Adaptive-Slope Squashing-Function-Based ANN for CSI Estimation and Symbol Detection in SFBC-OFDM System

Kapoor, Divneet Singh; Kohli, Amit Kumar

doi:10.1007/s13369-020-05207-w

Cited by 5 publications

(5 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The moving robot is required to explore the area, following the rules of social engagement with the humans, while simultaneously constructing the map. A modified version of SLAM has been presented in the present paper that uses adaptive squashing function [34] to optimize the non-linearities in the navigation path of the robot in a social environment.…”

Section: Related Workmentioning

confidence: 99%

“…xy R t R t ++ as the position coordinates for the next motion step, as targeted from the stereo-vision camera image processing. The input-output relationship for the ANN structure is described as [34]:…”

Section: Robot Kinematics and Navigationmentioning

confidence: 99%

“…f is the nonlinear squashing-function, which is considered here as hyperbolic-tangent function with slope factor () q j  . To attain faster convergence of Neural Networks, which is vital for navigation, an adaptive squashing function that adjusts the slope of squashing in addition to other network parameters through back propagation of errors into the network is considered [34]. The update of network parameters including the slope of squashing function starts with assigning trivial random values to the weights and unity slope to the squashing function for each neuron.…”

Section: Robot Kinematics and Navigationmentioning

confidence: 99%

See 2 more Smart Citations

Map Making in Social Indoor Environment Through Robot Navigation Using Active SLAM

et al. 2022

View full text Add to dashboard Cite

Robotics has come a long way from industrial robotic arms and is all set to enter our homes. The capability of a robot to navigate in an unknown human populated environment with obstacles and making map simultaneously is one of the significant characteristics in the domain of autonomous robotics. Further, the problem of robot navigating in a social environment while ensuring human safety and comfort through social norms needs to be addressed. This article presents a solution for mapping of unknown terrains with dynamic obstacles using simultaneous localization in social environments through Adaptive Squashing Function based artificial neural network training, which is able to track the target orientation angles more efficiently as compared to conventional fixed slope squashing function based backpropagation training algorithm. The performance of different state of the art techniques have been compared with proposed work through simulation models. Simulation results demonstrated the effectiveness of the proposed algorithm in complex environment where the proposed algorithm converged in less than 50% of the iterations taken by the exhaustive search algorithms and approximately 33% of the iterations taken by random search algorithm. Further, the proposed approach was tested in the real world settings, wherein the robot was deployed to create map for the

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Robot Kinematics and Navigationmentioning

confidence: 99%

Section: Robot Kinematics and Navigationmentioning

confidence: 99%

See 1 more Smart Citation

Map Making in Social Indoor Environment Through Robot Navigation Using Active SLAM

et al. 2022

View full text Add to dashboard Cite

show abstract

“…A tanh activation function with adaptive slope was used in an multi-layer perceptron (MLP) architecture in [507]. The used activation function is defined as…”

Section: Adaptive Slope Hyperbolic Tangentmentioning

confidence: 99%

On Transformative Adaptive Activation Functions in Neural Networks for Gene Expression Inference

Kunc

Kléma²

2019

Preprint

View full text Add to dashboard Cite

Motivation: Gene expression profiling was made cheaper by the NIH LINCS program that profiles only ∼1, 000 selected landmark genes and uses them to reconstruct the whole profile. The D-GEX method employs neural networks to infer the whole profile. However, the original D-GEX can be further significantly improved. Results: We have analyzed the D-GEX method and determined that the inference can be improved using a logistic sigmoid activation function instead of the hyperbolic tangent. Moreover, we propose a novel transformative adaptive activation function that improves the gene expression inference even further and which generalizes several existing adaptive activation functions. Our improved neural network achieves average mean absolute error of 0.1340 which is a significant improvement over our reimplementation of the original D-GEX which achieves average mean absolute error 0.1637

show abstract

“…The CMA algorithm operates in a simple unsupervised manner and was shown to be effective for non-linear channels. ANN uses an adaptive-slope squashing function (ASF) [46] for the estimation of channel state information (CSI) and detection of the symbol. The application of support vector machines to the equalization of communication networks that have been distorted by additive white Gaussian noise, inter-symbol and co-channel interference has been well reported in [47].…”

Section: Introductionmentioning

confidence: 99%

Training Strategy of Fuzzy-Firefly Based ANN in Non-Linear Channel Equalization

Mohapatra¹,

Rout²,

Bisoy

et al. 2022

IEEE Access

View full text Add to dashboard Cite

Channel equalization is remaining a challenge for the researcher. Especially for the non-linear channel as well as the extremely dispersive channel, an effective channel equalizer is required. It is common knowledge that non-linear channel equalizers based on the neural networks (NN) outperform adaptive filter-based linear equalizers. To train NN equalizers, gradient-descent-based approaches like the back-propagation algorithm are often utilized, although they have drawbacks such as trapping of local minima, slower convergence, and compassion to log in. In this work, we presented a novel training strategy using a fuzzy firefly algorithm (FFA) for channel equalization. By using proper network topology and parameters, the suggested training system offers stronger exploitation and exploration skills, as well as the ability to solve the local minima issue. The performance of the equalizer can be analyzed by estimating two parameters i.e. MSE and BER. To exhibit the suggested technique's resilience in performance, the burst error situation was used, and the outcomes showed that the strategy is more effective in managing such situations than previous methods. The outcomes of the proposed method are presented through simulation, Furthermore, it proved that the suggested method validates a wide range of SNR, and also it outperforms the existing NN-based equalizers.

show abstract

Adaptive-Slope Squashing-Function-Based ANN for CSI Estimation and Symbol Detection in SFBC-OFDM System

Cited by 5 publications

References 42 publications

Map Making in Social Indoor Environment Through Robot Navigation Using Active SLAM

Map Making in Social Indoor Environment Through Robot Navigation Using Active SLAM

On Transformative Adaptive Activation Functions in Neural Networks for Gene Expression Inference

Training Strategy of Fuzzy-Firefly Based ANN in Non-Linear Channel Equalization

Contact Info

Product

Resources

About