Diabetic retinopathy (DR) is a human eye disease that affects people who are suffering from diabetes. It causes damage to their eyes, including vision loss. It is treatable; however, it takes a long time to diagnose and may require many eye exams. Early detection of DR may prevent or delay the vision loss. Therefore, a robust, automatic and computer-based diagnosis of DR is essential. Currently, deep neural networks are being utilized in numerous medical areas to diagnose various diseases. Consequently, deep transfer learning is utilized in this article. We employ five convolutional-neural-network-based designs (AlexNet, GoogleNet, Inception V4, Inception ResNet V2 and ResNeXt-50). A collection of DR pictures is created. Subsequently, the created collections are labeled with an appropriate treatment approach. This automates the diagnosis and assists patients through subsequent therapies. Furthermore, in order to identify the severity of DR retina pictures, we use our own dataset to train deep convolutional neural networks (CNNs). Experimental results reveal that the pre-trained model Se-ResNeXt-50 obtains the best classification accuracy of 97.53% for our dataset out of all pre-trained models. Moreover, we perform five different experiments on each CNN architecture. As a result, a minimum accuracy of 84.01% is achieved for a five-degree classification.

Classification and regression are the major applications of machine learning algorithms which are widely used to solve problems in numerous domains of engineering and computer science. Different classifiers based on the optimization of the decision tree have been proposed, however, it is still evolving over time. This paper presents a novel and robust classifier based on a decision tree and tabu search algorithms, respectively. In the aim of improving performance, our proposed algorithm constructs multiple decision trees while employing a tabu search algorithm to consistently monitor the leaf and decision nodes in the corresponding decision trees. Additionally, the used tabu search algorithm is responsible to balance the entropy of the corresponding decision trees. For training the model, we used the clinical data of COVID-19 patients to predict whether a patient is suffering. The experimental results were obtained using our proposed classifier based on the built-in sci-kit learn library in Python. The extensive analysis for the performance comparison was presented using Big O and statistical analysis for conventional supervised machine learning algorithms. Moreover, the performance comparison to optimized state-of-the-art classifiers is also presented. The achieved accuracy of 98%, the required execution time of 55.6 ms and the area under receiver operating characteristic (AUROC) for proposed method of 0.95 reveals that the proposed classifier algorithm is convenient for large datasets.

This paper presents the employment of a DPA attack on the NIST (National Institute of Standards and Technology) standardized AES (advance encryption standard) protocol for key retrieval and prevention. Towards key retrieval, we applied the DPA attack on AES to obtain a 128-bit secret key by measuring the power traces of the computations involved in the algorithm. In resistance to the DPA attack, we proposed a countermeasure, or a new modified masking scheme, comprising (i) Boolean and (ii) multiplicative masking, for linear and non-linear operations of AES, respectively. Furthermore, we improved the complexity involved in Boolean masking by introducing Rebecca’s approximation. Moreover, we provide a novel solution to tackle the zero mask problem in multiplicative masking. To evaluate the power traces, we propose our custom correlation technique, which results in a decrease in the calculation time. The synthesis results for original implementation (without countermeasure) and inclusion of countermeasure are given on a Zynq 7020 FPGA (Artix-7 device). It takes 424 FPGA slices when implemented without considering the countermeasure, whereas 714 slices are required to implement AES with the inclusion of the proposed countermeasure. Consequently, the implementation results provide the acceptability of this work for area-constrained applications that require prevention against DPA attacks.

This article presents a latency-optimized implementation of the SIMECK lightweight block cipher on a field-programmable-gate-array (FPGA) platform with a block and key lengths of 32 and 64 bits. The critical features of our architecture include parallelism, pipelining, and a dedicated controller. Parallelism splits the digits of the key and data blocks into smaller segments. Then, we use each segmented key and data block in parallel for encryption and decryption computations. Splitting key and data blocks helps reduce the required clock cycles. A two-stage pipelining is used to shorten the critical path and to improve the clock frequency. A dedicated controller is implemented to provide control functionalities. For the performance evaluation of our design, we report implementation results for two different cases on Xilinx 7-series FPGA devices. For our case one, the proposed architecture can operate on 382, 379, and 388 MHz frequencies for Kintex-7, Virtex-7, and Artix-7 devices. On the same Kintex-7, Virtex-7, and Artix-7 devices, the utilized Slices are 49, 51, and 50. For one encryption and decryption computation, our design takes 16 clock cycles. The minimum power consumption is 172 mW on the Kintex-7 device. For the second case, we targeted the same circuit frequency of 50 MHz for synthesis on Kintex-7, Virtex-7, and Artix-7 devices. With minimum hardware resource utilization (51 Slices), the least consumed power of 13.203 mW is obtained for the Kintex-7 device. For proof-of-concept, the proposed SIMECK design is validated on the NEXYS 4 FPGA with the Artix-7 device. Consequently, the implementation results reveal that the proposed architecture is suitable for many resource-constrained cryptographic applications.

An unmanned free-swimming submersible (UFSS) is designed to perform certain tasks in water without interposing humans. The vehicle’s control is achieved by integrating mathematical (analog) and non-mathematical (embedded) controllers. The main goal of integrated controllers is to overcome the environmental disturbances and noise of the sensor data. These disturbances, as well as the noise data, are generated during steering, diving, and speed control. The amplitude of disturbances and noise varies with the depth and intensity of water waves. This article presents a robust hybrid linear quadratic regulator (HLQR) controller for UFSS. The presented controller targets the desired state of the UFSS in the presence of a disturbing environment. The hybrid approach is achieved by employing: (1) two linear quadratic regulators or controllers and (2) a mathematical structure of the Riccati equation. Consequently, the proposed HLQR controller is integrated into the UFSS system to evaluate the response in terms of settling time, rise time, overshoot, and steady-state error. Furthermore, the robustness of the HLQR is investigated by considering the feedback to step response and hydrodynamic disturbances. The implementation results reveal that the proposed controller outperforms state of the art controllers, such as proportional-integral-derivative and lead-compensator controllers.

The tele-presence robot is designed to set forth an economic solution to facilitate day-to-day normal activities in almost every field. There are several solutions to design tele-presence robots, e.g., Skype and team viewer, but it is pretty inappropriate to use Skype and extra hardware. Therefore, in this article, we have presented a robust implementation of the tele-presence robot. Our proposed omnidirectional tele-presence robot consists of (i) Tricon ultrasonic sensors, (ii) Kalman filter implementation and control, and (iii) integration of our developed WebRTC-based application with the omnidirectional tele-presence robot for video transmission. We present a new algorithm to encounter the sensor noise with the least number of sensors for the estimation of Kalman filter. We have simulated the complete model of robot in Simulink and Matlab for the tough paths and critical hurdles. The robot successfully prevents the collision and reaches the destination. The mean errors for the estimation of position and velocity are 5.77% and 2.04%. To achieve efficient and reliable video transmission, the quality factors such as resolution, encoding, average delay and throughput are resolved using the WebRTC along with the integration of the communication protocols. To protect the data transmission, we have implemented the SSL protocol and installed it on the server. We tested three different cases of video resolutions (i.e., $320\times 280$, $820\times 460$ and $900\times 590$) for the performance evaluation of the video transmission. For the highest resolution, our TPR takes 3.5 ms for the encoding, and the average delay is 2.70 ms with 900 × 590 pixels.