In response to various privacy risks, researchers and practitioners have been exploring different paradigms that can leverage the increased computational capabilities of consumer devices to train machine (ML) learning models in a distributed fashion without requiring the uploading of the training data from individual devices to central facilities. For this purpose, federated learning (FL) was proposed as a technique that can learn a global machine model at a central master node by the aggregation of models trained locally using private data. However, organizations may be reluctant to train models locally and to share these local ML models due to required computational resources for model training at their end and due to privacy risks that may result from adversaries inverting these models to infer information about the private training data. Incentive mechanisms have been proposed to motivate end users to participate in collaborative training of ML models (using their local data) in return for certain rewards. However, the design of an optimal incentive mechanism for FL is challenging due to its distributed nature and the fact that the central server has no access to clients’ hyperparameters information and the amount/quality data used for training, which makes the task of determining the reward based on the contribution of individual clients in FL environment difficult. Even though several incentive mechanisms have been proposed for FL, a thorough up-to-date systematic review is missing and this paper fills this gap. According to the best of our knowledge, this paper is the first systematic review that comprehensively enlists the design principles required for implementing these incentive mechanisms and then categorizes various incentive mechanisms according to their design principles. In addition, we also provide a comprehensive overview of security challenges associated with incentive-driven FL. Finally, we highlight the limitations and pitfalls of these incentive schemes and elaborate upon open-research issues that required further research attention.

In this letter, we propose a novel joint channel estimation and data detection scheme for massive multiple-input multiple-output (MIMO) uplink, which operates in a time-varying fading channel. More specifically, at the receiver of the proposed scheme, a data frame is divided into multiple blocks, and in each block, a demodulated data block is used for updating channel state information (CSI) in an iterative manner. Furthermore, the initial CSI in the block of interest is given by the estimated CSI in the previous block, hence allowing accurate tracking of CSI in a time-varying channel without imposing additional pilot insertion inside the data frame. Since the length of the divided blocks affects both the achievable channel tracking and data detection performances, it is optimized so as to maximize the discrete-input continuous-output memoryless channel’s (DCMC) capacity derived in this letter. It is demonstrated that the DCMC capacity of the proposed scheme is capable of nearly achieving those of the perfect CSI counterpart without imposing any substantial pilot overhead.

Aviation is an vital global industry that generates economic growth, creates jobs and facilitates national and international trade and tourism. During the Covid-19 pandemic, the world had seen a sharp decline in commercial air travel due to the worldwide travel restrictions, lock-down orders and widespread fear of infection. However, there is a rapid increase in the use of UAVs for surveillance, communication, delivery and transport. And the aviation industry is believed to be able to provide valid contribution to the post-pandemic economic recovery. In order to reduce the carbon emission from aviation industry, there is a requirement for more electric aircraft. Hydrogen and fuel cell system's capability of creating a clean road transportation have been widely tested. In this paper, the status of hydrogen and fuel cell utilised in aviation is reviewed, and the challenges and issues are identified for a better integration of the hydrogen and fuel cell system in aviation.

Remote ice detection has emerged as an application of Radio Frequency (RF) sensors. While antenna-based “RFID” sensing can detect various measurands, antenna-based sensors are not currently designed based on a systematic methodology, and in most cases may have a low sensitivity requiring specialist hardware or broadband interrogation signals, incompatible with spectrum regulations. Here, we develop a systematic methodology for designing an antenna-based sensor, applicable to measurands inducing a dielectric change in the near-field of the antenna. The proposed methodology is applied to designing printable antennas as highly-sensitive sensors for detecting and measuring the thickness of ice, demonstrating best-in-class sensory response compared to more complex antenna designs. Antenna design is investigated systematically for wireless interrogation in the 2.4 GHz band, where it is found that a loop antenna outperforms a dipole owing to its more distributed capacitance. The antenna’s realized gain was identified as the optimum parameter-under-test, with “positive” sensing proposed as a method of improving linearity and immunity to interference. The developed loop antenna sensor exhibits resilience to interference and applicability to different real-world deployment environments, demonstrated through over 80% average ice thickness measurement accuracy and at least 5 dB real-time sensitivity to ice deposition.

