Explore the vast range of titles available.

Deploying neural networks on encrypted data requires efficient evaluation of nonlinear activations, especially the ReLU function, without decryption. While the CKKS homomorphic encryption scheme supports packed arithmetic over approximate numbers efficiently, its approximate semantics make direct nonlinear evaluation difficult, and polynomial surrogates often introduce approximation error and non-discrete outputs. In this work, we present a task-specific, non-interactive construction for discrete ReLU evaluation in CKKS by combining modulus-switch-based discretization with interpolation-driven lookup-table (LUT) evaluation. We instantiate this design in two complementary schemes. The first uses trigonometric Hermite interpolation and functional bootstrapping to compute a discrete sign indicator, which is then combined with the encrypted input through conditional multiplication to obtain the ReLU output; this variant is compact and suitable for lightweight settings. The second uses iterative most-significant-bit (MSB) bootstrapping to support larger plaintext moduli and higher-precision regimes through repeated digit extraction. A common enabler of both schemes is a discretization step that maps approximate CKKS plaintexts to a finite integer representation; exactness in our setting therefore refers to exact evaluation over this discretized representation, while the deviation from the original CKKS plaintext is governed by the discretization error analyzed in Lemma 1. Experiments on encrypted MNIST inference and the accompanying LUT/storage analysis indicate that the proposed schemes preserve competitive accuracy relative to polynomial-approximation baselines while maintaining manageable auxiliary storage under the reported parameter settings. These results suggest that interpolation-based discrete activation is a promising alternative to polynomial approximation for selected CKKS-based encrypted inference tasks.

This research investigates the potential of a game-theoretic-based Active Yaw Control (AYC) strategy to enhance power generation in wind farms. The proposed AYC strategy in this study replaces traditional look-up tables with a trained Artificial Neural Network (ANN) that determines the optimal yaw misalignment for turbines under time-varying atmospheric conditions. The study examines a hypothetical 3x2 rectangular arrangement of NREL 5-MW wind turbines. The FAST.Farm simulation tool, utilizing the dynamic wake meandering (DWM) model, is employed to assess both the power performance and structural load on the wind turbines. When tested with two different inflow directions and ambient turbulence (10%), the AYC strategy demonstrated a maximum increase in total power output of 2.6%, although it affected individual turbines differently. It also exhibits an increase in some structural loads, such as tower-top torque, while some components experience a slight reduction in load. The results underscore the effectiveness of the ANN-guided game-theoretic algorithm in improving wind farm power generation by mitigating the negative impact of wake interference, offering a scalable and efficient method for optimizing large-scale wind farm. However, it is essential to evaluate the overall impact of AYC on wind farm efficiency in terms of both Annual Energy Production (AEP) and structural loading under various atmospheric conditions.

This paper presents a solution to satisfy the increasing requirement of real-time secure image transmission over public networks. The main advantage of the proposed cryptosystem is high efficiency. The confusion and diffusion operations are both performed based on a lookup table. Therefore, the time-consuming floating point arithmetic in chaotic map iteration and quantization procedures of traditional chaos-based image cipher can be avoided. Besides, this cryptosystem possesses satisfactory resistance to noise perturbation and loss of cipher data, which are inevitable and unpredictable in real-world channels. The channel disturbance and the deliberate damage from the opponents are both tolerated. The recovered image from the damaged cipher data has satisfactory visual perception. Simulations prove the advantages of the proposed scheme, which render it a good candidate for real-time secure image applications.

The technology of reconfigurable intelligent surfaces (RISs) has been showing promising potential in a variety of applications relying on Beyond-5G networks. RIS can indeed provide fine channel flexibility to improve communication quality of service (QoS) or restore localization capabilities in challenging operating conditions, while conventional approaches fail (e.g., due to insufficient infrastructure, severe radio obstructions). In this paper, we tackle a general low-complexity approach for optimizing the precoders that control such reflective surfaces under hardware constraints. More specifically, it allows the approximation of any desired beam pattern using a pre-characterized lookup table of feasible complex reflection coefficients for each RIS element. The proposed method is first evaluated in terms of beam fidelity for several examples of RIS hardware prototypes. Then, by means of a theoretical bounds analysis, we examine the impact of RIS beams approximation on the performance of near-field downlink positioning in non-line-of-sight conditions, while considering several RIS phase profiles (including directional, random and localization-optimal designs). Simulation results in a canonical scenario illustrate how the introduced RIS profile optimization scheme can reliably produce the desired RIS beams under realistic hardware limitations. They also highlight its sensitivity to both the underlying hardware characteristics and the required beam kinds in relation to the specificity of RIS-aided localization applications.

