Technological advances have led to the emergence of wireless sensor nodes in wireless networks. Sensor nodes are usually battery powered and hence have strict energy constraints. As a result, energy conservation is very important in the wireless sensor network protocol design and the limited power resources are the biggest challenge in wireless network channels. Link adaptation techniques improve the link quality by adjusting medium access control (MAC) parameters such as frame size, data rate, and sleep time, thereby improving energy efficiency. In this paper we present an adaptive packet size strategy for energy efficient wireless sensor networks. The main goal is to reduce power consumption and extend the whole network life. In order to achieve this goal, the paper introduces the concept of a bounded MAB to find the optimal packet size to transfer by formulating different packet sizes for different arms under the channel condition. At the same time, in achieve fast convergence, we consider the bandwidth evaluation according to ACK. The experiment shows that the packet size is adaptive when the channel quality changes and our algorithm can obtain the optimal packet size. We observe that the MAB packet size adaptation scheme achieves the best energy efficiency across the whole simulation duration in comparison with the fixed frame size scheme, the random packet size and the extended Kalman filter (EKF).

We present a label-free relative 3D surface reconstruction method via polarization-constrained ring illumination. A programmable 12-sector ring LED sequentially illuminates the transparent sample to generate direction-sensitive intensity responses, while a division-of-focal-plane polarization sensor records linear polarization components without mechanical rotation. The ring-illumination responses are used to estimate initial surface-gradient cues, and polarization features, including the degree and angle of linear polarization, are incorporated into a height-field optimization model to suppress unstable reconstruction in weak-texture and low-contrast regions. We validate the method on 20 µm standard microspheres and living yeast cells. The microsphere-calibrated model reduces full-region height RMSE from 2.16 µm to 0.45 µm. For a budding yeast cell, single-sector and overlapping paired-sector 12-direction encodings yield consistent relative-height maps ( r = 0.970; NRMSE = 6.03%), while polarization-weighted regularization reduces weak-interior high-frequency RMS by 10.8%.

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.

Structured illumination microscopy (SIM) offers advantages such as low phototoxicity and high imaging speed for live-cell super-resolution imaging. However, under low-light imaging conditions, parameters such as the spatial frequency, direction, and phase of the structured light are susceptible to external system interference, leading to estimation errors and amplified reconstruction artifacts. Although existing physics-driven and deep learning-based SIM reconstruction methods can significantly improve imaging quality under low signal-to-noise ratio conditions, deep learning networks—lacking the constraints of physical prior mechanisms—are prone to overfitting when subjected to noise interference. Meanwhile, physics-driven algorithms rely on accurate parameter estimates and exhibit insufficient reconstruction stability when test parameters deviate from training settings. To address this issue, this study proposes DART-SIM, a deep-learning-based SIM reconstruction method that is robust to structured-light parameters. Within a deep learning framework, this method establishes a physically guided mechanism with parameter robustness, reducing the model’s dependence on calibration parameters. Experimental results demonstrate that DART-SIM can effectively suppress reconstruction artifacts caused by parameter perturbations under low-light conditions, restore fine structures, and enhance the robustness and stability of SIM reconstruction.

Transparent cells often exhibit weak and spatially heterogeneous directional-intensity responses in non-interferometric microscopy, making morphology reconstruction vulnerable to low signal, incomplete angular information, and boundary artifacts. We present a polarization-reliability-adaptive soft-constrained ring-illumination method (PRAS-RIM) for reconstructing the apparent relative 3D morphology of transparent samples. The principal acquisition configuration comprises a programmable 12-sector ring source (P12), a PET diffuser, a fixed circular polarizer, a 40× objective with NA = 0.65, and a division-of-focal-plane polarization camera. Calibrated directional intensity provides the primary geometric evidence, whereas the angle of polarization (AoP) is represented in double-angle form and incorporated only as a pixelwise reliability-weighted soft constraint. Across 29 microsphere measurements, complete PRAS-RIM reduced the mean object-level RMSE from 2.77 ± 0.19 μm for the intensity-only baseline to 0.59 ± 0.07 μm. A yeast test set comprising 51 cells from three distinct static fields acquired within the same experiment demonstrated cross-field consistency and spatially varying reliability, but was not used for absolute-height validation because no independent geometric ground truth was available. These results support PRAS-RIM as an interpretable framework for robust label-free reconstruction of relative morphology without claiming optical phase, optical path length, absolute cell height, or a three-dimensional refractive-index distribution.