This research proposes a realistic mutual authentication protocol using non-ideal PUF in an RFID environment. Initially, we evaluate and show the weaknesses of two popular schemes, Gope et al.’s and Feng Zhu et al.’s schemes; based on our investigation, their schemes fail to achieve several security features and cannot withstand various attacks. Later, we propose a new scheme to resolve the problem their schemes and provide the solution for the RFID security problem generically. Analyzes informal and formal are used to ensure that our proposal fulfills security properties and withstands various attacks.

The Contribution to the 9th European Workshop on Visual Information Processing. The work is focuses on instance relevance estimation process in domain of multiple instance learning.

The key to efficient waste management is to ensure segregation of waste and resource recovery. Waste is usually segregated on the basis of whether it is biodegradable or non biodegradable. A major challenge with respect to this is that waste is not segregated before collection and is thrown into dustbins nonetheless. These end up as huge piles of waste in dump yards, which needs to be segregated. Waste segregation is currently being done manually by the Municipal Corporation. These results in unsanitary working conditions for the people who need to perform this task. Despite being provided with the necessary equipment, they run the risk of catching infections from the waste they work with. Automation of this process will be beneficial for the people who work on this, by reducing health hazards.

In this work, non-coherent massive MIMO differential phase-shift keying modulation (DPSK) detection is considered to get rid of the complexity of channel estimation. However, most of the well-performing DPSK detection techniques require high computational complexity at the receiver. The use of deep-learning is proposed for detecting the transmitted DPSK symbols over a single-user massive MIMO system. We provide a multiple-symbol differential detection implementation using deep-learning. Two deep-learning-based multiple-symbol differential detection receiver designs are proposed and compared with differential detection (DD), decision-feedback differential detection (DFDD), and multiple-symbol differential detection (MSDD) for the same system parameters. Where multiple-symbol differential sphere detection (MSDSD) is used to implement MSDD. The results show that the proposed deep-learning-based classification neural networks outperform decision-feedback differential detection and achieve an optimal performance compared to conventional multiple-symbol differential detection implemented by multiple-symbol differential sphere detection.

We describe a fast three-round mutual authentication protocol for parties A and B belonging to the same coalition group. Parties A and B keep their own independent long-term private keys that are used in the process of authentication and can be used for other purposes. The scheme assumes an initial setup with a trusted third party T. This party initiates another secret information that includes factors of a large RSA modulus. For authentication, both parties must demonstrate each other the knowledge of their private keys without revealing them and the ability to factorize a large RSA modulus. Thus, the protocol based on the suggested scheme provides reciprocal authentication. The scheme possesses all desirable properties of an interactive proof, i.e., completeness, soundness, and zero-knowledge. The security of the protocol relies on assumptions of difficulty of the RSA factorization and existence of a cryptographic hash function.

This paper proposes a solution that makes voltage scaling possible by simply using HLS tools provided by vendor to improve energy efficiency of FPGAs by 2x. Chip manufacturers define voltage margins on top of the “best-case” operational voltage of their chips to ensure reliable worst case functionality. The margins guarantee correct-ness of operation, but at the cost of performance and power efficiency. Violating the margins is tempting to save energy, but might lead to timing errors. This paper proposes an algorithmic solution that enables reliable removal of the margins by detecting errors on-the-fly. In contrast to previous approaches that require special hardware to detect timing errors, the proposed method is fully implementable using High Level Synthesis tools without reliance on additional hardware. The approach is demonstrated using a32×32matrix-matrix multiplication and a simple multi-layer neural network implemented on two Xilinx ZC702Field-Programmable-Gate-Array (FPGA) System-on-Chip (SoC)platforms, showcasing its utility in detecting errors that may originate from different sources of logic circuity, clock tree and memory. Results show that the energy dissipation is halved, while the implementation is clocked at 2.5x faster than specified by the design tool of the vendor.

This paper applies a Spatiotemporal Graph Convolutional Recurrent Neural Network which is a tight combination of a Graph Neural Network (GNN) to a Recurrent Neural Network (RNN) architecture for air pollution forecasting in long-term for the entire city. Our model can effectively learn the spatial and temporal features of the air pollution data and its influential factors (e.g. weather, traffic, external areas) at the time. Our method achieves better performance than a state-of-the-art ConvLSTM model in air pollution forecasting and a hybrid GNN-based model that separates GNN and RNN in discrete layers.

