Farmland WSN Incremental Deployment Strategy Based on Signal Attenuation Model

Dayong Liu, Chunjiang Zhao, Huarui Wu, Li Fei-Fei2016 · DOI

Abstract

This article discusses the partition node incremental deployment method in the wireless sensor network, which is divided into partition structure deployment, incremental deployment nodes and node position adjustment. It suggests how the partition node incremental deployment method might influence the running of the farmland wireless sensor networks, and gives illustration of how the connective and stable communication were provided against the instability of the farmland wireless sensor networks.

Introduction

Wireless Sensor Networks (WSNs) are the backbone of precision agriculture, but deploying them effectively in a field is challenging. Most deployment algorithms assume an open and stable environment, whereas real-world farmland is filled with non-uniform obstacles like crops that grow at different heights and densities, significantly attenuating radio signals.

Traditional static or random deployment methods often lead to "blind spots" or inefficient clusters that drain batteries unevenly. This paper proposes the Partition Node Incremental Deployment Method (PNIDM), a three-stage strategy designed to balance structured management with dynamic adaptation to signal propagation realities.

Methodology

The PNIDM approach moves from a rigid structure to a self-organizing one through three distinct steps:

The three stages of the PNIDM process: (a) Partition Structured Deployment, (b) Node Supplement, and (c) Node Position Adjustment.
  1. Partition Structured Deployment: The monitoring area is first divided into a grid of quadrilaterals where initial nodes establish a baseline network.
  2. Node Supplement: The system measures the "communication situation" ($P_{rdc}$) between initial nodes, accounting for average received power (RSSI), packet loss, and receiver sensitivity. If the signal is too weak in a specific sub-region, additional nodes are added to bridge the gap.
  3. Node Position Adjustment: To optimize the final layout, the authors employ a Virtual Force Algorithm (VFA). Nodes are treated as if they are connected by virtual springs. If nodes are too far apart (low signal strength), a "gravitational" force pulls them closer; if they are too close (high signal strength), a "repulsive" force pushes them apart.
Illustration of the Virtual Force Algorithm showing how force vectors from neighboring nodes are superposed to determine the final adjustment of node i.

Experimental Results

The study conducted a real-world experiment in a 100m × 100m corn field using TI CC1101EMK433 wireless modules. The initial grid used 15m quadrangles, with a target packet loss probability (PLP) of 10%.

The proposed algorithm (labeled as DTLDA) was compared against a baseline called KMSA:

  • Energy Efficiency: DTLDA showed significantly lower average node residual energy consumption compared to KMSA over 2500 rounds.
  • Network Longevity: In the first 2000 rounds, DTLDA maintained a higher number of "alive" nodes. The eventual node failure pattern suggested a better power balance across the entire network compared to the baseline.
  • Deployment Density: For the test case, the algorithm determined that 80 nodes (including 44 supplemental nodes) were required to maintain stable connectivity under the measured attenuation conditions.

Key Insights and Future Work

The core innovation of PNIDM is its environment-awareness. By measuring actual packet loss and RSSI in the field, the deployment adapts to specific crop density rather than relying on pure geometry. Using VFA allows the network to self-organize, "nudging" nodes into mathematically ideal positions for both connectivity and longevity.

Future work will address the dynamic nature of crops as they grow taller throughout a season and explore hybrid deployment methods that combine battery-powered nodes with solar-powered nodes to increase system autonomy.

Summary updated

Abstract

This article discusses the partition node incremental deployment method in the wireless sensor network, which is divided into partition structure deployment, incremental deployment nodes and node position adjustment. It suggests how the partition node incremental deployment method might influence the running of the farmland wireless sensor networks, and gives illustration of how the connective and stable communication were provided against the instability of the farmland wireless sensor networks.
BibTeX
@misc{liu2016farmland,
  author = {Dayong Liu and Chunjiang Zhao and Huarui Wu and Li Fei-Fei},
  title = {{Farmland WSN Incremental Deployment Strategy Based on Signal Attenuation Model}},
  year = {2016},
  doi = {10.2991/icmemtc-16.2016.2},
}
Figure