Degree Type

Creative Component

Semester of Graduation

Fall 2018

Department

Electrical and Computer Engineering

First Major Professor

Dr. Ahmed Kamal

Degree(s)

Master of Science (MS)

Major(s)

Computer Engineering

Abstract

Increase in the usage Internet of Things has driven lot of importance to wireless sensor networks. Wireless sensor network consists of sensor nodes with low power and low transmission range. Sensor power is the crucial part because if the power goes down, the sensors die out and will not be available for communication. This project deals with sensor nodes which are deployed in an area and there is an external source available for harvesting power. The harvested energy keeps the sensor nodes powered and the communication in the network can sustain for longer time.

This work talks about the simulation methods of a multi-agent reinforcement algorithm that aims at minimizing the energy consumption in a wireless sensor network. The simulation is performed in Cooja simulator of Contiki Operating System. Contiki OS is an open source Real Time Operating System (RTOS) for low power wireless devices. Internet of Things has many features including power awareness which is providing mechanisms for the estimation of power consumption in the network. Developing and debugging large wireless networks is really difficult and this is made much easier by providing environment to develop applications on fully emulated devices using Cooja, the Contiki network simulator.

The simulation is carried out for two scenarios on a network where are two sensor nodes and one sink node. The results are studied for both the cases and seen that overall energy consumption on the network is reduced for the case where the q-learning algorithm is used.

Copyright Owner

Naregudam, Nirupana

File Format

Word

Share

COinS