Author: Maria Rene Arandia Jimenez Major: Industrial Engineering 

My research is about optimizing the design of heterogeneous wireless sensor networks (WSN) within security applications. I have been doing this research under the mentorship of Dr. Kelly M. Sullivan, who is a professor in the industrial engineering department.  

A WSN is a collection of sensor nodes that are scattered over an area. These sensors detect spatial changes that are collected as data, and this data, is stored and forwarded to a user. WSN have several applications, such as in healthcare and weather monitoring, and are also used for security applications, mainly to spot and detain threats moving into a certain region. Sensors within a WSN all have the same radii coverage, which is the radius that is monitored by each sensor node. This radius also describes the range in which a sensor can forward data to sensors nearby. My goal is to maximize network coverage and increase connectivity efficiency of WSNs. One of my proposals is to consider heterogeneous WSN, which have sensor nodes with different sensing radii and can detect several security threats. Hopefully, this will also decrease the expenditure on sensor nodes.  

I identified two important parameters, which are the number of sensors Type A and Type B, and their corresponding radii, as it is a heterogeneous network. This will help me determine the overall coverage of the network when I create a Multiple Objective Optimization Model (MOOM). 

Before performing a MOOM, I began by simulating WSNs a in MATLAB. Dr. Sullivan guided me on how to write code in MATLAB, because I did not have previous experience with this software. First, I started by graphing random points in a 1×1 area, and I then established a radius for all of them. As WSNs are usually deployed randomly in remote areas, it made sense to have the sensors graphed randomly. Dr. Sullivan suggested that I should also graph sensor deterministically, by using different lattice structures as how sensors should be arranged. Then, to test coverage, I graphed target nodes, which represent a point in space that can be sensed by the WSN. In the future, this will help me evaluate different sensor arrangements, because the more target nodes the proposed WSNs cover, the better sensor arrangement they will have. 

Dr. Sullivan and I were also very curious to know more in depth how sensor nodes work, what they are made of, and more technical facts about them, so we reached out to some electrical engineering professors. Dr. Sullivan does not want the research to go off on a tangent with this information we will receive, but he also encouraged me to learn about how the sensors work, for general knowledge in this research. 

Now that I gained coding skills in MATLAB, the next step is to test many possible sensor arrangements. I will create an evaluating criterion that will take into account the sensors Type A and Type B coverage, WSN overall coverage, sensor connectivity, and number of sensors. This criterion will help me determine what sensor arrangements are the best, and which one is the optimal. I will also create the aforementioned MOOM, which will give me a theoretical answer. I will compare what I tested to the MOOM suggestion and analyze the differences.