An Electrocommunication System Using FSK Modulation and Deep Learning Based Demodulation for Underwater Robots

Qinghao Wang, Ruijun Liu, Wei Wang, Guangming Xie

TL;DR

This paper introduces a novel electrocommunication system for underwater robots that leverages 2FSK modulation and a deep learning-based demodulator. This system significantly extends communication range to over 10 meters at 5 Kbps with low power consumption, addressing a critical bottleneck for small underwater robots.

What Problem Does This Paper Address?

Underwater communication is notoriously difficult for small robots due to stringent power and size constraints. Traditional methods like acoustic and optical communication have limitations such as Doppler shifts, multi-path effects, and susceptibility to water turbidity. Electromagnetic communication is heavily attenuated in water, limiting its range. Electrocommunication, inspired by weakly electric fish, offers a promising alternative by using electric fields. However, previous electrocommunication systems for robots suffered from very short communication distances, limiting their practical application. This paper aims to overcome this limitation by developing a more robust and longer-range electrocommunication system.

Methods

The core of this work lies in two key innovations:

  • 2FSK Modulation: The system employs Binary Frequency Shift Keying (2FSK) modulation. This method uses two distinct frequencies to represent binary data ('0' and '1'), offering better anti-interference capabilities and lower complexity compared to other modulation schemes like ASK.
  • Deep Learning Demodulation: A neural network is used for demodulating the received electric field signals. This deep learning approach is designed to classify the modulated signals as binary data, effectively handling complex environmental noise and improving communication quality. The neural network architecture consists of an input layer, two hidden layers, and an output layer, with a specific number of neurons in each.

Additionally, the authors derived a new mathematical model for electrocommunication that accounts for the electric field distribution outside the near-field region and the influence of frequency on the receiver, improving upon existing dipole field models.

Datasets

Not discussed in the paper. The paper focuses on simulations and experimental validation using self-generated signals.

Key Insights and Ideas

  • Deep Learning for Robustness: Using deep learning for demodulation is a significant departure from traditional methods. It allows the system to learn and adapt to complex underwater noise environments, leading to a substantial reduction in bit error rate, especially at lower signal-to-noise ratios.
  • 2FSK's Practical Advantage: The choice of 2FSK is justified by its balance of anti-interference capability and implementation simplicity, making it suitable for resource-constrained underwater robots.
  • Beyond the Near-Field: The development of a more comprehensive electrocommunication model that extends beyond the near-field region is crucial for understanding and achieving longer communication distances.
  • Power-Distance Trade-off: The paper highlights that increasing transmitting power exponentially increases communication distance, making it impractical for small robots. Their approach focuses on improving signal processing and modulation to achieve longer ranges without excessive power demands.

Key Results

  • Extended Communication Distance: The new system achieves stable communication up to 10 meters. This is a substantial improvement over previous systems, which had much shorter ranges.
  • Data Transfer Rate: The system supports a data transfer rate of 5 Kbps.
  • Low Power Consumption: The communication system operates with a power consumption of less than 0.1 W.
  • Improved Bit Error Rate: Simulations and experiments show that the deep learning demodulator significantly outperforms traditional coherent demodulation, especially in noisy conditions. For instance, at an SNR of -2 dB, the deep learning method achieves a bit error rate of zero, while traditional methods require an SNR greater than 5 dB for the same performance.
  • Model Validation: The derived electrocommunication model was validated against experimental data, showing good agreement in predicting received voltage distribution with distance.

Limitations and Future Work

  • Limited Frequency Bandwidth: The authors acknowledge that the frequency utilization rate of 2FSK modulation is still limited, suggesting room for improvement.
  • Future Directions:
    • Frequency Division Multiplexing: Exploring frequency division multiplexing is proposed as a future direction to improve bandwidth utilization.
  • Deep Learning Framework Optimization: Further research into optimizing the deep learning framework to further reduce the bit error rate is suggested.
  • Large-Scale Cooperative Communication: The system's potential for large-scale underwater cooperative communications of robots is highlighted.

Summary updated

Abstract

Underwater communication is extremely challenging for small underwater robots which typically have stringent power and size constraints. In our previous work, we developed an artificial electrocommunication system which could be an alternative for the communication of small underwater robots. This paper further presents a new electrocommunication system that utilizes Binary Frequency Shift Keying (2FSK) modulation and deep-learning-based demodulation for underwater robots. We first derive an underwater electrocommunication model that covers both the near-field area and a large transition area outside of the near-field area. 2FSK modulation is adopted to improve the anti-interference ability of the electric signal. A deep learning algorithm is used to demodulate the electric signal by the receiver. Simulations and experiments show that with the same testing condition, the new communication system outperforms the previous system in both the communication distance and the data transmitting rate. In specific, the newly developed communication system achieves stable communication within the distance of 10 m at a data transfer rate of 5 Kbps with a power consumption of less than 0.1 W. The substantial increase in communication distance further improves the possibility of electrocommunication in underwater robotics.
BibTeX
@inproceedings{wang2020electrocommunication,
  author = {Qinghao Wang and Ruijun Liu and Wei Wang and Guangming Xie},
  title = {{An Electrocommunication System Using FSK Modulation and Deep Learning Based Demodulation for Underwater Robots}},
  booktitle = {IEEE/RSJ International Conference on Intelligent Robots and Systems},
  year = {2020},
  url = {https://openreview.net/forum?id=sZllmqtIO4I},
}
Figure