T. Abdul Razak ¹, G. Najeeb Ahmed ²

Affiliations
1 Research Department of Computer Science, Jamal Mohammed College, Tiruchirapalli - 620 024, Tamil Nadu, IN
2 Research and Development, Bharathiar University, Coimbatore - 641046, Tamil Nadu, IN

Abstract

Data mining and machine learning techniques help us to better and deeper understanding of collected data. Meta-learning techniques extend this concept by providing methods for knowledge discovery process automatization. Meta-learning introduces various interesting concepts, including data meta-features, meta-knowledge, algorithm recommendation systems, autonomous process builders, etc. All these techniques aim to improve usually expensive and demanding data mining analysis. This paper focus on general overview of basic data mining, machine learning and meta-learning techniques, while focusing on state-of-the-art, basic formalisms and principles, interesting applications and possible future development in the field of meta-learning.

Keywords

Data Mining, DMA, KDD, Machine Learning, Meta-Learning

Full Text

   

J. Anthoniraj ¹, T. Abdul Razak ²

Affiliations
1 Bharathidasan University, Trichy - 620024, Tamil Nadu, IN
2 Jamal Mohamed College, Trichy - 620020, Tamil Nadu, IN

Abstract

Wireless Sensor Network consists of independent sensor nodes that are responsible for monitoring physical conditions. WSN is deployed in hostile environments. So it is vulnerable to capture and compromise by an attacker. An adversary can launch various types of attacks on WSN that can be classified as layer-dependent attack and layer-independent attacks. The layer-independent attacks are Sybil attack and Clone attack. In Clone attack an adversary can capture a sensor and creates clone of a captured node. These clone nodes are deployed in network area. It is difficult to detect clone nodes in the sensor network. There are so many clone detection protocols are available in static and mobile sensor networks. It is more challenging to detect clone attack in mobile WSN compared to the static WSN. The proposed clone attack detection protocol called as Cluster Based Clone Attack Detection (CBCD) discovers the clone nodes available in Mobile WSN. In this protocol sensor network divided into clusters. All the clusters have a cluster head and random number of sensors nodes. This protocol detect clone node according to the movement of sensor nodes. The clone is identified when sensor node move within the cluster or other clusters in the sensor network. Theoretical analysis and simulations have been conducted to evaluate the protocol in terms of clone detection time, clone detection ratio, memory consumption.

Keywords

Clone Attack, Cluster based Clone Detection, Wireless Sensor Network.

Full Text