Fog computing emerged to meet to the needs of modern IoT applications, such as low latency, high security, etc. To this end, it brings the network resources closer to the end user. The properties of fog computing, such as heterogeneity, distribution, and resource limitations, have challenged application deployment in this environment. Smart service placement means deploying services of the IoT applications on fog nodes in a way that their service quality requirements are met and fog resources are used effectively. This paper proposes an efficient application deployment method in fog computing using communities. In contrast to previous research, the proposed method uses more factors than topological features to distribute network capacity more evenly between communities. This results in efficient use of network resources and better fulfillment of application requirements. In addition, according to our argument, using multiple criteria to prioritize applications will lead to better deployment and more effective use of resources. For this purpose, we use the number of application requests besides the deadline factor for application prioritization. Extensive simulation results showed that the proposed method significantly outperforms the state-of-the-art methods in terms of meeting deadlines, decreasing delays, increasing resource utilization, and availability by about 17, 33, 7, and 11 percent, respectively.

EVOLUTIONARY game theory has been known as the most suitable tool for modeling the dynamics of strategic interactions. In this regard, EVOLUTIONARY algorithms present the new approach for learning and decision modeling of bounded rationality factors. The objective of this study is providing the new model of searching for optimal strategies in iterated prisoner's dilemma (IPD) using genetic algorithm. For this purpose, by simulating competition between Iran and Saudi Arabia in OPEC oil Coalition, we used 12 strategy types over 20 runs of genetic algorithm for maximizing individual’s scores and also minimizing competitor fitness scores. Results show that “Tit for Tat” with the highest average fitness in both competitions known as the optimal strategy. The other strategis like; Soft Majority, Trigger & TF2T are next in ranking. The strategy “All D” is known as inefficient strategy in competition with the lowest productivity.

EVOLUTIONARY altruism has been a serious problem for EVOLUTIONARY theory from its earliest times at mid-nineteen century. Group selection theory was its solution until 1960s.But at that time it was criticized so that it was no longer the solution to that problem. Since then theories put forward to solve the problem by individual and/or gene selection theories. Kin selection, reciprocal altruism, EVOLUTIONARY game theory, the selfish gene, and indirect reciprocity are the main theories of this sort. Altruism according to these theories is apparent not real. Multi-level selection theory is introduced since 1970s that defends "new" group selection theory. In addition cultural group selection theory has come into the field. Altruism according to these theories is real. I agree with the second type of theories that real altruism has been evolved. But perhaps more in agreement with the first type I see it in the "shadow" of EVOLUTIONARY egoism.