NEW SDDO-BASED BLOCK CIPHER FOR WIRELESS
SENSOR NETWORK SECURITY
Nguyen Hieu Minh, Do Thi Bac, Ho Ngoc Duy
Abstract: Wireless sensor networks (WSNs) are exposed to a variety of attacks. The quality and complexity of attacks are rising day by day. Limitations in computational and battery power in sensor nodes are constraints on the diversity of security mechanisms. This paper concerns the problem of using of the switchable data-dependent operations (SDDOs) oriented to the design of fast cipher suitable to applications in constrained environments. The new SDDO-based block cipher using this approach presented and estimated. The security estimations show that cipher MD-64 is less likely to suffer intrusion of differential cryptanalysis than currently used popular WSN ciphers like DES, Camellia and so on. Moreover, FPGA synthesis result for hardware implementation (FPGA) proves that new cipher MD-64 is very efficient, especially for WSN
International Journal of Computer Science & Network Security
-VOL.10 No.3, March 2010;pp 54-60.
FAST HASH FUNCTIONS BASED ON CONTROL PERMUTATION
Do Thi Bac, Nguyen Hieu Minh, Молдовян У.А., Молдовяну П.
Abstract: А Using variable bit permutations there are designed hash functions prossessing high performance while hardware, firmware, and software implementation. Controlled bit permutation operation have been selected in correspondence with earlier proposed topology of the permutation network for implementing on its base the controlled bit permutation in struction oriented to embedding in the mass microprocessors. Due to the last fact the designed algorithms for computing cryptographic checksums are perspective for application as fast software-suitable hash function.
Вопросы защиты информации. 2, 2-6, 2010.
FAST BLOCK CIPHERS BASED ON SWITCHABLE SUBSTITUTION-PERMUTATION NETWORKS
Ho Ngoc Duy, Do Thi Bac, Молдовян А.А., Молдовян Д.Н.
Abstract: There are designed 64 bit and 128-bit block ciphers using switchable substitution permutation networkt . Due to the switchablitity of the used networks it is provided possibility to perform direct or inverse cryptographic transformation cost. There are investigated the statistic properties of the constructed ciphers and their security against diferential analysis
Вопросы защиты информации. 2, 7-17, 2010.
EQUIVALENCE AND STABILITY OF TWO-LAYER CELLULAR
NEURAL NETWORK SOLVING
Vu Duc Thai, Pham Thuong Cat
Abstract: Design and prove the stability and similarity between oringal mathematical and CNN model
Proceeding of ICARCV2010 (Singapore), Tr. 704-709
HEALTH INSURANCE DOCUMENT EXCHANGE SYSTEM IN VIETNAM
Nguyen Hai Minh, Hangchan Kim, Byoung-Kee Yi, Il Kon Kim, Yun Sik Kwak
Abstract: We propose a design for a system to support exchange of relevant clinical documents between healthcare providers and health insurance agencies based on international standards with Vietnam-specific considerations.
Proceedings of the international conference CJKMI 2010 on Japan, pp 73:74.
VECTOR FIELDS IN EXPRESSING HAIRSTYLES
Nguyen Van Huan, Do Nang Toan
Abstract: Build up a hairs model and hairs style is the most important period in a virtual human simulation system. So, to build a ensured quality simulation system, propose a effective solution is very important. This article talking about a static vector area application in bulding hairs models and hairs style by using area vector will allow to display different hairs style and rapid created, ensured quality and general.
The 2nd IEEE International Conference on Advanced Computer Control –ICACC 2010, 541-546, 2010.
IMPROVING BONE-CONDUCTED SPEECH RESTORATION IN NOISY ENVIRONMENT BASED ON LP SCHEME, 2010 .
Phung Trung Nghia, Masashi UNOKI, Masato AKAGI
Abstract: The restoration of bone-conducted (BC) speech is a very important issue for robust speech communication in extremely noisy environments. In our previous studies, we proposed a blind restoration method based on linear prediction (LP) scheme with a prediction method based on the simple recurrent neural network (SRN). However, SRN-based prediction is not suitable for training with the large corpus, which is necessary for real applications. The over-training problem of SRN makes difficult to train various kinds of BC speech in one session. In addition, it is difficult to adapt the SRN model to BC in unkown noisy environment. in order to build the open dataset BC speech restoration. Thus, in this paper, a prediction method based on the Gaussian mixture model (GMM) was used instead of SRN. A method for estimating the LP residual ratio was also proposed. We then investigate whether how the proposed method works to restore BC speech in high noisy environments. Objective and subjective evaluations were carried out to evaluate the improvements in sound quality and intelligibility of the restored speech. The results revealed that our proposed method
LSF-GMM outperforms the previous LSF-SRN and both of the two LP-scheme-based methods can adequately improve the voice quality and intelligibility of BC speech, in clean and high noisy environments
Proceeding of the apsipa 2010, student section, p. 10, dec, singapore 2010.
SPEED CONTROL OF 3-PHASE ASYNCHRONOUS MOTOR USING ARTIFICIAL NEURAL NETWORK
Pham Thuong Cat, Le Hung Linh, Pham Minh Tuan
Abstract: Speed control of an alternating current (AC) motor has been one of the difficult control problems. Many approaches have been proposed to solve these problems. In this paper, a speed control method is proposed using artificial neural network (ANN) with an online self-learning algorithm to compensate uncertain parameters in the dynamics model of AC motor. The global asymptotic stability of the control system is proved using Lyapunov stability method. Simulation results on MATLAB show the reliability and accuracy of the proposed method.
2010 8th IEEE International on Control and Automation Xiamen, China, June 9-11,2010. p832-p836
IMPROVEMENT OF SUCCESSFUL LOOKUP RATIO OF CHORD DISTRIBUTED HASH TABLES (DHTS) IN WIRELESS
Nguyen Chan Hung, Vu Thanh Vinh, Ngo Hoang Giang
Abstract: One of the most critical problems of peer to peer (P2P) multimedia applications is the effect of churn to the efficiency of locating data items, which is a fundamental function of P2P networks. Especially, when P2P network extends to the mobile environment, churn effect may heavily affect the whole network and disrupt normal service. In this paper, we adapt a mechanism named atomic ring maintenance for Chord DHT to mitigate the effect of churn. By implementing two mechanisms: 1) a new locking mechanism to ensure the stability of neighbor nodes and eliminate concurrent joining and 2) modifying the pointer stabilization process, we try to mitigate the inconsistency of node pointers to increase DHT performance. Simulation results show significant improvement in successful ratio of lookups for modified Chord protocol in high churn condition while latency of successful lookups is as good as that in basic Chord. We also observed the phenomenon of topology partitioning under churn and investigate the correlation between successful lookup ratio and network size and churn rate.
Proceedings of International Conference on Advanced Technologies for Communications, Ho Chi Minh, Viet Nam, 2010, pg 119-126.