Data Fusion in IoT

Data fusion is the process of integrating multiple data sources to produce more consistent, accurate, and useful information than that provided by any individual data source (i.e. any individual IoT).

Data fusion in the context of the Internet of Things (IoT) is a crucial technique that enhances the richness of information collected from various IoT devices.

Let us delve into data fusion in some details:

1. Multi-Sensor Data Fusion:

• Objective: The primary goal of multi-sensor data fusion is to integrate data from multiple sensors that monitor the same target or phenomenon.

• Purpose: By analyzing and synthesizing this combined data, we can achieve a more comprehensive and accurate understanding of the target being observed.

• Application: This technique is widely used in IoT scenarios where data from different sensors need to be combined to make informed decisions.

1. Levels of Data Fusion in IoT:

• Low-Level Data Fusion:

• Raw data generated by physical objects (such as sensors) are directly fed into the fusion process.

• The aim is to provide better information by combining these raw measurements.

• Middle-Level Data Fusion:

• Different features extracted from heterogeneous raw data are fused.

• Relevant features are identified by applying diverse data fusion methods.

• High-Level Data Fusion:

• Decisions are fused from input decisions to obtain the most optimum one.

• This level considers the overall context and provides a holistic view of the situation.

1. DFIOT: Data Fusion for the Internet of Things:

• Methodology: A novel data fusion method called DFIOT has been proposed in reference [1].

• Approach:

• DFIOT relies on the Dempster–Shafer (D–S) theory and an adaptive weighted fusion algorithm.

• It considers the reliability of each device in the network and resolves conflicts between devices during data fusion.

• Benefits:

• Improved reliability, accuracy, and conflict management compared to other methods.

• Achieves up to 99.18% accuracy on benchmark artificial datasets and 98.87% accuracy on real datasets with minimal conflict.

• Application perspective: Significant energy savings (up to 90%) when using DFIOT.

In summary, data fusion in IoT plays a pivotal role in handling the vast and diverse data generated by interconnected devices, leading to better decision-making and more efficient resource utilization.

References:

1. DFIOT: Data Fusion for Internet of Things (https://link.springer.com/article/10.1007/s10922-020-09519-y)

2. Data Fusion and Management in IoT: Enhancing Information Accuracy and Consistency (https://dzone.com/articles/data-fusion-and-management-in-iot-enhancing-inform)