The inherent selfishness of each node for the enhancement of message successful delivery ratio and the network overall performance improvement are reflected in the contradiction relationship of competition and cooperation in delay/disruption tolerant networks (DTN). In particular, the existence of malicious node aggravates this contradiction. To resolve this contradiction, social relationship theory and group theory of social psychology were adopted to do an in-depth analysis. The concrete balancing approach which leveraged Nash equilibrium theory of game theory was proposed to resolve this contradiction in reality. Thus, a new congestion control routing algorithm for security defense based on social psychology and game theory (CRSG) was put forward. Through the experiment, this algorithm proves that it can enhance the message successful delivery ratio by more than 15% and reduce the congestion ratio over 15% as well. This algorithm balances the contradiction relationship between the two key performance targets and made all nodes exhibit strong cooperation relationship in DTN.
We consider a wide range of non-convex regularized minimization problems, where the non-convex regularization term is composite with a linear function engaged in sparse learning. Recent theoretical investigations have demonstrated their superiority over their convex counterparts. The computational challenge lies in the fact that the proximal mapping associated with non-convex regularization is not easily obtained due to the imposed linear composition. Fortunately, the problem structure allows one to introduce an auxiliary variable and reformulate it as an optimization problem with linear constraints, which can be solved using the Linearized Alternating Direction Method of Multipliers(LADMM). Despite the success of LADMM in practice, it remains unknown whether LADMM is convergent in solving such non-convex compositely regularized optimizations. In this research, we first present a detailed convergence analysis of the LADMM algorithm for solving a non-convex compositely regularized optimization problem with a large class of non-convex penalties. Furthermore, we propose an Adaptive LADMM(Ada LADMM) algorithm with a line-search criterion. Experimental results on different genres of datasets validate the efficacy of the proposed algorithm.
