Modeling the Wi-Fi Network in the Statler Auditorium
Three times a week, over six-hundred Cornell students pile into the Statler auditorium and get ready to play a game of their own; getting a good wi-fi connection. For the sake of simplification, we can boil down a player’s role in the Statler Auditorium Wi-Fi network into two nodes: the device and the modem for the internet. A player chooses between two different modems to connect to. Either the student can connect to the default Eduroam network, or the student can presume the identity of a guest and connect to the Cornell Visitor network. In the image below, I modeled the possibilities for a Networks student who is struggling to use the internet in the Statler auditorium with an undirected graph pointing to each possible node.
The Mbps (megabits per second) speed of the internet is inversely proportional to the traffic x / 200, which depends on the number of students (x) connected to each wi-fi network. Suppose that 600 students gradually fill up the Statler auditorium. By experience, I know that a device tends to default to the eduroam network, so everyone’s devices become a part of route E. Route E becomes quite crowded, with a traffic value of 600 / 200 or greater. The internet is now astronomically slow for everyone connected to it. During my first lecture, I desperately tried to connect to Canvas in a vain attempt to download the day’s lecture slides. I realized that perhaps I could connect to Cornell Visitor instead, and I was then able to use the internet. Let us assume that I was the only student who thought of this at the time. As a result, route G initially had a traffic of 1/200. The minimal traffic meant much faster wi-fi for me. However, by the next lecture, a lot of students had discovered the alternate strategy. If half of the students in class connected to Cornell Visitor, both Routes E and G had traffic of 300 / 200. As a result, both routes became congested, and everyone experienced the same astronomically slow speeds. Since every student is playing in a self-interested manner with no awareness to how other players are acting, the traffic for both routes tends to become high. Even if equilibrium is reached, the inadequate routers prevent there from being any winners. Only outside interference, such as professors telling the students to disconnect devices they do not need, can lead to reduced traffic.
Picture Sources:
http://ntgr.co.in/home/products/networking/wifi-routers/R6260.aspx
https://www.pinterest.com/pin/466755948860807846/
Text Sources:
https://www.cs.cornell.edu/home/kleinber/networks-book/networks-book-ch08.pdf