Cooch Behar Government Engineering College

A flexible, intensity-modulated optical bend-sensor was realized using a TiO2 similar to gel-wax nanocomposite core and clear silicone-rubber cladding in a rectangular macro-waveguide format. A 940 nm infrared (IR) light emitting diode (LED) and Si PIN photodiode enabled simple photometric readout with electromagnetic-interference immunity in a thin, conformal package. An integrated analytical framework linked LED emission, coupling, wavelength-dependent absorption, macro-bending radiation loss, and photodiode responsivity to predict photocurrent versus bend radius; durability effects up to 10,000 cycles were captured by a polynomial update of a single core-dependent parameter. Fabrication employed cast silicone-rubber layers and a 0.25 wt% TiO2-loaded gel-wax core, with vacuum curing to improve optical quality. Electrical characterization established the LED operating point, and bending tests over the bend-radius (R) 1.0-75.0 cm validated the model with a maximum deviation within +/- 5 %. The practical linear range was R = 1-30 cm (full measurable range R = 1-75 cm), with model-to-measurement deviation within +/- 5 %; the pointwise sensitivity decreased monotonically from 120.7 to 34 mu A cm(-1) as R increased from tight to gentle bends. After durability testing, the straight-state current drifted modestly (approximate to 4.11 % at 5000 cycles; approximate to 14.33 % at 10,000 cycles), while the sensitivity remained nearly unchanged; dynamic response/recovery times were not measured in this quasi-static study. The architecture supports straightforward electronics, scalable geometry, and a pathway to mass production via 3-D printing, positioning the device for wearable joint-angle tracking, soft-robotic hinge feedback, field curvature monitoring, etc.

This work reports the design, fabrication, and comprehensive characterization of a matrix-coupled optoelectronic resistor array based on white InGaN light-emitting diodes (LEDs) and CdS photoresistors, developed as a tunable electronic material platform for wide-range analog resistance control. The device incorporates an M × N (4 × 4 demonstrated) LED–photoresistor matrix housed within a mirror-coated polymeric spacer, enabling distance-dependent optical coupling that governs the electronic response of the CdS detectors. Detailed materials processing steps—including polylactic acid-based spacer fabrication, protected-silver reflective coatings, hot-melt encapsulation, and black nitrocellulose optical isolation—are presented to highlight the role of structural and interfacial materials in optical confinement and thermal stability. The optical–electronic behavior is described through an analytical model integrating a modified Shockley LED law, power-law radiance scaling, and Beer–Lambert attenuation through the spacer medium. Model parameters were extracted from measured LED I–V characteristics and fitted light-dependent resistor (LDR) resistance–voltage datasets, achieving excellent agreement with experiments (typical pointwise error < ± 3

This paper focuses on a low-complexity precoding weight suitable for multi-user massive multiple-input multiple-output (MU-mMIMO) in user equipments’ (UEs’) mobility environment. Null-space expansion (NSE) method achieves better inter-user interference (IUI) suppression performance than channel prediction schemes by steering several nulls for interfering UTs in the high mobility environment. However, NSE faces challenges in effectively managing the computation complexity of the weight design. In this work, we incorporate a Richardson iteration (RI) in zero-forcing (ZF)-based NSE method to reduce computation burden. Moreover, deep-unfolding (DU) is utilized to further accelerate the convergence of RI-based NSE method. Computer simulation demonstrates the superiority of the proposed scheme in terms of IUI suppression performance and computation complexity.

This paper proposes a low-complexity minimum mean square error (MMSE) detection scheme based on the block successive over-relaxation (BSOR) method, which avoids direct channel matrix inversion. By exploiting the circulant and sparse structure of the orthogonal time frequency space (OTFS) channel, we employ fast Fourier transform (FFT)-based diagonal block inversion and matrix-vector multiplications to reduce complexity from cubic to near-linear order. Additionally, we introduce an efficient Gram matrix computation and a matched filtering technique using FFTs, and propose a matrix compression scheme that significantly reduces storage from $\mathcal{O}\left( {{M^2}{N^2}} \right){\text{ }}$ to $\mathcal{O}(\alpha MN),$ where α≪M denotes the sparsity level, where M and N are the numbers of delay and Doppler bins, respectively. Simulation results show that the proposed BSOR detector achieves near-MMSE performance with significantly reduced computation complexity.

This paper proposes a low-complexity cross-domain approximate stochastic expectation propagation (CD-SEPA) based detector for multiple-input multiple-output (MIMO) systems with orthogonal time-frequency space (OTFS) modulation. The expectation propagation (EP) detectors in the delay-Doppler (DD) domain incur high computational complexity. To reduce complexity, the proposed detector adopts a cross-domain iterative detection framework that jointly exploits the sparsity of the time domain (TD) channel and the discrete symbol constraints in the DD domain. In the time domain, a Chebyshev-accelerated weighted Neumann series approximation (Cheby-wNSA) is utilized to efficiently compute the posterior mean, reducing the matrix inversion complexity. The resulting posterior statistics are then transformed to the DD domain, where a randomly selected subset of factors is updated at each iteration. Furthermore, the detector exploits a pruned neighborhood of constellation points around the current symbol estimate to further reduce computational complexity. The simulation results show that the proposed CD-SEPA detector achieves near-exact EP performance with significantly reduced computational complexity.