This work aims to alleviate the memory requirements of the recent wake engineering model described in Criado Risco et al. [1]. The original model relies on a RANS-based look-up table of three-dimensional velocity deficit and added turbulence intensity fields computed for a stand-alone turbine under a wide variety of conditions. The objective is to develop an alternative to the model of Criado Risco et al. [1], particularly in terms of added turbulence intensity, for which little research has been carried out to date. To achieve this, a one-dimensional analytical expression is fitted to the look-up table and generalized to higher dimensions. The turbulence intensity model is then coupled to a velocity deficit model and implemented in PyWake, an open-source wake engineering software. Overall, the new turbulence intensity model is found to provide a reliable description of the RANS look-up table data while reducing by half the memory requirements of the original model. This conclusion is extended to multiple wake situations, for which this work also establishes a direct link between the adequate superposition method and the definition chosen to describe the added turbulence intensity in the wake.

In this study, we present a method of chroma enhancement in the CIELAB color space and compare it with that in the RGB color space. Color image enhancement using the CIELAB color space has the disadvantage that the color gamut problem occurs because the conversion to the RGB color space is necessary to display the image. However, since the CIELAB color space is based on human visual perception, the quality of the resulting images is expected to be higher than that of the RGB color space. In the method using the CIELAB color space, we introduce a lookup table to reduce the calculation costs. Experiments comparing image enhancement results obtained from two color spaces are performed using several digital images.

This paper proposes CML-Net, a novel collaborative multi-lookup table network, tailored for real-time enhancement of severely degraded low-light images. By introducing a cascade of 1D and 4D lookup tables within a single channel, CML-Net expands the receptive field and enhances the ability to process local pixel information. A lightweight global enhancement module utilizing parallel Vision State-Space Modules is designed for fast global information extraction, providing adaptive gamma and color correction parameters. Experimental results demonstrate that CML-Net outperforms state-of-the-art methods, achieving an average rank of 2.2 and 1.8 on full-reference and non-reference datasets, respectively, while maintaining real-time processing capabilities. Deployment tests on mobile devices showcase its potential for edge device applications.

Bilateral filtering is a widely used image smoothing filter that preserves image edges while also smoothing texture. In previous research, the focus of improving the bilateral filter has primarily been on constructing an adaptive range kernel. However, recent research has shown that even slight noise perturbations can prevent the bilateral filter from effectively preserving image edges. To address this issue, we employ a neural network to learn the kernel parameters that can effectively counteract noise perturbations. Additionally, to enhance the adaptability of the learned kernel parameters to the local edge features of the image, we utilize the edge-sensitive indexing method to construct kernel parameter lookup tables (LUTs). During testing, we determine the appropriate spatial kernel and range kernel parameters for each pixel using a lookup table and interpolation. This allows us to effectively smooth the image in the presence of noise perturbation. In this paper, we conducted comparative experiments on several datasets to verify that the proposed method outperforms existing bilateral filtering methods in preserving image structure, removing image texture, and resisting slight noise perturbations. The code is available at https://github.com/FightingSrain/AdaBFLUT .

Noise contaminates and distorts measurements from wireless sensor networks (WSNs). The sensor node's computations and energy consumption are increased due to the noise in the signal, resulting in the sensor node's shorter lifespan. Therefore, efficient design is required to achieve noise reduction. The finite impulse response (FIR) filter is a significant part of signal preprocessing in the WSN to remove noise. Signal preprocessing could assist in minimizing the amount of energy used in communication between nodes while also improving data transmission efficiency. The multiplier block in an FIR filter involves the generation and addition of partial products (PP), which consumes a large area and enormous power. The most common adaptive filtering technique utilizes the least mean squares (LMS) algorithms. The multiply and accumulate (MAC) process is the spine of LMS adaptive filters. Using multiple MAC units can boost the system's speed, but the cost of the system rises as the multipliers take up a lot of space and consume more energy. When LMS is implemented as a completely reliable architecture on a hardware platform, the multiplier evolves into a bottleneck for higher-order filters, resulting in significant size, expenditure, and energy needs, rendering the design unsuitable for practical implementation. FIR filters are frequently realized using distributed arithmetic (DA) based scheme. DA-based designs replace multipliers with look-up tables (LUT), where the precomputed PPs is saved. LUTs are required for weight update and filtering processes. The size of the LUT rises exponentially with the increasing order of filters. To minimize the size of a LUT, offset binary coding (OBC) is utilized, resulting in smaller memory size. The proposed architecture is based on DA in which PPs of filter coefficients are pre-calculated, and these coefficients are saved in LUTs, and through the use of OBC, the weighted coefficients are updated. The proposed method does not decompose the LUT into two smaller LUT to achieve area and energy efficiency. Typically, each iteration requires recalculating all of the LUT's address locations. This research presents a novel method that iteratively updates the contents of LUT without rotating the address, making the FIR filter more energy efficient. Random access memory-based LUT is utilized to eliminate the physical address rotation. A method for reducing LUT access time is proposed. The proposed design significantly reduces area and energy consumption compared to the existing systems. Also, it saves a large amount of sliced LUT and flip-flops compared to conventional methods. Therefore, high-performance adaptive filtering applications like IoT-based WSN can benefit the most from the suggested approach.