Seyed Sadra Kashef

Transmission Parameter-based Demodulation in Visible Light Communications using Deep Learning

This paper presents a compact bandpass filter (BPF) with a triple rectangular split resonator optimized for high performance at a central frequency of 3.682 GHz. The filter achieves an exceptional voltage standing wave ratio (VSWR) of 1.088 and a return loss of 27.48 dB, demonstrating superior impedance matching. Additionally, the filter exhibits minimal insertion loss with S21?=0.38dB, ensuring efficient signal transmission. The design boasts a sharp roll-off rate of 87 and a narrow transition band of 0.196 GHz, making it suitable for high-selectivity applications. The compact size of the filter, measuring only 24?mm×24?mm, enhances its applicability in modern communication systems with limited space requirements.

In this paper, the physical layer security (PLS) in visible light communication (VLC) is investigated where the eavesdropper (ED) is assisted with the reconfigurable intelligent surface (RIS). Here, in the relay-aided system model, the ED is equipped with a new RIS type. The model also accounts for the received signal-dependent noise. For various trusted relaying schemes such as decode-and-forward (DF), amplify-and-forward (AF), and cooperative jamming (CJ), closed-form expressions of the lower and upper bounds on secrecy rate regions are derived under power constraints. Moreover, the effect of channel state information (CSI) uncertainty is analysed, as a practical condition. Considering this effect, it is demonstrated that the CJ scheme resists eavesdropping by increasing the noise and interference on the illegitimate user. The simulation results show that applying the RIS technique does not increase the signal-to-noise ratio of the ED in the CJ scheme in comparison with DF and AF schemes.

Contamination of oil, particularly by dissolved water, is a very common problem in the failure of step-down transformers used by electricity providers as this degrades the insulating property of the oil. In this paper, the use of D-shaped optical fibers functionalized with Graphene Oxide is presented to detect the water content in transformer oil. The synthesis of graphene oxide was achieved by a modified version of Hummer's method. Subsequently, the drop-casting process was used to apply the graphene oxide onto the D-shaped fibre. The coating thickness attained in the samples was around 200 nm. Side polishing in a single-mode fiber engages an evanescent field that increases its sensitivity as an optical sensor. A few layers of graphene oxide coating on D-fiber exhibit a quick response time and high sensitivity to moisture content present in transformer oil, which proves to be a hopeful solution in real-time monitoring and maintenance of transformer insulation systems. It manifested that the experimental results had a high sensitivity to different water contents in transformer oil for the D-shaped fiber coated with GO. The GO-coated fibers exhibited a sensitivity of 0.5677 dB/ppm, which is relatively high compared with the sensitivity in the case of uncoated D-shaped fibers.

Wireless body area networks (WBANs) can be utilized for remote observation of patients and old individuals at home as medical applications. Visible light communication (VLC) has been proposed as an efficient technology for medical WBANs. By considering the key feature of VLC systems, VLC can be used in extra-WBAN links from node (patients) to access point. Since symbol error rates (SERs) is critical metric in medical WBANs, we propose joint power and subcarrier allocation scheme in VLC systems to minimize SERs subject to transmit power limitation, and required minimum rate for each user. Here, the improved channel model is utilized for VLC link between different numbers of patients and access point which is considered received signal power dependence noise (RSN). In addition, the effect of tilting assisted receiver on resource allocation is evaluated. Moreover, to address real-world challenges like patient movement and obstructions, we propose incorporating wall-mounted mirrors to reflect light signals and thus maintain reliable communication, considering both line-of-sight (LoS) and non-line-of-sight (NLoS) transmission. The numerical results indicate that RSN and receiver tilting have a significant impact on the allocation of resources. Furthermore, it is shown that outage probability can approach zero using the NLoS link of proposed wall-mounted mirrors.

This paper presents firstly the literature review and then investigates the design, simulation, and performance analysis of a compact microstrip bandpass filter (BPF) for wireless communication, IoT, and radar systems as candidate for TWIN OPEN LOOP RESONATORS in our future studies of filter and diplexers. Using ROGER RT (RO3003), an open-loop square with two SIRS at the top of the resonator structure is investigated for the filter, which is optimized to operate at the center frequency of 3,82 GHz with an external quality factor of 57,88. Regarding S-parameter results, the performance is good, with an insertion loss of 0,41 dB, while a high return loss is achieved at 21,66 dB. Such characteristics guarantee good selectivity and low signal distortion at an upper band of 83,74 dB /GHz with a transition band of 0,269 GHz. The proposed BPF has a transmission zero of 71,53 dB at 4,427 GHz with a compact size of 18 mm × 18 mm.

In this work, we evaluate the efficiency of different one-dimensional and two-dimensional multiple signal classification (MUSIC) scenarios in direction-finding (DF) for the azimuth and the elevation angles through the Monte Carlo simulations. In addition, a new second order differential array (SODA) based on two L-shaped array (TLSA) geometry is proposed to enhance the accuracy of the DF algorithm in the presence of mutual coupling effects. Moreover, we studied the effect of the elevation estimation error on the azimuth DF accuracy. For this purpose, the root mean square error values of the azimuth and elevation angles with different arrangements of a SODA-based nine elements TLSA geometry are calculated. It shows that DF-based on the proposed sparse TLSA has a better performance than uniform TLSA in accuracy and complexity. The simulations show that establishing a one-dimensional MUSIC DF algorithm for two independent linear arrays of monopole antennas can estimate the azimuth and the elevation angles precisely with acceptable time consumption due to one-dimensional peak search of the Pseudo-spectrum function.

The optical sensing devices has been fabrication based on D-Shaped Fiber coated with Graphene Oxide (GO). The single mode Step Index fiber (SMF-28) used resembl The optical sensing devices has been fabrication based on D-Shaped Fiber coated with Graphene Oxide (GO). The single mode Step Index fiber (SMF-28) used resembl