This paper investigates and enhances unmanned aerial vehicle (UAV) relay-assisted free-space optics (FSO) optical wireless communication (OWC) systems under the effects of pointing errors (PEs) and atmospheric turbulences (ATs). The incorporation of UAVs as buffer-aided moving relays in the conventional FSO (CFSO) relay-assisted systems is proposed for enhancing the performance of PEs through AT. Using M-PSK (phase shift keying) and M-QAM (quadrature amplitude modulation), the impact of PEs on transmission quality is evaluated in this work. We evaluate and optimize the symbol error rate, outage probability (OP), and signal-to-noise ratio (SNR) for the UAV-to-ground station-based FSO communications systems. The spatial diversity-based relay-assisted CFSO systems can enhance the performance of the UAV-UAV FSO links. In this paper, a new FSO (NFSO) channel model for the hovering UAV-FSO OWC fluctuations under the PEs, AT effects, jitter, deviation, receiving an error, and wind resistance effects are established. To improve the performance of hovering UAV-based FSO relay OWC systems. We reduce the influence of UAV-FSO OWC fluctuations under PEs and AT effects. By receiving incoherent signals at various locations, the spatial diversity-based relay-assisted NFSO systems can significantly increase the system's redundancy and enhance connection stability. Numerical results show that to achieve a bit-error-rate (BER) of $$\le 10^{ - 5}$$≤ 10 - 5 , the required SNR is ≥ 23 dB when the wind variance of the UAVs $$\sigma_{\alpha }^{2}$$σ α 2 increases from 0 to 7 mrad with FSO link distance L = 2000 m. The required SNR is ≥ 25 dB when the wind variance $$\sigma_{\alpha }^{2}$$σ α 2 is 1 mrad at an OP of $$10^{ - 6}$$10 - 6 . To obtain an average BER of $$10^{ - 6}$$10 - 6 , the SNR should be 16.23 dB, 17.64 dB, and 21.45 dB when $$\sigma_{\alpha }^{2}$$σ α 2 is 0 mrad, 1 mrad, and 2 mrad, respectively. Using 8-PSK modulation without PEs requires 23.5 dB at BER of $$10^{ - 8}$$10 - 8 while 16-QAM without PEs requires 26.5 dB to maintain the same BER of $$10^{ - 8}$$10 - 8 . Compared with 16-QAM without PEs, the SNR gain of 8-PSK without PEs is 3 dB. The results show the relay-assisted UAV-FSO system with five stationary relays can achieve BER $$10^{ - 8}$$10 - 8 at 25 dB SNR in the ideal case and $$10^{ - 5}$$10 - 5 at 27 dB SNR with AT and PE at FSO length 1000 m. The results show the relay UAV-FSO system outperforms the CFSO at the BER and SNR performance. The effects of UAV’FSO s fluctuation increase when the UAV-FSO link length, $${\text{L}}_{{{\text{fso}}}}$$L fso increases. The results of the weak turbulence achieve better SER compared with MT and ST. The obtained results show that decreasing $${\text{L}}_{{{\text{fso}}}}$$L fso can compensate for the effects of UAV-FSO link fluctuation on the proposed system. Finally, we investigated the CFSO relay-assisted UAV-FSO system with aided NFSO-UAVs spatial diversity-based relay-based on NFSO OWC and revealed the benefits of the resulting hybrid architecture.

Maize, a fundamental crop globally, is particularly susceptible to a range of leaf diseases, which can result in substantial yield reductions and economic challenges for agricultural producers. Prompt and precise identification of these diseases is critical to minimizing their adverse effects on food security. This study investigates the application of ensemble machine learning methodologies to improve the robustness and accuracy of maize leaf disease classification. For this proposed experiment, the standard dataset has been utilized, dataset contains 3857 images belonging to blight, Common rust, gray leaf spot, and healthy leafs. By using this dataset three kinds of features (Gray-level co-occurrence matrix (GLCM), Local Binary Pattern (LBP) and Gabor) were extracted. This proposed experiment was carried out in three categories i.e., Single, Double and Multiple combination of features. These extracted features are submitted to three machine learning algorithms, such as s (SVM), kNN, and NN. In single feature Gabor with NN Classifier has given 85.40% as highest accuracy, in the Bi-features Gabor with LBP using NN algorithm has record the 88.00% as an output result, at last in Tri-features (Gabor + GLCM+LB) SVM has raised as a highest recognition accuracy as 88.80%.

