Kongu Engineering College

About: Kongu Engineering College is a based out in . It is known for research contribution in the topics: Computer science & Cluster analysis. The organization has 2001 authors who have published 1978 publications receiving 16923 citations.

Smart Sensor Network-Based Autonomous Fire Extinguish Robot Using IoT

Abstract: Fire explosion is among the main reasons for death in the world. The urban spaces have a lot of population, many systems have control over fire detection but not over control of fire due to lack of functionalities. The operation of the robot depends on the android application on the smartphone. It can also be communicated using Wireless fidelity technology. The motion detection technology is embedded in it, which can identify the objects or obstacles. With Arduino microcontroller and IoT technology, this robot can send emergency alerts in critical conditions, explore the compounds, and effectively check for the fire. The entire robot was contained in the sensor-based network, and if the sensors fail to detect, the robot still will operate and extinguish fire with the help of an inbuilt camera. Finally, this robot is tested for operation in automated mode and live video buffering capability.

Modelling and Simulation of Artificial Neural Network Based Harmonic Elimination Technique for Solar-Fed Cascaded Multilevel Inverter

Abstract: This paper presents the harmonic elimination technique for a fifteen level cascaded multilevel inverter (CMLI) supplied with the seven varying DC input sources. Here, the input supply sources are obtained from the solar Photovoltaic’s (PV). The harmonic equations are solved and switching angles are obtained in such a way that the fundamental is kept constant and lower order harmonics are minimized or eliminated. The data set with varying input DC voltages and switching angles are trained with Artificial Neural Network (ANN). This trained network is integrated with CMLI to improve the power quality. The modeling of solar panel which serves as the input to the CMLI is developed. A eleven level CMLI with five varying DC input sources and a set of multiple carrier Pulse Width Modulation (PWM) methods are also given for comparison. The results show that ANN based method provides a significant improvement with respect to the reduction of Total Harmonic Distortion (THD).

RFDCAR: Robust failure node detection and dynamic congestion aware routing with network coding technique for wireless sensor network

Abstract: Wireless sensor networks is an attractive concept that is being implemented in all fields of work with diverse applications. Hence it is not a surprise that there are several wireless routing algorithms available and they mainly focus on reducing the consumption of energy in WSN, the direction of the fact that a sensor node point work on batteries. But algorithms do not study on these energy deficient nodes and their collision effects. There are various reasons for node failure that can fall under mechanical or electrical problems, battery depletion, environmental degradation or hostile tampering. But the most common failure of nodes occur due to limited energy availability. Failure caused due to a group of nodes can minimize the network paths. These activities can lead to failures in the subset of acting nodes resulting in a disconnected or no path situation from the network. This algorithm introduces the multipath node disjoint routing by combining local and global procedures for adaptive route. The capable nodes in the network are located using Lyapunoy optimization technique through network coding technique that enhances the operation and lifetime of the entire network. Through weight of the packet and along with the packet receiving ratio the algorithm separates the packets and route them to a different path to the base station thereby improving delivery and optimizing time and energy. Simulations are conducted in a NS3 environment and proved that this algorithm is efficient in performance than the existing methods.

An efficient lightweight cryptographic scheme of signcryption based on hyperelliptic curve

Abstract: The emerging trend now in network security is lightweight cryptography which is the need of the hour due to the growth of wireless technology. As the internet technology grows day by day identity authentication in wireless network is more important. Generally, these wireless systems are in demand of security and resource (power). In order to meet these constraints an important lightweight scheme called Remote user authentication with signcryption is proposed with security features such as confidentiality, integrity (originality of sender and receiver), message and user authentication, non-repudiation, forward secrecy and public verifiability. Signcryption is a new paradigm that fulfils the functions of signature and encryption in one logical step and therefore the cost in terms of computation is reduced compared to the scheme signature-then-encryption. Till today many Remote user authentication schemes based on Elliptic curve, El-Gamal and RSA have been proposed. This paper points out the limitations of existing scheme such as lack in security aspects for example, password guessing attack, impersonation attack, forward secrecy, computation overhead, larger memory requirements. The proposed scheme based on signcryption using hyper elliptic curve that fulfils all the gaps of existing system.