Statistical Modeling of Energy Harvesting in Hybrid PLC-WLC Channels

Victor Fernandes, Thiago Nogueira, H. Vincent Poor, Moisés V. RibeiroSustainability 2021 · DOI

TL;DR

This paper develops statistical models for energy harvesting (EH) in hybrid Power Line Communication (PLC) and Wireless Communication (WLC) channels, specifically within the 0-100 MHz frequency band. They found that the log-normal distribution accurately models harvested energy from both the primary signal and noise, enabling easier simulation for energy-efficient network research.

What Problem Does This Paper Address?

The increasing demand for ubiquitous connectivity, driven by 5G, 6G, IoT, and Smart Cities, is leading to a significant surge in energy consumption by telecommunication networks. To combat this, researchers are exploring energy-efficient solutions, including hybrid communication systems that combine different technologies like PLC and WLC. While these hybrid channels offer new communication possibilities, they also present an opportunity for energy harvesting (EH) from the ambient electromagnetic radiation. However, to effectively design and implement EH systems in these hybrid environments, accurate statistical models of the harvested energy are crucial. This paper aims to fill this gap by providing such models for in-home hybrid PLC-WLC channels.

Methods

The core of this research involves statistical modeling of harvested energy. The authors utilized a dataset from a measurement campaign to analyze the energy harvested from both the "free-of-noise" (FoN) received signal and the additive noise within the hybrid PLC-WLC channel. They employed the maximum likelihood value (LLV) criterion to fit various statistical distributions (Log-normal, Beta, Exponential, and Rayleigh) to the harvested energy data. The distribution that yielded the highest LLV was selected as the best model. The study also considered five different "power masks" for power allocation, which dictate how transmission power is distributed across frequencies. The total harvested energy was modeled as the sum of the FoN signal's harvested energy and the additive noise's harvested energy, assuming statistical independence.

Datasets

The study used a dataset comprising 175,428 estimates of hybrid PLC-WLC channel frequency responses (CFRs) and millions of samples of PLC and WLC additive noises. This data was collected from a measurement campaign conducted in seven residences in Juiz de Fora, Brazil. The paper references previous work ([18]) for details on the measurement setup, methodology, and equipment.

Key Insights and Ideas

  • Log-Normal is King for EH: The log-normal distribution emerged as the most accurate statistical model for both the "free-of-noise" received signal's harvested energy and the additive noise in hybrid PLC-WLC channels. This finding simplifies the modeling process significantly.
  • Sum of Log-Normals: The total harvested energy can be effectively modeled as the sum of two independent log-normal random variables, representing the FoN signal and the additive noise. This additive property makes the overall model tractable.
  • Simplicity for Sustainability: By providing these statistical models, the research makes the phenomenon of energy harvesting in hybrid PLC-WLC channels "easy-to-simulate." This is crucial for developing energy-efficient and self-sustainable networks for applications like IoT and Smart Grids.
  • Power Masks Matter: The choice of power mask significantly influences the harvested energy. While some masks are simpler to implement (e.g., uniform distribution), others that leverage channel state information (CSI) can potentially yield better EH performance.
  • Distance Affects EH: The distance between the WLC device and the power outlet (short-path vs. long-path channels) impacts the accuracy of the statistical models, with shorter paths showing better agreement with measured data.

Key Results

  • The log-normal distribution consistently provided the highest LLVs across various power masks and channel types for both FoN received signals and additive noises.
  • The study demonstrated that the statistical models developed show good agreement with the measured data, with discrepancies generally within acceptable limits (e.g., up to 21% in some cases).
  • The research provides specific parameters (mean and standard deviation) for the log-normal distributions under different scenarios, enabling quantitative analysis of harvested energy.
  • The paper shows that the modeling of harvested energy can be used to predict achievable data rates in hybrid PLC-WLC systems, further validating the models' utility.

Limitations and Future Work

  • Model Accuracy and Outliers: The authors acknowledge that statistical modeling inherently has limitations in precisely capturing outliers, which can lead to discrepancies between model predictions and measured data, especially in scenarios with significant noise components.
  • CSI Dependency: Power masks that rely on complete channel state information (CSI) are more theoretical and complex to implement in real-world scenarios compared to simpler uniform power masks.
  • Future Work: The paper suggests that these models can be extended to facilitate further research into energy-efficient and self-sustainable networks for IoT, Smart Grids, and Industry 4.0. The ease of simulation provided by these models is a key enabler for such future investigations. The authors also highlight the potential for ambient EH approaches where transmission power is very low, which could lead to higher relative contributions from additive noise.

Summary updated

Abstract

This work introduces statistical models for the energy harvested from the in-home hybrid power line-wireless channel in the frequency band from 0 to 100 MHz. Based on numerical analyses carried out over the data set obtained from a measurement campaign together with the use of the maximum likelihood value criterion and the adoption of five distinct power masks for power allocation, it is shown that the log-normal distribution yields the best model for the energies harvested from the free-of-noise received signal and from the additive noise in this setting. Additionally, the total harvested energy can be modeled as the sum of these two statistically independent random variables. Thus, it is shown that the energies harvested from this kind of hybrid channel is an easy-to-simulate phenomenon when carrying out research related to energy-efficient and self-sustainable networks.
BibTeX
@inproceedings{fernandes2021statistical,
  author = {Victor Fernandes and Thiago Nogueira and H. Vincent Poor and Moisés V. Ribeiro},
  title = {{Statistical Modeling of Energy Harvesting in Hybrid PLC-WLC Channels}},
  booktitle = {Sustainability},
  year = {2021},
  doi = {10.3390/su14010442},
}
Figure