Showing papers by "Narula Institute of Technology published in 2012"

An application of genetic algorithm method for solving patrol manpower deployment problems through fuzzy goal programming in traffic management system: a case study

Abstract: This article demonstrates a fuzzy goal programming (FGP) approach with the use of genetic algorithm (GA) for proper deployment of patrol manpower to various road-segment areas in urban environment in different shifts of a time period to deterring violation of traffic rules and thereby reducing the accident rates in a traffic control planning horizon. To expound the potential use of the approach, a case example of the city Kolkata, West Bengal, INDIA, is solved.

An improved flood warning system using WSN and Artificial Neural Network

Abstract: The operation of the surface water system for flood control is very important and crucial to minimizing the impacts of flood during the real-time flood events. The model considered here is assumed for a sophisticated flood warning system. Many works have been done with different parameters using Artificial Neural Network (ANN) but this paper considers six parameters related to flood effecting causes directly or indirectly. The input parameters are theoretically bounded and prediction simulated using ANN. The output results obtained are perfectly matched with the taken model and satisfactorily acceptable. The output gives early warning as well as flood situation for disaster management and preparedness to combat aftermath of flood.

Artificial Neural Networks for Cognitive Radio: A Preliminary Survey

Abstract: The cognitive radio is the emerging field where an efficient use of spectrum is desired by detecting white spaces. The Artificial intelligence is the core of the cognitive engine which observes the external and internal environment parameters and acts to increase the QoS of the communication system. In this paper a survey of how artificial neural networks can be used to develop this intelligent core. Various techniques recently suggested by researchers e.g. automatic modulation classification, pattern recognition based cognitive transmitter / receiver , ANN based learner and spectrum predictor have been discussed in this paper

Effect of strongly coupled plasma on the doubly excited states of heliumlike ions

Abstract: The effect of strongly coupled plasma occurring in astrophysical context has been studied for the first time to estimate the energy levels of the autoionizing states of highly stripped astrophysically important ions Al11+, Si12+, P13+, S14+ and Cl15+ and also C4+ isoelectronic to helium. The transition energies corresponding to 1s 2:1Se → 2s 2:1Se, 2p 2:1De, 2s2p:1Po, 2s3d:1De and 2p3d:1Fo are analyzed with respect to different plasma densities using the ion sphere (IS) model of strongly coupled plasma. Transition energies are obtained from the position of the poles of a variational functional based on frequency dependent perturbation calculation probing the collective oscillation modes of the plasma embedded two electron ions. For the free ions corresponding to zero plasma coupling our calculated data agree well with those obtained from spectroscopic data while for the plasma embedded ions the data are new but follow systematic trend. The work has been performed in the domain of linear response theory. The analytical wave function of the doubly excited states have been obtained and may be useful for calculating various transition properties of the plasma embedded ions and also for estimating the rate coefficients for dielectronic recombination processes which play a major role in maintaining equilibrium in high temperature astrophysical or laser produced plasmas.

Efficient clustering techniques to optimize the system lifetime in Wireless Sensor Network

Abstract: Mobility of sensor node in Wireless Sensor Network (WSN) is one of the key advantages of wireless over fixed communication system. But to track the sensor node in the heterogeneous network is more challenging and difficulties. In heterogeneous system, generally power consumption is more then homogeneous system. Coordination in distributed sensor network the implementation of clustering is an important technique and clusters of bounded size which is the total number of nodes in a specific cluster, is an important parameter in clustering algorithms which are very much effective in reducing energy consumption by minimizing the neighborhood of a node. Communication cost is also an important parameter for computation in a large area. Clustering techniques is Wireless Sensor Networks (WSNs) compare to random sampling is less costly due to the saving of time in journeys, reduction in number of transmissions and receptions at each node, identification, contacts etc. Which are valuable for increasing the overall network life, scalability of WSNs. Clustering sensor nodes is an effective and efficient technique for achieving all the requirement. In this paper we study the comprehensive theoretical aspects of the clustering problem to energy optimization in wireless sensor networks.

Investigation of cross sensitivity of single and double electrode of admittance type level measurement

Abstract: Level measurements of liquid material are one of the important parameters in a process industry. Low cost admittance type liquid level measuring system using single electrode with conducting vessel and double electrode with insulating vessel already exist. In the present work, attempts have been made to study the cross sensitivity of temperature in the admittance type liquid level measurement. The cross sensitivity in the admittance measurement due to temperature is considered for admittance calculations. Error curves due to cross sensitivity for both single electrode and double electrode are simulated using MATLAB. The cross sensitivity due to temperature has been found significant and can be used for simultaneous measurement of temperature and liquid level.

