UWB On-Body Channel Modeling and Analysis for Diverse Indian Dance Forms

In this work, an ultrawideband (UWB) on-body channel has been modeled and empirically analyzed across various on-body channel links, with multiple variations introduced by different Indian dance forms (IDFs). A comprehensive system-level analysis is conducted using root mean square (rms) delay spread ( $\sigma _{\tau }\text {)}$ and bit error rate (BER) to quantify the system performance. Experimental results demonstrate that the proposed UWB wireless body area network (WBAN) has signal reliability when measured through a cumulative distribution function (cdf), representing a unique pattern for various IDFs and BER $\lt 10^{-{4}}$ . First-order statistics, such as $\sigma _{\tau }^{}$ and BER, give an average view, which requires further investigation. Level crossing rate (LCR), average fade duration (AFD), and fade probability [Pr(F)] are second-order statistics that enhance the channel analysis by providing the detailed information on signal variability over time. In addition, a BER versus channel model (CM) analysis is conducted for Nakagami, Rician, Rayleigh fading, and proposed CMs. The proposed CM exhibited superior performance due to its flexibility in accommodating various fading conditions, highlighting its suitability to track the dance performance. The validity of the proposed CM is further confirmed using akaike information criteria (AICs), reinforcing its suitability for accurate on-body channel characterization. The classification of IDFs is achieved using statistical features derived from the $S_{{21}}$ parameter and evaluated via fivefold cross validation, yielding a mean classification accuracy of 94.88%. The proposed system achieves better performance than convolutional neural network (CNN)- and long short-term memory (LSTM)-based classifiers and is suitable for low-complexity, real-time implementation. This information is vital for optimizing system performance in fluctuating environments.