Dynamic Cooperative Communications with Mutual Information Accumulation for Mobile Robots in Industrial Internet of Things

Daoyuan Sun, Zefan Liu, Xinming Zhang

TL;DR

This paper introduces a novel dynamic cooperative communication algorithm for mobile robots in industrial settings. It leverages mutual information accumulation and rateless coding to reduce communication delays, a critical challenge for mobile robots in the Industrial Internet of Things (IIoT). The proposed two-step algorithm efficiently adapts to robot mobility, outperforming existing heuristic approaches in simulation.

What Problem Does This Paper Address?

The Industrial Internet of Things (IIoT) relies heavily on mobile robots for tasks like monitoring, data collection, and material handling. These robots need seamless, low-latency communication with the cloud for receiving instructions and transmitting data. However, their mobility introduces significant challenges:

  • Dynamic Topology: As robots move, the communication network topology changes constantly, requiring rapid adaptation of routing paths.
  • Delay Sensitivity: Industrial applications often have strict latency requirements (40-500 ms) for mobile robots.
  • Limited Onboard Resources: Mobile robots have limited computing power, making complex routing calculations on-device impractical.

Existing solutions often struggle to balance these factors, either by not adequately addressing mobility or by incurring high computational overhead. This paper aims to bridge this gap by enabling efficient, low-latency communication for mobile robots in IIoT environments by utilizing existing sensor nodes as relays.

Methods

The core of the paper is a dynamic cooperative transmission algorithm designed for mobile robots in IIoT. This algorithm is built upon two key concepts:

  1. Mutual Information Accumulation: Instead of discarding received data that isn't immediately decodable, nodes accumulate "mutual information" from multiple transmissions. This allows for faster decoding once sufficient information is gathered, reducing the need for retransmissions and thus lowering delay.
  2. Rateless Coding: This type of forward error correction (FEC) coding allows nodes to decode the original data after receiving any sufficient amount of coded packets. This is crucial for mutual information accumulation, as it means nodes can combine information from various sources without strict synchronization or packet ordering.

The proposed algorithm is a two-step approach:

  • Static Part (Fixed Deployment Node Routing Algorithm): This part calculates an initial, low-delay routing path from the sink node to all fixed sensor nodes in the WSN. It uses a modified Dijkstra algorithm that accounts for the information accumulation process, rather than just summing hop delays. This step is performed infrequently.
  • Dynamic Part (Dynamic Path Selection): This part handles the mobility of the robot.
  • Sink to Mobile Robot: The algorithm leverages the pre-calculated static paths. It searches for neighboring sensor nodes of the mobile robot and selects the path with the shortest delay to the robot.
  • Mobile Robot to Sink: A distributed greedy algorithm is employed. The mobile robot broadcasts information, and neighboring nodes compete to relay it. The node with the highest priority (determined by its pre-calculated delay from the sink) and that successfully decodes the information takes precedence. This approach aims for low time complexity.

Datasets

Not discussed in the paper. The authors used simulations in MATLAB R2020b with randomly generated topologies.

Key Insights and Ideas

  • Leveraging WSNs as Relays: The paper effectively demonstrates how existing fixed sensor nodes in an IIoT environment can be repurposed as intelligent relays for mobile robots, avoiding the need for dedicated infrastructure.
  • Mutual Information Accumulation for Mobility: The core innovation is applying mutual information accumulation not just for static WSNs but specifically to address the challenges posed by mobile robots, where dynamic pathfinding is essential.
  • Rateless Coding's Role: Rateless codes are identified as a key enabler for mutual information accumulation, allowing nodes to flexibly combine information from diverse sources without strict packet ordering or retransmission requests.
  • Two-Step Algorithm for Efficiency: The separation into a static (computationally intensive but infrequent) and a dynamic (lightweight and frequent) part is a smart strategy to handle both network infrastructure and robot mobility efficiently.
  • Greedy Approach for Mobile-to-Sink: The distributed greedy algorithm for the mobile robot to sink communication is a pragmatic choice to keep time complexity low, acknowledging that perfect optimality might be sacrificed for real-time responsiveness.

Key Results

  • Reduced Delay: Simulations show that the proposed dynamic cooperative transmission algorithm achieves lower end-to-end delay compared to existing heuristic algorithms (Heuristic 1 and Heuristic 2) for both static and dynamic scenarios.
  • Performance with K: For the dynamic path selection (sink to mobile robot), increasing the parameter K (number of neighbors considered) initially reduces delay, stabilizing around K=15-20, indicating a good balance between exploration and computational cost.
  • Mobile-to-Sink Efficiency: The distributed greedy algorithm for mobile-to-sink communication consistently shows lower average delay than the static route, highlighting its effectiveness in dynamic scenarios.
  • Time Complexity: The algorithm is designed to have lower time complexity, particularly for the dynamic parts, making it suitable for real-time adaptation in mobile robot scenarios. The fixed deployment part has a complexity of O(n³), while the dynamic parts are significantly lower (e.g., O(Km) or O(1)).

Limitations and Future Work

  • Channel Quality Variability: The authors acknowledge that in complex industrial environments, wireless channel quality can be unpredictable and dynamically changing, posing a challenge for routing algorithms due to hysteresis. They suggest increasing the update frequency of the routing algorithm for fixed nodes to mitigate this.
  • Theoretical Upper Bound: The paper uses Shannon's theorem for information accumulation rate, which provides a theoretical upper bound. Future work could explore efficient rateless coding schemes that balance decoding delay and encoding overhead, and analyze the impact of packet loss.
  • Single Robot Abstraction: The current study abstracts the problem to a single mobile robot and a single sink node. Future research should extend the algorithm to handle multiple mobile robots, which introduces significant complexity in MAC protocols and concurrent transmission management.
  • MAC Protocols: The paper doesn't delve into specific MAC protocols for managing concurrent transmissions, especially with multiple mobile robots. This is identified as an area for future exploration.

Summary updated

Abstract

Mobile robots play an important role in the industrial Internet of Things (IIoT); they need effective mutual communication between the cloud and themselves when they move in a factory. By using the sensor nodes existing in the IIoT environment as relays, mobile robots and the cloud can communicate through multiple hops. However, the mobility and delay sensitivity of mobile robots bring new challenges. In this paper, we propose a dynamic cooperative transmission algorithm with mutual information accumulation to cope with these two challenges. By using rateless coding, nodes can reduce the delay caused by retransmission under poor channel conditions. With the help of mutual information accumulation, nodes can accumulate information faster and reduce delay. We propose a two-step dynamic algorithm, which can obtain the current routing path with low time complexity. The simulation results show that our algorithm is better than the existing heuristic algorithm in terms of delay.
BibTeX
@inproceedings{sun2024dynamic,
  author = {Daoyuan Sun and Zefan Liu and Xinming Zhang},
  title = {{Dynamic Cooperative Communications with Mutual Information Accumulation for Mobile Robots in Industrial Internet of Things}},
  booktitle = {Sensors},
  year = {2024},
  url = {https://openreview.net/forum?id=BLj02cTeZd},
}
Figure