An SOCP Estimator for Hybrid RSS and AOA Target Localization in Sensor Networks.

Area of Science:

• Wireless Sensor Networks
• Localization Algorithms
• Signal Processing

Background:

• Target localization in 3D WSNs is challenging, especially with unknown target transmit power.
• Existing methods for unknown transmit power often involve recursive computations or high computational costs, leading to increased execution times.

Purpose of the Study:

• To develop a computationally efficient, single-iteration algorithm for target localization in 3D WSNs.
• To address the challenge of unknown target transmit power without recursive methods or excessive computational load.

Main Methods:

• A hybrid approach utilizing Angle of Arrival (AOA) and Received Signal Strength (RSS) measurements.
• Derivation of a non-convex least squares (LS) estimator, transformed into a convex one using second-order cone programming (SOCP) relaxation.
• Exploitation of the angle between vectors via inner product definition to enhance estimator tightness.

Main Results:

• A single-iteration solution with moderate computational complexity is proposed, avoiding lengthy execution times.
• The proposed method achieves performance comparable to more complex state-of-the-art methods.
• Outperforms existing methods for a small number of anchor nodes (N), which is practically significant for reducing network costs.

Conclusions:

• The developed algorithm offers an efficient and accurate solution for target localization in 3D WSNs with unknown transmit power.
• The SOCP relaxation and inner product exploitation provide a tighter and computationally feasible estimator.
• The method's effectiveness with fewer anchor nodes makes it valuable for practical WSN deployments.