On the capacity of UWB-based wireless sensor networks

Yi Shi, Yiwei Thomas HouComputer Networks 2008 · DOI

TL;DR

This paper tackles the fundamental question of maximum data throughput (network capacity) in Ultra-Wideband (UWB) wireless sensor networks. The authors develop a cross-layer framework to jointly optimize routing, scheduling, power control, and rate assignment, deriving closed-form expressions for network capacity in both single-hop and multi-hop scenarios, and establishing an upper bound for arbitrary network configurations.

What Problem Does This Paper Address?

Wireless sensor networks (WSNs) are increasingly important for applications like surveillance and monitoring, where data from numerous sensors needs to be reliably transmitted to a central base station. Ultra-Wideband (UWB) technology offers unique advantages for WSNs, including simple design and low power spectral density. However, a critical challenge is determining the maximum amount of data that can be collected by the base station, especially given the potential bottleneck near the sink. This paper addresses this by investigating the network capacity of UWB-based sensor networks, considering both specific network layouts and general bounds across all possible configurations. They aim to answer: "What is the maximum aggregate data rate a base station can receive, and how can we achieve it?"

Methods

The authors employ a cross-layer optimization approach, integrating aspects of routing, scheduling, power control, and rate assignment. For a given network instance, they first analyze the single-hop scenario where all sensors directly communicate with the base station. Here, they show that maximum capacity is achieved when nodes transmit at maximum power across all available sub-bands. This insight is then extended to multi-hop networks. A key contribution is demonstrating that for multi-hop networks, the maximum capacity is achieved when only nodes within one hop of the base station transmit, effectively reducing the problem to a single-hop scenario. They derive closed-form expressions for network capacity and optimal rate assignments. Finally, they analyze how network capacity changes with the number of nodes and establish a theoretical upper bound for capacity across any arbitrary network instance.

Datasets

Not discussed in the paper. The paper focuses on theoretical analysis and provides examples with simulated network instances.

Key Insights and Ideas

  • Cross-layer optimization is crucial: The paper highlights that achieving maximum network capacity in UWB sensor networks requires a holistic approach, jointly considering routing, scheduling, power control, and rate assignment, rather than optimizing each layer in isolation.
  • Single-hop simplicity for multi-hop capacity: A surprising and elegant result is that for multi-hop UWB sensor networks, the maximum network capacity is achieved by effectively ignoring all nodes beyond the first hop to the base station. This simplifies the complex multi-hop problem to a single-hop one.
  • Maximum power for maximum capacity: In both single-hop and multi-hop scenarios (for the relevant nodes), the optimal strategy for maximizing network capacity involves transmitting at the maximum allowed power on all available sub-bands.
  • UWB's interference resilience matters: The paper implicitly leverages UWB's inherent low interference characteristics, which allows for higher node densities and capacities compared to traditional narrow-band networks without capacity degrading as drastically.
  • Capacity increases with density (up to a point): The study demonstrates that adding more sensor nodes, particularly as one-hop neighbors to the base station, generally increases the overall network capacity, though this increase is bounded.

Key Results

  • Closed-form capacity expressions: The paper provides explicit mathematical formulas for the maximum network capacity in both single-hop and multi-hop UWB sensor networks for a given network instance.
  • Optimal rate vector identification: For each scenario, the authors identify the specific rate at which each node should transmit to achieve the maximum capacity.
  • Network capacity bound: An upper bound for the network capacity of any arbitrary UWB sensor network instance is derived as $W \log_2(e)$, where $W$ is the total spectrum bandwidth. This signifies that capacity doesn't grow indefinitely with node density.
  • Examples with simulated networks: The paper includes numerical examples for 20-node and 30-node networks, illustrating the calculation of capacity and optimal rates, and showing how capacity increases with node addition. For a 20-node network, the capacity was found to be 317.96, increasing to 541.88 for a 30-node network.

Limitations and Future Work

  • Idealized assumptions: The analysis relies on certain idealized assumptions, such as perfect channel state information and simplified propagation models. Real-world scenarios might introduce complexities not fully captured.
  • Focus on aggregate rate: The primary metric is the aggregate data rate at the base station. While important, other performance metrics like latency, fairness, or energy efficiency are not the main focus.
  • Future work: The authors suggest exploring more complex network topologies, considering energy constraints for sensor nodes, and investigating dynamic network scenarios. Further research could also delve into the impact of different UWB modulation schemes and interference models.

Summary updated

BibTeX
@article{shi2008capacity,
  author = {Yi Shi and Yiwei Thomas Hou},
  title = {{On the capacity of UWB-based wireless sensor networks}},
  journal = {Computer Networks},
  year = {2008},
  doi = {10.1016/j.comnet.2008.05.015},
}
Figure