Cloud-LoRa: Enabling Cloud Radio Access LoRa Networks Using Reinforcement Learning Based Bandwidth-Adaptive Compression

Muhammad Osama Shahid, Daniel Koch, Jayaram Raghuram, and Bhuvana Krishnaswamy, University of Wisconsin-Madison; Krishna Chintalapudi, Microsoft Research; Suman Banerjee, University of Wisconsin-Madison

21st USENIX Symposium on Networked Systems Design and Implementation (NSDI 24)

This research paper introduces Cloud-LoRa, a novel approach to enhancing the performance of LoRa networks by leveraging cloud computing principles.

Key Innovations in Edge Computing Context:

Cloud Radio Access for LoRa: Cloud-LoRa shifts signal processing tasks from resource-constrained edge devices (LoRa gateways) to the cloud. This offloads computationally demanding tasks like:
- Signal detection and decoding: Leveraging powerful cloud servers for more robust signal processing, improving reception sensitivity and reducing error rates.
- Joint decoding across multiple gateways: Combining signals from multiple gateways in the cloud enables more accurate decoding and improved coverage.
Bandwidth-Adaptive Compression: To address the increased bandwidth requirements of cloud-based processing, Cloud-LoRa incorporates a reinforcement learning-based compression algorithm. This dynamically adjusts the compression rate based on real-time network conditions, optimizing data transmission and minimizing backhaul costs.
Improved Performance: Cloud-LoRa demonstrates significant improvements in signal-to-noise ratio (SNR) gains and network throughput compared to traditional LoRa deployments. This translates to:
- Extended coverage: Reaching devices further away from gateways.
- Increased reliability: Enhancing the robustness of LoRa communication in challenging environments.
- Improved energy efficiency: Reducing energy consumption at the edge by offloading processing tasks.

In essence, Cloud-LoRa exemplifies how edge computing principles can be applied to revolutionize the performance of existing wireless technologies like LoRa. By leveraging cloud resources and intelligent compression techniques, it opens up new possibilities for IoT applications that require extended coverage, improved reliability, and reduced energy consumption.

Read the paper here.