Enabling opportunistic energy trading between overlapping energy harvesting wireless sensor networks

Teng Jiang, Geoff V. Merrett, Nick R. HarrisENSsys@SenSys 2014 · DOI

TL;DR

This paper introduces Opportunistic Energy Trading (OET) for wireless sensor networks (WSNs) that harvest energy. It proposes a framework for inter-network energy trading, allowing networks to share resources to overcome the limitations of uncontrollable energy harvesting and finite battery life, especially during periods of low harvest.

What Problem Does This Paper Address?

Wireless Sensor Networks (WSNs) are increasingly deployed in environments where battery replacement is impossible. Energy harvesting offers a solution for indefinite operation, but harvested energy is often unpredictable, varying significantly in both space and time. Traditional methods manage energy within a single network's boundaries, struggling with these variations. For instance, solar-powered networks face complete energy depletion at night, while battery-powered networks might have excess energy that goes to waste. This paper addresses the challenge of enabling WSNs to go beyond their individual limitations by allowing them to trade energy resources with neighboring networks, thus creating a more robust and resilient system.

Methods

The authors extend their previous work on Opportunistic Energy Trading (OET) by removing simplifying assumptions. They propose a framework for OET that operates at higher layers of the protocol stack, enabling communication and understanding between networks. Key components include:

  • Service Management Module (SMM) and Resource Management Module (RMM): Standardized modules appended to the application layer that allow networks to advertise their services and describe their available resources (e.g., energy, network coverage). This enables networks to discover and understand each other's capabilities.
  • Addressing Translation Module (ATM): This module encapsulates packets with a standardized Cross-Boundary Addressing (CBA) header for cross-network transmissions. It translates these headers into local routing headers, facilitating communication between networks that might use different underlying protocols.
  • Cooperation Management Layer (CML): A cross-layer module responsible for collecting cost-related metrics and configuring lower layers to provide services. It also acts as a bridge for information exchange between layers.
  • Cost Monitoring and Recording: A mechanism to quantify the cost of services provided. The paper proposes using the number of hops taken to route a neighboring network's packets as a proxy for energy cost, allowing for financial reimbursement.

The paper also investigates the impact of varying degrees of overlap between co-located networks, from 0% (no overlap) to 100% (fully overlapping), on the effectiveness of OET.

Datasets

The authors used simulated data for their evaluation. Specifically, they modeled solar energy availability using radiation power data provided by Humboldt State University (40.99°N, 124.08°W, elevation 36m) for the period of 07-11 January.

Key Insights and Ideas

  • Inter-network cooperation is crucial for energy-harvesting WSNs: The unpredictable nature of energy harvesting necessitates looking beyond individual network boundaries for robust energy management.
  • Higher-layer support is essential for practical OET: Simply enabling low-level packet forwarding isn't enough; networks need to understand each other's services and resources to effectively trade energy.
  • Overlap significantly impacts OET effectiveness: The degree of overlap between networks directly influences how effectively energy can be traded. Greater overlap means more potential for inter-network routing and thus better energy sharing.
  • Cost quantification is key for fair trading: Establishing a clear metric for the "cost" of energy services (like routing packets) is vital for a sustainable trading model, allowing for financial reimbursement.
  • OET can significantly improve energy utility in solar-powered networks: By allowing solar networks to "buy" energy at night, their operational lifetime and packet delivery rates can be substantially improved.

Key Results

  • Improved Packet Delivery in Solar Networks: In a simulated case study, a solar-powered network (Network A) experienced a significant increase in successfully delivered packets as the overlap percentage with a battery-powered network (Network B) increased. Without OET (0% overlap), packet delivery dropped drastically overnight due to energy depletion. With 100% overlap, packet delivery was maintained throughout the night.
  • Minimal Impact on Battery-Powered Networks: The battery-powered network (Network B) experienced a slight decrease in residual battery energy (around 2.5% over 5 days with 100% overlap), indicating that it was selling energy to Network A. This cost is recorded to allow for financial reimbursement.
  • Cost Increases with Overlap: The cost recorded by Network B (measured in hops) increased with the overlap percentage, as more packets were routed for Network A.
  • Overlap is a Key Factor: The study demonstrated that the effectiveness of OET is directly proportional to the degree of overlap between networks, with higher overlap leading to better energy sharing and improved performance for the energy-harvesting network.

Limitations and Future Work

  • Limited Scope of Trading: The current work focuses on a solar-powered network buying energy from a battery-powered network. The authors suggest future work on bi-directional OET, where networks can accumulate cost to buy back services.
  • Simplified Routing: The simulations used a collection tree routing protocol without dynamic updates, which is not energy-aware. The authors acknowledge that real-world networks would likely employ more sophisticated routing.
  • Topology Assumptions: While the paper explores different grid topologies (3x8 and 8x3), real-world network topologies can be more complex.
  • Practical Implementation: The authors aim for a practical implementation of OET with real scenarios and topologies in future work.
  • Privacy and Security: The paper notes that privacy and security issues related to cross-network communication are outside the scope of this work but remain areas for future investigation.

Summary updated

BibTeX
@inproceedings{jiang2014enabling,
  author = {Teng Jiang and Geoff V. Merrett and Nick R. Harris},
  title = {{Enabling opportunistic energy trading between overlapping energy harvesting wireless sensor networks}},
  booktitle = {ENSsys@SenSys},
  year = {2014},
  doi = {10.1145/2675683.2675688},
}
Figure