An Optimization Routing Algorithm for Green Communication in Underground Mines

Heng Xu, Qiyue Li, Jianping Wang, Guojun Luo, Chenghui Zhu, Wei SunSensors 2018 · DOI

TL;DR

This paper proposes a novel optimization routing algorithm (ORAGC) for green communication in underground mines. By combining a Degree Constraint Minimum Spanning Tree (DCMST) approach for optimal routing path selection with a Genetic Algorithm (GA) for efficient node clustering, ORAGC significantly reduces energy consumption and extends network lifetime in dynamic, heterogeneous underground environments.

What Problem Does This Paper Address?

Underground mines rely on sensor networks for safety, production monitoring, and localization. These networks often consist of numerous mobile sensor nodes powered by batteries, making energy efficiency (green communication) a critical challenge. The harsh and dynamic nature of mines, with their complex geological structures and potential for accidents, further complicates network management. Existing solutions often fail to adequately consider the heterogeneity of nodes (varying initial energy, functionality) and the complex environmental factors, leading to suboptimal energy consumption and reduced network lifespan. This paper aims to address this by developing a routing and clustering strategy that accounts for these heterogeneous factors to maximize network longevity.

Methods

The proposed ORAGC algorithm employs a two-stage approach:

  1. Degree Constraint Minimum Spanning Tree (DCMST): This algorithm is used to identify optimal routing paths. DCMST considers not only the minimum edge weights (representing energy consumption and transmission distance) but also imposes degree constraints on nodes. This helps to prevent network congestion and excessive energy drain from multi-path connections, ensuring a more robust and energy-efficient path selection. The weights are derived from energy consumption models that account for data acquisition, processing, transmission, and idle states.
  2. Genetic Algorithm (GA): Following the DCMST-based routing path identification, a GA is employed for network clustering. The GA optimizes the cluster structure by treating chromosomes as cluster configurations and genes as nodes within clusters. It considers heterogeneous factors such as residual energy, node roles (anchor, multifunction, ordinary), and connectivity to balance energy consumption and nodal loads across the network. The GA uses crossover and mutation operations to explore the solution space and find an optimal clustering arrangement that minimizes overall energy consumption and extends network lifetime.

Datasets

Not discussed in the paper. The authors describe a simulated underground mine environment with specific parameters for tunnels, intersections, node types, and transmission ranges.

Key Insights and Ideas

  • Heterogeneity is Key: The paper strongly emphasizes that simply optimizing for one factor (like initial energy) is insufficient for underground mine networks. Accounting for diverse node capabilities, energy levels, and environmental dynamics is crucial for true green communication.
  • DCMST for Robust Routing: Using DCMST for routing goes beyond basic MST by incorporating degree constraints. This is vital in dynamic environments to prevent single nodes from becoming bottlenecks or energy drains, thereby ensuring better connectivity and lower overhead.
  • GA for Adaptive Clustering: The GA's ability to handle complex, multi-objective optimization makes it well-suited for dynamically clustering heterogeneous nodes. It allows for flexible cluster head selection, even enabling mobile nodes to act as cluster heads when anchor nodes fail, a critical consideration in accident scenarios.
  • Integrated Approach: The synergy between DCMST for pathfinding and GA for clustering is a core insight. This combined approach ensures that routing paths are energy-efficient and that the resulting clusters are balanced and sustainable.

Key Results

The ORAGC algorithm demonstrated superior performance compared to several existing routing algorithms (ETLE, HEER, EEUC, ERP, DCHGA) across four simulated scenarios representing different underground mine environments (long tunnels vs. intersections, intact vs. damaged anchor nodes).

  • Network Lifetime Extension: ORAGC significantly extended network lifetime. For instance, in scenario 1, the first node failed after 1910 rounds and the last node after 14,780 rounds, showing improvements of 7.3% and 15.02% respectively over the DCHGA algorithm. Similar improvements were observed across all scenarios.
  • Energy Consumption Reduction: The algorithm effectively balanced energy consumption, preventing premature node failures and thus prolonging the overall network lifespan.
  • Scenario Impact: The results indicated that the lack of anchor nodes had a more significant impact on network lifetime than the number of intersections, highlighting the importance of adaptable clustering strategies.

Limitations and Future Work

  • Simulation-Based: The results are based on simulations, and real-world deployment in actual underground mines would be necessary to validate these findings fully.
  • Parameter Tuning: While the paper mentions setting parameters like mutation and crossover probabilities, the sensitivity of the algorithm to these parameters in diverse real-world conditions is not deeply explored.
  • Scalability: While the time complexity is analyzed, further investigation into the algorithm's performance with extremely large-scale networks could be beneficial.

The authors suggest that their approach provides a solid foundation for green communication in challenging environments like underground mines. Future work could involve exploring more advanced energy consumption models, incorporating real-time environmental data for dynamic re-clustering, and investigating the impact of different communication protocols.

Summary updated

Abstract

With the long-term dependence of humans on ore-based energy, underground mines are utilized around the world, and underground mining is often dangerous. Therefore, many underground mines have established networks that manage and acquire information from sensor nodes deployed on miners and in other places. Since the power supplies of many mobile sensor nodes are batteries, green communication is an effective approach of reducing the energy consumption of a network and extending its longevity. To reduce the energy consumption of networks, all factors that negatively influence the lifetime should be considered. The degree constraint minimum spanning tree (DCMST) is introduced in this study to consider all the heterogeneous factors and assign weights for the next step of the evaluation. Then, a genetic algorithm (GA) is introduced to cluster sensor nodes in the network and balance energy consumption according to several heterogeneous factors and routing paths from DCMST. Based on a comparison of the simulation results, the optimization routing algorithm proposed in this study for use in green communication in underground mines can effectively reduce the network energy consumption and extend the lifetimes of networks.
BibTeX
@inproceedings{xu2018optimization,
  author = {Heng Xu and Qiyue Li and Jianping Wang and Guojun Luo and Chenghui Zhu and Wei Sun},
  title = {{An Optimization Routing Algorithm for Green Communication in Underground Mines}},
  booktitle = {Sensors},
  year = {2018},
  doi = {10.3390/s18061950},
}
Figure