Experimental Validation of Cooperative RSS-based Localization with Unknown Transmit Power, Path Loss Exponent, and Precise Anchor Location

Yingquan Li, Bodhibrata Mukhopadhyay, Jiajie Xu, Mohamed‐Slim AlouiniIEEE Trans. Wireless Communications 2024

TL;DR

This paper introduces CTUP (Cooperative Localization Techniques with Unknown Parameters), a robust framework for locating wireless sensor nodes using only Received Signal Strength (RSS). Unlike previous methods that assume perfect environmental knowledge, CTUP simultaneously estimates target locations while accounting for three major real-world "messy" factors: unknown transmit power, fluctuating path loss exponents, and imprecise anchor node coordinates.

What Problem Does This Paper Address?

In Wireless Sensor Networks (WSNs), knowing where a sensor is located is often as important as the data it collects. While GPS is the go-to solution, it’s power-hungry and fails indoors or in "urban canyons." RSS-based localization is a popular alternative because it’s cheap and doesn't require complex time synchronization.

However, standard RSS models are brittle. They typically assume we know exactly how much power a node is transmitting and exactly how fast that signal decays in a specific environment (the Path Loss Exponent, or PLE). Furthermore, they assume "anchor" nodes (nodes with "known" positions) are perfectly placed. In reality, anchor positions are often determined by standard GPS, which can be off by several meters. This paper tackles the challenge of achieving high-accuracy localization when all these variables are unknown or uncertain.

Methods

The authors formulate the localization problem as a Maximum Likelihood (ML) estimation. Because the raw ML problem is non-convex and non-linear (making it computationally "hard" to solve directly), they apply several mathematical relaxation techniques:

  1. Taylor Series Expansion: Used to linearize the relationship between distance and signal strength.
  2. Semidefinite Relaxation (SDR) & Epigraph Method: These transform the "hard" non-convex problem into a "tractable" convex one.
  3. Mixed SDP-SOCP: The resulting problem is solved using a hybrid of Semidefinite Programming (SDP) for precision and Second-Order Cone Programming (SOCP) for computational efficiency.

The authors propose four variants (CTUP-1 through CTUP-4) depending on which parameters (Transmit Power, PLE, or both) are unknown, all while consistently accounting for anchor location uncertainty.

Datasets

A standout feature of this research is its move beyond pure simulation. The authors created an indigenous experimental dataset:

  • Hardware: 50 Bluetooth Low Energy (BLE) nodes.
  • Scale: Deployed over a massive 640 m × 180 m outdoor area.
  • Ground Truth: Node positions were verified using RTK-GPS, providing centimeter-level accuracy.
  • Anchor Uncertainty: To simulate real-world errors, they recorded anchor positions using standard GPS, intentionally introducing the meter-level "noise" typical of consumer-grade positioning.

Key Insights and Ideas

  • Cooperation is Key: By using "target-to-target" links (cooperative localization) rather than just "anchor-to-target" links, the system gains significantly more data points, which helps "average out" the errors from unknown parameters.
  • The "Three Unknowns" Matter: The paper demonstrates that ignoring anchor location uncertainty or assuming a static PLE leads to significant "localization drift." By treating these as variables to be estimated rather than constants, the model becomes much more resilient.
  • Mathematical Tractability: The transition from a non-convex ML estimator to a mixed SDP-SOCP solver allows for high-accuracy results without the prohibitive computational cost of traditional global optimization.

Key Results

  • Superior Accuracy: In both simulations and real-field tests, CTUP consistently outperformed state-of-the-art methods (like IRGDL and FCUP).
  • Noise Resilience: CTUP-1 showed significantly better performance than existing SDP-based methods when signal noise ($\sigma$) exceeded 3 dB, largely because it explicitly mitigated anchor location uncertainty.
  • Real-World Validation: Using their 50-node BLE deployment, the authors showed that as the number of anchor nodes increases, the Normalized Root Mean Squared Error (NRMSE) of the location estimate drops significantly, proving the scalability of the cooperative approach.
  • Parameter Estimation: Beyond just finding locations, CTUP-4 successfully estimated the unknown Transmit Power and PLE with high precision, matching the theoretical Cramer-Rao Lower Bound (CRLB).

Limitations and Future Work

While CTUP is robust against environmental noise and coordinate errors, the authors note a few areas for growth:

  • Computational Complexity: While more efficient than pure SDP, the mixed SDP-SOCP approach still has a higher CPU runtime than simpler (but less accurate) gradient descent methods.
  • Security: The current model assumes all nodes are "honest." The authors suggest that future work should focus on making these techniques resilient against malicious nodes that might transmit false RSS data to spoof the system.
  • Dynamic Environments: While the PLE is estimated, the current model may need further refinement for high-mobility scenarios where the environment changes rapidly.

Summary updated

Abstract

Received signal strength (RSS)--based cooperative localization has gained significant attention due to its straightforward system architectures and cost-effectiveness. In this paper, we propose Cooperative Localization Techniques (with Unknown Parameters), referred to as CTUP(s), which consider uncertainty in anchor nodes' locations and assume the transmit power and \textcolor{blue}{path loss exponent (PLE)} to be unknown. Unlike prior studies, CTUP(s) address unknowns by estimating these parameters, along with the location of target nodes. The non-convex and non-linear nature of the maximum likelihood (ML) estimator of the problem is addressed through relaxation techniques, employing Taylor series expansion, semidefinite relaxation (SDR), and the epigraph method. The resulting problem is solved using semidefinite second-order cone programming (SDP-SOCP), leveraging the precision of SDP and the simplicity of SOCP. We deployed an extensive network comprising 50 BLE nodes covering an area of 640~m $\times$ 180~m to gather RSS data. The precise location of the nodes is obtained using real-time kinematics global positioning system (RTK-GPS), which is treated as the ground truth. Furthermore, to replicate real-world scenarios, we recorded the positions of the anchor nodes using a standard GPS, thereby introducing uncertainty into the anchor node locations. Extensive simulation and hardware experimentation demonstrate the superior performance of CTUP compared to existing techniques.
BibTeX
@article{li2024experimental,
  author = {Yingquan Li and Bodhibrata Mukhopadhyay and Jiajie Xu and Mohamed‐Slim Alouini},
  title = {{Experimental Validation of Cooperative RSS-based Localization with Unknown Transmit Power, Path Loss Exponent, and Precise Anchor Location}},
  journal = {IEEE Trans. Wireless Communications},
  year = {2024},
}
Figure