Ankush Rai

Quasi Opportunistic Supercomputing for Geospatial Socially Networked Mobile Devices

due to several reasons such as attentions given to design features and to modification from littler to medium sized system that results in trivial enhancement in performance.In spite of the tremendous achievement of grids which run parallel jobs, there still exists demand of requesting applications to avail opportunistic computing for considering the right selection of prerogative supercomputers. The objective of this research is to build a framework for on the fly computational commencing on socially convolved terrestrial grids to facilitate the sharing of computing workload over wireless networks as virtual supercomputers of unprecedented power. This will assist computing in heterogeneously interconnected mobile assets, Internet of Things, in crowdedscenarios for variable ranges like: stadiums, shopping centers, and so forth.

Membrane Computing Based Scalable Distributed Learning and Collaborative Decision Making for Cyber Physical Systems

Machine to Machine based communication and coordination in cyber physical systems for manufacturing application is a daunting problem. Here, the difficulty is in developing a multi task learning algorithm for finding collaborative key points within high dimensional decision space. In order to overcome such a problem a gradient based method is required to enable distributed control and collaborative decision making. This paper presents membrane computing based gradient method for online multitask learning and controlling the distributed coordination to achieve stochastic planning to formulate optimize semantic network to derive a local optimal policy. Experimental verification is achieved through implementation of several concurrent functioning of mobile robotic platforms.

Multi-user networked framework for virtual reality platform

Networked Virtual Environments (NEV) are sets of disjointed virtual worlds which are topographically set apart from the users yet connected through the communication network to give them the illusion of being in the same virtual location. NVEs are increasingly receiving consideration from business and research point of view. Contrasted with various sorts of possible applications, NVEs require a high responsiveness to ensure continuous streaming of live data for full immersion. It is as yet difficult to develop such applications since it require to create virtual reality framework over distributed databases; as every user needs locally adequate data flow to concurrently construct a scene as with other users and able to accommodate multi-user interactions in the virtual environment. Such information requires simultaneous computation and communication of streaming data. In this study we built up a prototype of a NVE in view of the blend of Minority Game (MG) and Online Induction Algorithm (OIA) with lossless transfer functions.

COMPUTATIONAL MODELLING AND PREDICTING SPREAD OF ARBOREAL EPIDEMIC

This study presents a modelling solution for the arboreal epidemic like Huanglongbing. Usually, the spread of such plant disease is modelled based upon the four parameters such as: susceptibility, exposure, infectiousness, detection and removed but such a model is deprived from the time as a dimension to model such variations. Owing to this the time for which infection, exposure, detection and removal time is censored form modelling studies of disease spread through heterogeneous plant species. The research presented in this work characterize such heterogeneous transmission with integration of temporal spatial modelling of latent period of season and effects on the host, infection period and dispersal parameters corresponding to the host age. The outcome form this research will enable to control arboreal epidemic.