Optimizing Resources and Increasing the Coverage of Internet-of-Things (IoT) Networks: An Approach Based on LoRaWAN

Area of Science:

• Computer Science
• Telecommunications Engineering
• Network Optimization

Background:

• Long Range Wide Area Networks (LoRaWANs) are crucial for low-power, wide-area IoT applications.
• Optimizing resource allocation in LoRaWANs is essential for reducing implementation and maintenance costs.
• Existing methods may not adequately address dynamic network conditions and energy efficiency.

Purpose of the Study:

• To propose a novel resource optimization methodology for LoRaWANs.
• To minimize total energy per useful bit (Ebit) and data collection time.
• To evaluate the impact of device scaling and repeater placement on network performance.

Main Methods:

• Variable Neighborhood Search (VNS) metaheuristic for determining optimal repeater number and location.
• Minimum-cost spanning tree algorithm for network cost reduction.
• Adjustment of parameters like Spread Factor (SF), bandwidth, and transmission power.

Main Results:

• VNS effectively determined repeater placement and quantity, enhancing coverage.
• Small SF values minimize Ebit in networks with few devices; large SFs are beneficial for extended reach.
• Device scaling over time necessitates equal time slot distribution per SF in high-node-count networks to ensure fair data collection.

Conclusions:

• The proposed methodology significantly reduces LoRaWAN implementation and maintenance costs.
• Optimal SF selection is critical and depends on network size and coverage requirements.
• Dynamic device scaling must be considered in fitness functions for robust LoRaWAN performance.