Fuzzy min-max neural network for satellite infrared image clustering

Abstract: The process of estimation of precipitation from satellite images begins with the detection and identification of convective clouds. Clustering of the satellite infrared images is required in order to estimate the cloud cover area. In this paper a neuro-fuzzy technique in the form of unsupervised fuzzy minmax clustering neural (FMMCN) network has been implemented for clustering satellite infrared image. Each cluster is in the form of an n-dimensional hyperbox defined by minimum and maximum points and a fuzzy membership function. FMMCN suits this application area because it is completely unsupervised and hence, unlabeled data can be used with it. Also the number of clusters is not required to be mentioned at the beginning as it is calculated dynamically.

A scalable neural-network modular-array architecture for real-time multi-parameter damage detection in plate structures using single sensor output

Abstract: A scalable modular neural network array architecture has been proposed for real time damage detection in plate like structures for structural health monitoring applications. Damages in a plate like structure are simulated using finite element method of numeric system simulation. Various damage states are numerically simulated by varying Young's modulus of the material at various locations of the structure. Transient vibratory loads are applied at one end of the beam and picked at the other end by means of point sensors. The vibration signals thus obtained are then filtered and subjected to wavelet transform (WT) based multi resolution analysis (MRA) to extract features and identify them. The redundant features are removed and only the principal features are retained using principal component analysis (PCA). A large database of principal features (the feature base) corresponding to different damage scenarios is created. This feature base is used to train individual multi layer perceptron (MLP) networks to identify different parameters of the damage such as location and extent (Young's modulus). Individually trained MLP units are then organized and connected in parallel so that different damage parameters can be identified almost simultaneously, on being fed with new signal feature vectors. For a given case, damage classification success rate has been found to be encouraging. The main feature of this implementation is that it is scalable. That is, any number of trained MLP units capable of identifying a certain parameter of damages can be integrated into the architecture and theoretically it will take almost the same time to identify various damage parameters irrespective of their numbers.

Three-body negative ions under Coulomb interaction

Abstract: The ground state energy eigenvalues of the symmetric three-body exotic negative ions p+π−π− and p+K−K− have been determined variationally for the first time using an explicitly correlated Hylleraas basis set. Ground state energies of Ps− and p+μ−μ− were also determined to check the accuracy of the present methodology by comparing those results to a few accurate earlier results for these systems.

Temperature induced mean based cloud motion prediction model from satellite infrared images

Abstract: Detection and identification of convective clouds can be done from thermal infrared (TIR) images (10.5–12.5µm) as clouds are associated with extremely low temperature. Clouds can be tracked from a given sequence of satellite images, and becomes useful for weather now-casting. There are several models available for cloud motion prediction. Adjusted Mean Based Prediction (AMBP) Model is one of them. A modified model of AMBP, based on temperature, for cloud movement tracking has been proposed and presented in this paper for better prediction.

Pattern Recognition for damage detection in aerospace vehicle structures

Abstract: A Pattern Recognition based damage detection scheme for aerospace vehicle structures is proposed. It involves capturing mechanical vibration signals from plate like structures using displacement sensors; removal of noise and extraction of features using Wavelet Transform based signal processing techniques, and training a Neural Network Ensemble to classify and identify the damages that appear in the structure. A few cases are studied. Encouraging results are found in classifying single damages in the structure. However success rates dropped in case of identifying multiple damages for the same structural form. A sensor placement strategy is then drawn out that improved the results significantly.

Microcontroller based automatic fault identification in a distribution system

Abstract: A delta-star transformer connected to distribution system has been analyzed with the approach of generalized theory of electrical machines and expressions for symmetrical and different unsymmetrical faults are derived. The detailed theoretical analysis shows that magnitude and phase angle of fault current varies depending on nature and types of faults. A microcontroller based continuous monitoring unit is developed to detect and identify types of faults in the transformer connected distribution system. Software executed in microcontroller evaluates nature of a fault based on measured magnitudes and phase angles of voltages and currents under faulty condition. Theoretical analysis and experimental results validate acceptability of the developed unit‥

Mathematical modelling of packet transmission through cluster head from unequal clusters in WSN

