Due to resource constraints and severe conditions, wireless sensor networks should be self-adaptive to maintain certain desirable properties, such as energy efficiency and fault tolerance. In this paper, we design a practical utility function that can effectively balance transmit power, residual energy, and network connectivity, and then we investigate a topology control game model based on non-cooperative game theory. The theoretical analysis shows that the topology game model is a potential game and can converge to a state of the Nash equilibrium. Based on this model, an energy-efficient and fault-tolerant topology control game algorithm, EFTCG, is proposed to adaptively constructs a network topology. In turn, we present two subalgorithms: EFTCG-1 and EFTCG-2. The former just guarantees network single connectivity, but the latter can guarantee network biconnectivity. We evaluate the energy-efficient effect of EFTCG-1. Meanwhile, we also analyze the fault-tolerant performance of EFTCG-2. The simulation results verify the validity of the utility function. EFTCG-1 can efficiently prolong the network lifetime compared with other game-based algorithms, and EFTCG-2 performs better in robustness, although does not significantly reduce the network lifetime.
Device to device (D2D) communication has recently been established in the literature as an effective means to increase the frequency spectrum and enhance the efficiency of energy consumption in future cellular systems. However, certain issues, resulting from reusing resources in the same cell, have caused serious perturbations. We study issues pertaining to D2D communication, such as dual mode selection, channel allocation and power control, aiming at the maximization of the overall throughput of the system, while at the same time ensuring that the generated interference is kept minimized. This is an NP-Hard problem that decomposes the optimization problem into two layers: the inner layer, where the DQN algorithm is used as an indicator of the optimal transmission power that should be allocated to the D2D pairs in accordance with their mode of operation, and the outer layer, where strategic decisions, such as which communication mode to use and how to allocate the channels, are made. We have proved the superiority of the proposed scheme, in terms of both system throughput and performance, through simulating experiments involving different scenarios.
When it comes to selecting an optimal defense strategy for an intrusion detection system of a wireless sensor network, such challenges as those brought about by the diversification of the attack methods and the expanded database of the attack patterns have to be dealt with. To overcome those challenges, this paper combines realistic bounded rationality with the incomplete information of the attack-defense players by employing evolutionary game theory as a tool. Firstly, an attack-defense evolutionary game model considering three types of population, in which attackers are subdivided by the source of the threat into external attackers and selfish nodes, is proposed. The sets of player types and the game strategies in our model can be extended from 2× 2 to n× m × l. The sensitivity of the evolutionary population to similar strategies, which reflects the efficiency change in the multi-agent learning process, is depicted by a replicator dynamic equation especially optimized for this purpose by the introduction of an enhanced cooperation mechanism. In essence, an optimal defense strategy selection algorithm is provided by calculating the evolutionary stable equilibrium and a description of the evolutionary trajectory of the players over time is obtained. Moderate security and proactive defense in the form of support decisions have been provided by our method for wireless sensor networks. Experimental results have verified the validity of our method. Moreover, the optimized algorithm has solved the problem that an excessively large database of attack patterns affects the speed of switching to the optimal defense decision strategy and the learning efficiency of the evolutionary game replicator dynamic mechanism is not fast enough.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.