Objective: To develop and validate an algorithm for predicting non-attendance to outpatient appointments. Results: We developed two decision tree models for dermatology and pneumology services (trained with 33,329 and 21,050 appointments, respectively). The prospective validation showed a specificity of 78.34% (95%CI 71.07, 84.51) and a balanced accuracy of 70.45% for dermatology; and 69.83% (95%CI 60.61, 78.00) - 65.53% for pneumology, respectively. When using the algorithm for identifying patients at high risk of non-attendance in the context of a phone-call reminder program, the non-attendance rate decreased 50.61% (P<.001) and 39.33% (P=.048) in the dermatology and pneumology services, respectively. Conclusions: A machine learning model can effectively identify patients at high risk of non-attendance based on information stored in electronic medical records. The use of this model to prioritize phone call reminders to patients at high risk of non-attendance significantly reduced the non-attendance rate.

A document by Gopakumar K . Click on the document to view its contents.

Abstract—In this work, a multi-level 42-sided polygonal space vector structure (SVS) for suppression of lower order harmonics for Open-End Induction Motor(OEIM) drive applications is proposed. The proposed power circuit topology consists of two inverters feeding an Open-End Induction Motor from either side. The main inverter fed with a single DC link providing active power for motor operation is switched at low switching frequency. The secondary inverter fed with a capacitive supply is switched at high frequency to suppress lower order harmonics upto 39th order, up to the base speed of operation allowing maximum utilization of the DC link. The advantages of lower order harmonic suppression in motor phase voltage, for polygonal space vector structures are combined with multi-level inverter topology. This results in lower switching losses in low frequency switching main inverter and low voltage secondary inverter. Use of a single DC link facilitates four quadrant operation of the inverter. The proposed scheme is validated for steady state and dynamic performance by experimental results.

Retinal images acquired using fundus cameras are often visually blurred due to imperfect imaging conditions, refractive medium turbidity, and motion blur. In addition, ocular diseases such as the presence of cataract also result in blurred retinal images. The presence of blur in retinal fundus images reduces the effectiveness of the diagnosis process of an expert ophthalmologist or a computer-aided detection/diagnosis system. In this paper, we put forward a single-shot deep image prior (DIP)-based approach for retinal image enhancement. Unlike typical deep learning-based approaches, our method does not require any training data. Instead, our DIP-based method can learn the underlying image prior while using a single degraded image. To perform retinal image enhancement, we frame it as a layer decomposition problem and investigate the use of two well-known analytical priors, i.e., dark channel prior (DCP) and bright channel prior (BCP) for atmospheric light estimation. We show that both the untrained neural networks and the pretrained neural networks can be used to generate an enhanced image while using only a single degraded image. We evaluate our proposed framework quantitatively on five datasets using three widely used metrics and complement that with a subjective qualitative assessment of the enhancement by two expert ophthalmologists. We have compared our method with a recent state-of-the-art method cofe-Net using synthetically degraded retinal fundus images and show that our method outperforms the state-of-the-art method and provides a gain of 1.23 and 1.4 in average PSNR and SSIM respectively. Our method also outperforms other works proposed in the literature, which have evaluated their performance on non-public proprietary datasets, on the basis of the reported results.

Polymer modified Quartz Tuning Fork sensor array with three sensors has been developed to classify 1,4-dimethoxy-2, 3- butanediol (BD), cyclohexanone (CH), ethanol, methanol and acetone with various concentrations. Linear Discriminant Analysis (LDA), Decision Tree (DT), and Random Forrest (RF) have been used and evaluated as the classifiers predicting the VOCs and their concentration. The classifiers were trained with three features: response time, recovery time, and shift in the resonant frequency of the sensors. The results have been compared and analyzed for the performance of each classifier for the present system. RF gives the best performance as a classifier with an accuracy of classification more than 95%. We present results of VOC classifier performance for the QTF sensor array.

This research is proposed to resolve the problems in the existing user authentication protocol. Our proposal fulfills security features and withstands well-known attacks.

We propose improved unequal-clustering and routing protocol (IUCR) protocol to solve both of these problems jointly. IUCR provide fixed area clustering derived from transmission range of network nodes. This clustering also develops strong network backbone that provides fail-over-proof routing. Efficient routing path is achieved by finding minimal hop-count with availability of alternate routing path.