Abstract: The network lifetime always a critical issue in sensor networks because the sensor nodes are characterized by restricted and non-replaceable energy supply. Thus maximizing lifetime of network by minimizing energy consumption poses a challenge in design of protocols. Therefore, proper organization of clustering and orientation of nodes within the cluster becomes one of the important issue to extend the lifetime of the whole sensor network through cluster head. We investigate the problem of energy consumption in ClusterHead (CH) rotation (i.e. re- clustering) in wireless sensor networks. The selection criteria of the CH are based on the residual energy of nodes within the clusters and minimum average distance from the Base Station. The total energy in delivering a packet from the sensor node to other nodes has been mathematically derived. Moreover, the expected number of packet retransmissions and the effect of it on the network is also discussed in our proposed energy model. In this paper we applied the approach for producing energy-aware unequal clusters with optimal selection of cluster head and discussed several aspects of the network mathematically and statistically. This work presents an analysis of its design and implementation aspects. The simulation results demonstrate that our approach of re-clustering in terms of energy consumption and lifetime parameters.

Trusted Routing in AODV Protocol Against Wormhole Attack

Abstract: This research aims to develop an enhanced version of AODV including trust and reputation that will ensure the transmission of messages in a secure way avoiding wormhole attackers in the network. Firstly, wormhole attackers are identified depending on their transmission related misbehavior with respect to other nodes in the network. Next, appropriate trust value is assigned to the misbehaving nodes. Furthermore the measure of trust is calculated by considering the past interaction between the considered node with other nodes in terms of direct and indirect interactions and related trust values. Finally, the trust evaluating node takes the decision to include or not to include the trustee node in discovered route and secures the path from wormhole attacker. The key concept to interpret the misbehavior of a node is based on its responsiveness towards incoming route requests in terms of associated processing time, latency, threshold time duration. To facilitate the trust computation and respective decision, the enhanced version assigns the tasks to different functional components: Context Analyser, Event Analyser, Trust Manager, Trust Repository, Notifier, Decision Manager, Trust Engine and Node Manager. Our analysis and simulation results show the effectiveness of our proposal against wormhole attack in AODV.

Are Meditative states distinguishable from the pre-meditative ones? A new alternative study

Abstract: Analysis of the HRV signals towards classification of pre-meditative and meditative states has been widely accepted among the research communities. However, none of the existing methods used so far successfully classifies the meditative and pre-meditative states. In this article, we introduce a new approach, which is based on Similarity Index introduced by A.C.C. Yang et. al. The method calculates the rank for each word of length four and compares probability distributions between each pair of subjects (both in meditative and pre-meditative states) by using Q-Q plot. The Q-Q plots show same probability distribution for the meditative states and different probability distributions for the pre-meditative states. Thus Q-Q plots clearly distinguish the meditative states from the pre-meditative ones.

Channel allocation in a cognitive radio network using non deterministic Q learning algorithm

Abstract: Cognitive radio technology has been emerged to provide the solution of improvement of spectrum utilization. In this paper we try to solve the problem of allocation of channels to the secondary users using non deterministic Q learning technique. Using Q learning technique, we try to provide prior knowledge to the secondary users about the channel usage pattern of licensed users so that a secondary user can efficiently select channels in a multi cell cognitive radio network. Firstly, a channel assignment algorithm based on non deterministic Q learning technique has been proposed in this paper. Using ageing technique, starvation of low priority users is removed from the algorithm. Simulation result shows that the proposed channel assignment algorithm is effective.

Throughput Analysis for Fixed Allocation of Channels in Cognitive Radio Network

Abstract: In this paper we discuss throughput analysis for fixed allocation of channels in a cognitive radio network. In fixed allocation of channels number of primary users allocated to a channel is fixed all the time. But the number of primary users is different for different channels. The secondary users or cognitive users sense the channel whether the channel is free or not. But sometimes when secondary users access the channel then primary users arrive and collision occurs. The transmission probabilities of primary and secondary users are different. The packet transmission delay and channel access delay of secondary users is considered in this paper. Our objective is to find the optimal number of secondary users to maximize the total throughput of primary and secondary users.

