Search

In this paper we proposed a novel technique for Atrial Activity (AA) decomposition in Electrocardiogram (ECG) of Atrial Fibrillation (AF). The main purpose of our proposed technique is to decompose AA signal by combining two statistical methods, Independent Component Analysis (ICA)-existing and Weighted Average Beat Subtraction (WABS)-new, for AF with multiple stable sources, respectively. We found the limits of BSS algorithms which are mostly used to extract AA signal, while beauty of our proposed algorithm is that it decomposes multi-lead AA signals from surface ECG with AF. Our proposed technique is verified with clinical data and the results demonstrate that our proposed method is feasible.

As wireless sensor networks continue to grow in usage for distinct applications. Storage (memory) requirement and security are major problems for these resource limited networks. This paper present a novel and an efficient key establishment and management scheme for WSNs, named Random Initial and Master Key (RIMK).We showed that our proposed technique, manages keys randomly with less storage (i.e. only two keys), less communication and computation overhead compared to existing Localized Encryption and Authentication Protocol (LEAP), in which every node uses four different keys in a static pattern and occupy greater storage space, which increases overhead and complexity in the network. Simulation results show that RIMK is more efficient and robust than LEAP. DOI: http://dx.doi.org/10.11591/telkomnika.v11i6.2703

Emerging trends in Internet of things (IoT) has caught the attention of every domain e.g., industrial, business, and healthcare etc. Sensor-embedded IoT devices are the key drivers for collecting large amount of data. Managing these large datasets is one of the critical challenges to be tackled. Continuous huge information collection through sensor-enabled devices is known as the massive IoT (mIoT). Thus, there is a need of self-adaptive artificial intelligence (AI)-based strategies to effectively cluster, examine and interpret the entire entities in the system. With increased data volumes and power hungry natured IoT devices it is a dire need to manage their power wisely. To fairly allot the power levels to the tiny portable devices it is important to integrate mIoT with AI-based techniques. To remedy these issues this paper proposes a novel cross-layer based energy optimization algorithm (CEOA) in mIoT system by examining the detailed features and data patterns. Experimental analysis reveals that proposed CEOA outperforms its competing counterpart i.e., Baseline in terms of efficient power management and monitoring.

International audience