OCR system. The reason of that relates to characteristics of  Arabic characters and the way of writing cursively. Furthermore, no rules  can control on handwriting way, different styles, sizes and curves make the  process of recognition is very complex. On other side, the key for reaching  to good recognition is by getting a correct segmentation. Actually, the way  of segmentation is important, because if there is a small part is not clear in  character that will reflect on recognition process. In this paper we aim to  enhance the accuracy of off-line Arabic Hand Written text segmentation.   Three stages are proposed to reach to highest ratio of segmentation. Line  segmentation is the first stage, where it is proposed to separate each line.  We depend on row density to predict spaces among lines. Second stage is  Object segmentation and it is proposed to segment each word or sub word.  Eight neighbors connectivity are used to detect connected pixels. Final  stage is shape segmentation which is proposed to segment sub word to  characters. The idea in this stage is finding segmentation points among  branch points in the baseline. To apply that we propose four threshold  values to investigate on each branch point.   The result was satisfactory  and the model proved a good ability to tackle different types of texts with  bad samples. Keywords : Arabic handwritten recognition; Segmentation; Image  processing; Pattern recognition. Learning Vector Quantization (LVQ) and k-Nearest Neighbor for Intrusion Classification Abstract:  Attacks on computer infrastructure are becoming an increasingly  serious problem nowadays, and with the rapid expansion of computer  networks during the past decade, computer security has become a crucial  issue for protecting systems against threats, such as intrusions. Intrusion  detection is an interesting approach that could be used to improve the  security of network system.  Different soft-computing based methods have  been proposed in recent years for the development of intrusion detection  systems. This paper presents a composition of Learning Vector Quantization artificial neural network and k-Nearest Neighbor approach to detect  intrusion. A Supervised Learning Vector Quantization (LVQ) was trained for  the intrusion detection system; it consists of two layers with two different  transfer functions, competitive and linear. Competitive (hidden) and output  layers contain a specific number of neurons which are the sub attack types  and the main attack types respectively. k-Nearest Neighbor (kNN) as a  machine learning algorithm was implemented using different distance  measures and different k values, but the results demonstrates that using  the first norm instead the second norm and using k=1 gave the best results  among other possibilities. The experiments and evaluations of the proposed method have been performed using the NSL-KDD 99 intrusion detection  dataset. Hybrid (LVQ_kNN) was able to classify the datasets into five classes at learning rate 0.09 using 23 hidden neurons with classification rate about  89%. Keywords : Intrusion Detection System; Learning Vector Quantization; k- Nearest Neighbor. Performance Evaluation for VOIP over IP and MPLS Abstract:  Corporates and multisite organizations are now applying VOIP  usage all over their branches, this made offices with no boundaries and  reduced a huge amount of co