Dynamic Resource Optimization for Decentralized Estimation in Energy Harvesting IoT Networks

Claudio Battiloro, Paolo Di Lorenzo, Paolo Banelli, Sergio Barbarossa

Abstract

We study decentralized estimation of time-varying signals at a fusion center, when energy harvesting sensors transmit sampled data over rate-constrained links. We propose dynamic strategies to select radio parameters, sampling set, and harvested energy at each node, with the aim of estimating a time-varying signal while ensuring: i) accuracy of the recovery procedure, and ii) stability of the batteries around a prescribed operating level. The approach is based on stochastic optimization tools, which enable adaptive optimization without the need of apriori knowledge of the statistics of radio channels and energy arrivals processes. Numerical results validate the proposed approach for decentralized signal estimation under communication and energy constraints typical of Internet of Things (IoT) scenarios.

Introduction

In the burgeoning world of the Internet of Things (IoT), sensor networks are often deployed in hard-to-reach places, requiring "set-and-forget" operation. These sensors are increasingly equipped with Energy Harvesting (EH) capabilities (e.g., solar, wind, or vibration) to extend their lifespan indefinitely. However, EH introduces a significant challenge: uncertainty. Traditional static optimization assumes known energy availability and channel conditions, but in reality, both are highly dynamic. This paper addresses the challenge of dynamically managing resources—bits, transmission power, and sampling sets—to maintain a stable network that provides the best possible estimate of a signal over time.

Methodology

The authors frame the problem through decentralized estimation of graph signals. Key technical components include:

  • Probabilistic Quantization: Sensors convert raw data into bits using a probabilistic uniform quantizer to model bandwidth-induced distortion.
  • Bayesian LMMSE Estimation: The Fusion Center reconstructs the global signal using a Linear Minimum Mean-Square Error (LMMSE) estimator.
  • Stochastic Lyapunov Optimization: Using "virtual queues" to represent battery levels, the authors employ a "drift-plus-penalty" function to transform long-term stability goals into simple, per-time-slot decisions.
  • Linearization for Low Complexity: A first-order Taylor expansion of the MSE function transforms non-convex problems into linear ones suitable for low-power IoT hardware.

Key Results and Discussion

Numerical simulations of a 50-node wireless sensor network validate the effectiveness of the proposed strategies. A central finding is the role of the control parameter $V$, which manages the tradeoff between estimation accuracy and battery stability.

BMSE convergence as a function of the control parameter V for different maximum energy harvesting rates.

As illustrated in the figure above, the Bayesian Mean-Square Error (BMSE) approaches the theoretical optimal performance ($BMSE^{opt}$) as $V$ increases. This convergence demonstrates that the system can achieve high precision if the hardware can tolerate larger battery fluctuations (and thus larger physical batteries).

Furthermore, the system exhibits remarkable robustness in non-stationary environments. The authors tested the algorithm under an "ON/OFF" energy harvesting profile to simulate real-world variability.

Real-time adaptation of BMSE during non-stationary energy harvesting conditions.

The temporal results show that the algorithm dynamically "throttles" estimation quality to keep the network alive during energy droughts. When energy harvesting stops, the BMSE increases as the system conserves power; when harvesting resumes, accuracy quickly recovers. This model-free adaptability is a key advantage for practical IoT deployments where environmental statistics are unknown.

Conclusion

This work provides a robust framework for decentralized estimation in EH-powered IoT networks. By leveraging stochastic optimization and linearization, the proposed strategies achieve near-optimal performance while ensuring battery stability. Future research will focus on extending these methods to decentralized machine learning and edge-based anomaly detection.

Summary updated

Abstract

We study decentralized estimation of time-varying signals at a fusion center, when energy harvesting sensors transmit sampled data over rate-constrained links. We propose dynamic strategies to select radio parameters, sampling set, and harvested energy at each node, with the aim of estimating a time-varying signal while ensuring: i) accuracy of the recovery procedure, and ii) stability of the batteries around a prescribed operating level. The approach is based on stochastic optimization tools, which enable adaptive optimization without the need of apriori knowledge of the statistics of radio channels and energy arrivals processes. Numerical results validate the proposed approach for decentralized signal estimation under communication and energy constraints typical of Internet of Things (IoT) scenarios.
BibTeX
@article{battiloro2021dynamic,
  author = {Claudio Battiloro and Paolo Di Lorenzo and Paolo Banelli and Sergio Barbarossa},
  title = {{Dynamic Resource Optimization for Decentralized Estimation in Energy Harvesting IoT Networks}},
  journal = {IEEE Internet of Things Journal},
  year = {2021},
  url = {https://openreview.net/forum?id=Hqp9UbeZBn},
}
Figure