Search

Nowadays, Internet of Things (IoT) has been widely employed in different applications, such as health care, manufacturing, and weather forecasting. However, due to sensor sensitivities, potential harsh environmental interference, and deception, IoT data is normally apt to be imperfect and erroneous. This paper presents an edge-based approach based on the Gaussian mixture model and fuzzy measure to detect anomalous data without prior knowledge or training to overcome such adverse issues. The experimental results demonstrate that the proposed approach is efficient and effective in detecting anomaly data and achieves detection accuracy ranging from 93% to 100%.

IoT data is prone to different kinds of failures (hardware, software, and communication failures). Fault detection and recovery are challenging problems due to sensing devices’ limitations and the deployment field’s nature. Furthermore, timely and accurate detection of faulty data and recovery is highly significant to IoT applications to ensure operational stability and execution efficiency. This paper presents a faulty data detection and recovery approach based on dynamic interval-valued evidence and Kalman filter to accomplish this objective. The proposed approach is edge-based and requires no training to perform faulty data detection and recovery. The simulation results reveal that the proposed approach is efficient and effective in fault detection and recovery.

Wireless sensor networks (WSNs) are critically resource‐constrained due to wireless sensor nodes' tiny memory, low processing units, power limitations, and narrow communication bandwidth. The data reduction technique is one of the most widely used techniques to reduce transmitted data over the wireless sensor networks and to minimize the sensor nodes' energy consumption, particularly, the entire network in general. This paper proposes a reliable dual prediction data reduction approach for WSNs. This approach performs data reduction through two phases: the data reduction phase (DRP) and data prediction phase (DPP). The DRP is mainly to decrease the number of transmissions between the sensor node and the sink node, thereby minimizing energy consumption. It also detects faulty data and discards them at the sensor node. The discarded faulty data at the sensor nodes are replaced by estimated values at the sink node to maintain data reliability. DPP runs at the sink node or base station, which works in synchronization with the sensor nodes. This phase is responsible for predicting the non‐transmitted data based on the Kalman filter. The simulation results demonstrate that the proposed approach is efficient and effective in data reduction, data reliability, and energy consumption.

This paper proposes a novel approach to predict the link in OppNets. We utilize the deep belief network (DBN) algorithm for the link prediction. Katz Index, Resource Allocation, Adamic Adar, and Common Neighbor metrics are used separately as input vectors to the DBN algorithm, where the input vectors contain only unrepeated links. Two different time slot of 10 and 30 min used to divide the dataset into snapshots in a structure of matrix. Therefore, the results obtained from the experiment shows that the proposed method comes up with more precise prediction than applying each metrics individually. Katz-DBN has shown high precision with 98% depending on 30 min time slot, while in 10 min time slot the precision is 97%.