Publication

Abstract : Intelligent and ubiquitous vehicular transit has become a necessity of the 21st century. A satisfactory solution seems to fixate around the development of self–-learning vehicles, with multiple interfaces providing the relevant parameters for the above stated learning. Therefore, any labyrinthine vehicle should also be supported with contemporary computational and communication technologies (Internet of Things, Wireless Sensor Networks) for maximum goodput. With supernumerary sensors being embedded within automobiles, an efficient and flexible paradigm is strongly desired. This study proposes the Multi-Level Asymmetric Nodal Interoperability protocol for usage in the latest generation of intelligent automobiles. Categorically, this paper is an attempt at examining novel and innovative ways for designing a hierarchical network model and a subsequent data link layer protocol data unit (frame) for cost-effective, reliable and fault tolerant transmission of data in intra-vehicular network. A machine learning based environment-adaptive method to detect link fault has also been proposed in the frame work of the network model using self-organizing maps. Furthermore, the proposed network protocol and supplementary algorithms have been proved capable of minimizing network overhead by 80% during transmission of Boolean values in a real-time environment when compared to IEEE 802.3 wired Ethernet standard.