Improved BER Performance Using Novel ELM Based Demodulator for CAP-VLC System

Prakash, Veda; Manimegalai, C. T.; Kalimuthu, K.; Gauni, Sabitha

doi:10.1109/access.2021.3124968

Cited by 2 publications

(2 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To show the reliability of the system we have added complexity analysis and time delay in Table 2. 13 The proposed research is targeted for using the learning models for increasing the efficiencies of VLC out of which we have iterated several ML algorithms out of which we have proposed the ELM as the best one. Table 2 shows the ELM has been chosen based on its less complexity and minimal delay.…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Design of matched filter for signal‐to‐noise ratio maximization to improve data rate

Prakash

Manimegalai

2022

Micro & Optical Tech Letters

Self Cite

View full text Add to dashboard Cite

The matched filter is an optimal linear filter for maximizing signal-to-noise ratio (SNR) in the presence of additive random noise. Learning-based matched filter is the proposed work. The proposed model supports (i) maximizing the SNR and (ii) setting the threshold, it is also called an intelligent matched dual tire filter. This improves the SNR by about 20 dB and the corresponding data rate to over 400 MBit/s. The bit error rate for 20 dB is very lower than FEC (10 −3 ) of up to 10 −7 . The accuracy of the proposed matched learning filter is proved to be above 95%. Indoor visible light communication is one of the applications, and it can also be used for image processing and outdoor communication such as satellite and radar communication. It is obvious for communication as it avoids signal degradation, lobing, and overfitting to achieve a better data rate. This requires funding support and infrastructure facility that can be used to deploy the proposed algorithm in the real-time scenario in the future.

show abstract

Section: Resultsmentioning

confidence: 99%

“…The solid line explains, for example, learning platform with 10e−7 and the output with different symbols marked are explained without learning platform up to 10e−6. To show the reliability of the system we have added complexity analysis and time delay in Table 2 13 …”

Section: Resultsmentioning

confidence: 99%