A Data Management Framework for Mobile Ad hoc Network (MANET)

Abstract

Mobile Ad hoc Network (MANET) is defined as an infrastructure-less network created spontaneously among a set of nodes. Data management is an issue of prime concern in MANET because of the massive amount of data exchanges among hosts of the network. Data management requires novel mechanisms to confront with issues of data discovery, data representation and cache management etc. To address them, this thesis has proposed a framework called Association Rules Based Network Layer Data and Information Management System for Mobile Ad hoc Network (ALADIN). The novelty of ALADIN lies in utilization of cross-layer design, correlation among data items and semantic data representation. The cross-layer information is utilized for data discovery and maintaining consistency of cached data. The correlation among data items is exploited for cache replacement and enhancing the performance of discovery phase. A multilevel ontology is utilized for ensuring the semantic interoperability among nodes in scalable fashion. The framework has been implemented in JIST/SWANS simulator. A set of experiments with different network settings and mobility models has been done to evaluate the framework. The results show the improvement in hit ratio, latency and scalability of the data management operations with the proposed framework.