Modern life is getting more complicated and people rely more on the intelligence embedded in their electronic gadgets. It is expected these gadgets will take larger roles in making, at least some simple, decisions instead of us. It is not difficult to see how many gadgets will be needed in an Internet of Things environment, or smart home settings, or any sort of connected devices. Interaction among these devices can be addressed using game theoretical models. However for a large number of devices interacting/playing with each other, the classical game models can be complicated. One way to approach this problem is by using evolutionary game theory (EGT). Evolutionary games deal with large number of players by making assumptions such as some common similarities in the players' interests, payoffs, and bounded rationality. Both of these assumptions seem to fit in modeling the large number of players'/devices' interaction. On the other hand, evolutionary games can model the user behavior in taking decisions when repeatedly played. Meaning that, each time a player does a move, the player observes the payoff and can compare it with the average payoff, and in the next play the player can choose a different move if it gives higher payoff and so on so forth. By using the concept of replicator dynamics, evolutionary games make it possible to observe how the choice dynamics is made. It can be looked at as learning until reaching to a very stable choice which is an evolutionary stable choice. This thesis first presents the problem of communications under a denial of service attack through a jamming threat. We consider the problem where the players try to communicate with a base station under the threat of jammers who, possible cooperatively, try to block their communications. The users have the option to work cooperatively too. The second problem this thesis deals with a generalized network model known as ephemeral network under the threat of a malicious attack with the absence of any central authority. The only control to the network is a set of rules which are agreed upon before setting a connection. Thirdly, we study the problem of advanced persistent threats (APTs), which is the problem of a powerful and stealthy attacker who wants to infiltrate the system. Evolutionary game theory is used by giving the players, the APT attacker and the system defender, the opportunity to adapt their decisions according to the replicator dynamics to reach to the robust decision,i.e, to choose the defend/attack strategy.The final part of this work uses evolutionary game theory to model the coexistence between WiFi and LTE-U technologies. We consider a scenario where there are two heterogeneous populations, one population represents the set of LET-U APs and the other one represents the set of WiFi AP. Furthermore, we assume that AP's belong to the same population do not interfere with each other. We study, under a given set of transmission strategies, the stability of the strategies that can appear in such a conflict. We specify the conditions under which, a coexistence with minimal interference can be established.