Cognitive radio (CR) is an intelligent wireless communication system and the core of it is the cognitive engine. Cognitive engine is expected to implement cognitive learning, inference, decision-making through the artificial intelligence technology to decide a specific radio configuration (i.e. carrier frequency, modulation type, power, etc.) according to the changing of environment. In this paper, a cognitive radio learning inference and decision-making engine based on Bayesian network (BN) is proposed to obtain the optimum configuration rules adapt to the variation of the environment with the learning and inference algorithm of Bayesian network. Simulation results show the feasibility and validity of modeling the cognitive learning inference and decision-making engine with Bayesian network.

Cognitive radio technology is proposed to solve the problem of the scarce radio spectrum resources. In view of the spectrum allocation in cognitive radio technology, this paper first analyses the three allocation rules based on graph coloring theory which include the sum reward rules, the max-min reward rules and the proportional-fair reward rules. The simulations of the three algorithms adopt the average user-reward. And we can gain the optimal rule among the three rules by the comparing simulation. This paper also proposes an improved algorithm based on graph coloring theory. The simulation results indicate that the proposed algorithm can reduce the computation effectively and have no effect on the system reward.

No abstract

Deep subspace clustering methods have achieved impressive clustering performance compared with other clustering algorithms. However, most existing methods suffer from the following problems: 1) they only consider the global features but neglect the local features in subspace self-expressiveness learning; 2) they neglect the discriminative information of each self-expressiveness coefficient matrix; 3) they ignore the useful long-range dependencies and positional information in feature representation learning. To solve these problems, in this paper, we propose a novel multi-scale deep subspace clustering with discriminative learning (MDSCDL) to obtain a high-quality self-expressiveness coefficient matrix. Specifically, MDSCDL bridges multiple fully-connection layers between encoder and decoder to learn multi-scale self-expressiveness coefficient matrices from global and local features, representing the more comprehensive relationship among data. By modeling the interdependencies of the multi-scale selfexpressiveness coefficient matrices, MDSCDL adaptively assigns discriminative weights for each matrix and fuses them with convolution operation. Moreover, to increase representation power, MDSCDL introduces the coordinate attention mechanism to extract the long-range dependencies and positional features for subspace self-expressiveness learning. Extensive experiments on the face and object datasets have shown the superiority of MDSCDL compared with several state-of-the-art methods.

Aiming at the supplier selection problem where the decision information is an interval intuitionistic fuzzy number and completely does know the attribute and decision maker's weight, this problem is reduced to a multi-attribute group decision problem. A decision method based on information entropy to determine the decision-maker's weight and the deviation maximization method to determine the attribute weight is proposed. Finally, this method is applied to the selection of automotive parts suppliers, and the results are compared with the relevant methods, which fully illustrates the effectiveness of this method.