On Euler Tours in Streaming Models and some Games on Graphs
In this thesis, we take a look at several graph theoretical problems. We present two streaming algorithms for finding an Euler tour in a graph, prove a tight bound on the capture time in the Bridge-Burning Cops and Robbers Game and solve an open problem for the Extreme Vertex Destruction Model. In Chapter 2, we consider the classical Euler tour problem and take a modern look at this problem in the context of the graph streaming model. Here, RAM is of size O(n polylog(n)), where n is the number of nodes, and the graph is given as a stream of its edges. With this restricted memory space, we give a one-pass algorithm for finding Euler tours in the graph streaming model. In Chapter 3, we regard a lesser-known streaming model, the so-called StrSort model, to tackle a downside of our algorithm mentioned above. The algorithm stores an Euler tour on an output tape in form of a successor function. The order of the edges is given, but the edges are not actually sorted in the order of the Euler tour. Therefore, further processing the tour with another streaming algorithm might become difficult. We give an algorithm for sorting the edges of a graph according to a found Euler tour, that has a preparation step in the graph streaming model and a processing step in the StrSort model. The so-called Bridge-Burning Cops and Robbers Game is the topic of Chapter 4. Here, every time the robber traverses an edge, this edge is deleted afterwards. We study winning strategies of a single cop and make statements on the maximum number of turns of such strategies. In Chapter 5, we study networks formed by selfish agents. When a node is ‘destroyed’, i.e. the adjacent edges are deleted, the network is damaged and some players lose the connection to each other. In the Extreme Vertex Destruction Model, we observe the impact of such a deletion on swap equilibrium graphs and make statements on the maximum amount of damage that can cause here.