Master of Science By Research

Permanent URI for this collectionhttp://192.248.9.226/handle/123/18682

Browse

Browsing Master of Science By Research by Author "Kulasekere, EC"

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • Thumbnail Image
    item: Thesis-Abstract
    An Electrical parametric model of human skin and blood glucose spectroscopy
    (2014-06-05) Wanasinghe, TR; Kulasekere, EC
    Diabetes is well known as a leading cause of death all around the world. Mainly, invasive methods are used for blood glucose monitoring in the current context. The monitoring is done either as an inpatient procedures or using home based measuring devices. Invasive or minimally invasive methods make it di_cult when it comes to frequent measurements required by diabetes patients. It also has other issues such as the associated pain, phobia, and the spread of diseases like AIDS. These issues are heightened in the case of home based monitoring devices. As a result many researchers have attempted to introduce non-invasive measuring techniques for home based glucose monitoring devices. However none of then have met the accuracy requirements for medical use. Dielectric spectroscopy (DS) is one such methods which has been proposed for non-invasive glycaemia monitoring. In DS, the variation of skin impedance has been used to derive an index representing blood glucose uctuation. As a re- sult of the lack of knowledge of the impedance characteristics of the skin and the tissue underneath, and its relation to the level of blood glucose, the consistency and accuracy of the measurements are questionable. The ensuing research pro- poses a theoretical framework for skin impedance variations with the blood glucose level and also provides experimental veri_cation of the same. This research also proposes an electrical parametric (impedance) model for human skin and blood glucose spectroscopy which consists of human skin, electrode-electrolyte interface and coupling capacitance between transmitter and receiver. Such a mathematical model of the physiological system will enable us to further analyze the relationship the physiological parameters have with the uctuation of the blood glucose levels for di_erent individuals. Moreover, the thesis analyzes the inuence from bio-sensor to sensitivity mea- surements and proposes a concentric annular ring slot antenna (CARSA) as a possible sensor for non-invasive blood glucose measurement via DS. Compared to early research of Cada_ et al. [1], CARSA showed a 13 fold increment of the measurement sensitivity. Further, it could be seen that, this sensitivity increment was 40 fold when the e_ective length of CARSA decreases from 10 cm to 6.5 cm. The thesis further highlights the importance of careful design of this sensor and proposes a rigorous mathematical model of its derivation.
  • Thumbnail Image
    item: Thesis-Abstract
    Intelligent Control of Distributed Decision Agents
    (2016-05-25) Jayakody, SA; Kulasekere, EC
    Advances in technology have enabled the manufacturing of massive numbers of deployable computing agents with integrated sensors and actuators. Networked multiple distributed agents in a remote environment will enable distinct event sensing, and information dissemination. Such a collection of deployed agents can perform as a distributed micro sensor network, which cooperates to solve at least one common application. The basic building block of such a network is its deployable agents, and those are considered to be autonomous, unreliable, and irregular in orientation. The interconnections are unknown and assembled in an ad hoc manner. Hence, intelligent control and expected dynamics in the system present unique challengers in the system design. This research presents an approach to organize an unstructured collection of autonomous agents into a cooperative sensor network spontaneously. Furthermore, intelligent control is achieved through instantly populated set of searching agents, followed by natural biological Ant systems. A set of independent searching agents called ants cooperate to find distinct sensor-events with the shortest possible routes concurrently. Ants cooperate using an indirect form of communication mediated by pheromone. Ants update pheromone on the edges of the network as local variables while they are in parallel search. This allows multiple users to sense distinct events simultaneously. Overall design minimizes total energy consumption and allows selfconfiguring, robust, and scalable sensor network design. Proposed framework simplifies coordination overhead of the network and facilitates the implementation of efficient, adaptive Ant based algorithm..