Entropy Correlation and Its Impacts on Data Aggregation in a Wireless Sensor Network

Area of Science:

• Computer Science
• Network Engineering
• Information Theory

Background:

• Wireless Sensor Networks (WSNs) offer potential for efficient communication protocols.
• Existing correlation models are often linear and distance-dependent, limiting their applicability.
• Exploiting correlation in WSNs requires robust correlation models.

Purpose of the Study:

• To propose a general, distance-independent entropy correlation model for WSNs.
• To evaluate the model's effectiveness in data aggregation schemes.
• To develop methods for data compression and representative data selection in WSNs.

Main Methods:

• Developed a general distance-independent entropy correlation model based on joint entropy and group size.
• Estimated the model using individual member entropy and pair-wise entropy correlation coefficients.
• Applied the model to analyze data compression and representative data schemes in WSNs.

Main Results:

• The proposed entropy correlation model effectively evaluates data aggregation schemes.
• Identified optimal routing strategies for data compression within WSNs.
• Proposed a method for determining the number and selection of representative nodes under distortion constraints.

Conclusions:

• The novel entropy correlation model is effective for WSNs.
• The model facilitates the development of efficient data reduction schemes.
• Practical validation confirms the model's utility and the effectiveness of the proposed schemes.