Performance analysis of Adaptive Power Control algorithm for varying target SIR

Abstract: In this paper we study the performance of two Power Control algorithms: the Adaptive Standard Power Control (ASPC) and the Standard Power Control (PCA) algorithms. Simulation of both algorithms is being done under three different network conditions including the case with varying SIR target. The results from our simulations are shown graphically. These results show that the performance ASPC is be better than PCA under the different network conditions considered in this paper. All simulations are done in MATLAB.

The movement of sensors within cluster in WSN is Elliptic in nature

Abstract: A wireless sensor network (WSN) has to maintain a desirable sensing area of coverage and periodically report sensed data to the administrative center (i.e., base station, router etc.). Then the coverage area of the movement of sensors and lifetime are two important problems in a WSN due to constraint of limited battery energy. Sensors are usually deployed randomly in coverage area of interest and therefore the sensor tracking become one of the biggest challenge in Wireless Sensor Networks (WSNs). Many works done to track the sensor nodes in WSNs and all previous theoretical analyses on the coverage area and lifetime are primarily focused on the random uniform distribution of sensors. Here is depicting a real analysis in WSNs, In this paper, we try to find out the coverage area of sensors and best of our knowledge we first provide a mathematical and analytical framework for the coverage area of the deployment sensor nodes in a WSN that follows a multivariable normal distribution or Gaussian distribution. The multivariate normal distribution is often used to describe correlated real-valued random variables each of which clusters around a mean value which gives the monitored region. We determine the randomly deployed sensor nodes, follows an Elliptic area in wireless sensor network.

Convective development at Kolkata (22.53° N, 88.33° E), India during pre-monsoon season using linear discriminant analysis technique

Abstract: In the present work, statistical indices are formed using the most effective different combinations of 22 and 20 thermodynamic and dynamic parameters for the convective development at Kolkata (22.53° N, 88.33° E), India during pre-monsoon season (Mar-May) utilizing the data of 12 years (1985-1996). A multivariate statistical technique, namely Linear discriminant analysis (LDA) has been utilized to 22 primarily selected parameters derived from the radiosonde observations of 0000 UTC for next 12 hours yields respectively 59.57% and 58.70% correct prediction for convective development (CD) and fair-weather (FW) in next three years (1997 to 1999). A similar analysis for radiosonde observations of 1200 UTC for next 12 hours yields 63.79% and 50% for CD and FW respectively. Another similar LDA analysis with the above data period utilizing 20 parameters [excluding Miller’s (1972) & George’s (1960) from the earlier set] built from the radiosonde observations of 0000 UTC for next 12 hours observation yield 63.83% & 56.21% correct prediction for CD and FW respectively in the next 3 years. The corresponding figures for 1200 UTC for next 12 hours are 54.41 % & 67.34% respectively. With a view to understand the effect of the parameters, namely convective available potential energy (CAPE) and Convective inhibition (CIN), a similar LDA analysis has been applied to 17 parameter set (including the ratio of CAPE and CIN and excluding (θes -θe) at each level constructed form the above; θes and θe being saturated equivalent potential temperature and equivalent potential temperature respectively). The radiosonde observations of 0000 UTC for next 12 hours yield 68.29% and 54.43% correct prediction for CD and FW respectively in the next 3 years. Next, 1200 UTC radiosonde observations for next 12 hours yield 77.08% and 57.44% correct prediction for CD and FW respectively. For all the 22, 20 and 17 set of parameters (both for 0000 UTC and 1200 UTC observations for next 12 hours), the efficient skill scores, namely, True skill score (TSS), Heidke skill score (HSS), Critical success index (CSI) are computed. The 17 parameter combination is most efficient and may be utilized in an effective manner for prediction purpose. The investigations reveal that both for 0000 UTC and 1200 UTC observations for next 12 hours, correct prediction improves with the inclusion of the parameter CAPE/CIN (17 parameter combination), especially for 1200 UTC observations. Also, afternoon predictions are more effective than morning predictions for 22 parameter, 20 parameter and 17 parameter cases.

Monte Carlo analysis to estimate the performance of adaptive Power Control algorithm for varying target SIR

Abstract: In this paper we study the performance of two Power Control algorithms: the Adaptive Standard Power Control (ASPC) and the Standard Power Control (PCA) algorithms. Simulation of both algorithms is being done under three different network conditions including the case with varying SIR target. In order to achieve robust estimates of the performance we have undertaken Monte-Carlo analysis of 500 runs for each case. These results show that the performance ASPC is be better than PCA under the different network conditions considered in this paper. All simulations are done in MATLAB.