Chun-Mei Chen

In order to remove the strong noise with complex shapes and high density in nuclear radiation scenes, a lightweight network composed of a Noise Learning Unit (NLU) and Texture Learning Unit (TLU) was designed. The NLU is bilinearly composed of a Multi-scale Kernel Module (MKM) and a Residual Module (RM), which learn non-local information and high-level features, respectively. Both the MKM and RM have receptive field blocks and attention blocks to enlarge receptive fields and enhance features. The TLU is at the bottom of the NLU and learns textures through an independent loss. The entire network adopts a Mish activation function and asymmetric convolutions to improve the overall performance. Compared with 12 denoising methods on our nuclear radiation dataset, the proposed method has the fewest model parameters, the highest quantitative metrics, and the best perceptual satisfaction, indicating its high denoising efficiency and rich texture retention.

Accurate and fast convergence to the optimal channel is a challenge in a cognitive radio sensor network (CRSN) when multiple cognitive wireless channels coexist. Some traditional wireless channel selection methods can be used to study the optimal channel selection. However, their convergence speed cannot meet the requirements because of vast computation and time accumulation. In this paper, a rapid channel selection strategy based on machine learning called MAB-CQ (multi-armed bandit-channel quality) is proposed. This strategy maps the channel selection problem to the improved multi-armed bandit (MAB) model. In the model, the second users (SUs) and the channels in the CRSN correspond to the players and the arms of MAB, respectively. The optimal channel is determined based on the UCB (upper confidence bound) of MAB-CQ for each player. In addition, the UCB equation is creatively defined to balance the exploration and exploitation problem. At the same time, to reduce the computation complexity, coefficients about the factors are used to narrow down the exploratory scope of our strategy. As a result, an accuracy optimal channel and a fast convergence speed are achieved by iterative execution of MAB-CQ. Extensive experimental results demonstrate that the MAB-CQ can converge to nearly 100% within the 10(5) time slots. By comparison, MAB-CQ has obvious advantages in cumulative rewards, computational complexity and convergence speed.

In this paper, we consider the multi-hop cooperative ad hoc networkswith dynamic network status, limited node energy, and fadingchannel environment.