Skin diseases, particularly melanoma and other types of pigmented lesions, constitute a significant portion of global health concerns due to their prevalence and potential severity. In recent years, deep learning (DL) has revolutionized image classification tasks in the medical domain, particularly using Convolutional Neural Networks (CNNs) for skin lesion analysis. However, traditional CNNs are limited to capturing spatial features, often overlooking sequential patterns and complex contextual cues inherent in dermatological imagery. This study explores the automated classification of pigmented skin lesions using the HAM10000 dataset, a diverse collection of 10,015 dermatoscopic images spanning seven diagnostic categories. Addressing challenges in computational dermatology, we leverage MobileNet-V2 and InceptionV3 deep learning architectures, optimized via transfer learning and advanced preprocessing techniques. Comparative evaluation is performed between baseline CNN models and their Long Short-Term Memory (LSTM)-augmented variants to assess improvements in classification performance through sequential feature modeling. Results indicate that LSTM integration enhances contextual feature learning, improving accuracy for underrepresented lesion classes, with InceptionV3+LSTM achieving the highest classification accuracy.

Green and smart buildings (SBs) are promising solutions to the fast-growing and ongoing urbanization around the globe and are the composition of huge services. The SB can acquire a tremendous amount of information and perform proper handling of many broadcasting residents accordingly. With the abrupt increase in the high-quality data transmission demands over long reach for the large number of connected residents in SBs, conventional data transmission techniques are inadequate to provide adequate services. Thus, in this paper, to improve the intelligence and reliability of SBs with high overall efficiency, cost-effectiveness, and security, a hybrid passive optical network (PON) and visible light communication (VLC) indoor broadcasting system is proposed. The bidirectional hybrid PON-VLC consists of integrated time and wavelength division multiplexing (TWDM) and wavelength division multiplexing (WDM) with optical code division multiplexing (OCDMA) PON with red green blue employed light emitting diodes (RGB LEDs) based VLC system. Seven OCDMA codes are utilized in the proposed work at a 10/2.5Gbps per channel transmission rate over hybrid fiber-VLC channels to strengthen the system performance. The simulation results show that the modified new zero cross-correlation (MNZCC) code outperforms other codes and provides a faithful 100km fiber length and 5m VLC range for 200 active residents. Moreover, VLC distance can be incremented up to 10m wireless range by increasing the received aperture diameter of 25cm. Also, the numerical analysis, as well as comparative performance, depicts the superiority of PON-VLC under the impact of impairments and noise in fiber-VLC link-based SBs.

A smooth evolution of 16×100Gbps hybrid mode division multiplexing (MDM)/coherent-next generation passive optical network (C-NGPON) system based on wavelength complement coding is proposed. Probabilistically shaped quadrature amplitude modulation (PS-QAM) modulation established alone as the key solution to embrace advanced coherent transponders. The simulation results show that the proposed hybrid MDM/C-NGPON system can offers a high split ratio of 512 in uplink and 256 in downlink as compared to conventional passive optical networks (PONs). For legacy PON signal coexistence with PON signals as well as complement signal the minimum received power of -19dBm for uplink and − 9dBm for downlink can be obtained effectively. Also, optical to signal noise ratio (OSNR) of 20dB can be achieved forPS-512QAM as compared to uniform-shaped 128QAM at 10KHz laser linewidth and it also offers power penalty of 0.5dBunder normalized generalized mutual information (NGMI) limit of 0.857. In addition, the system offers a long-haul transmission distance of 250km considering a shaping factor of 4.7 in uplink and 4.8 in downlink, at NGMI limit. Besides this, the system throughput upto 800Gbps for uplink and 700Gbps for downlink can also be obtained successfully. Finally, it is shown that proposed work shows superiority over other existing systems.