Spray: A Spectrum-efficient and Agile Concurrent Backscatter System

Abstract

Recent works have achieved considerable success in improving the concurrency of backscatter network. However, they do not optimize the balance between throughput and spectrum occupancy, both of which serve as pivotal parameters in concurrent transmissions. Moreover, these works also introduce complex components on tag thereby increasing both power consumption and deployment costs. In this paper, we propose Spray, a tag-lightweight system to achieve high throughput and narrow band occupancy with low power. The key idea is to incorporate an agile channel allocating and scheduling mechanism into the backscatter network. This approach allows for efficient spectrum utilization and concurrency without the need for energy-intensive components. To optimize throughput in the presence of the challenge of harmonic interference, we introduce a novel algorithm that determines the channels with an optimal combination of central frequencies and bandwidths. Additionally, we propose a fair scheduling strategy to ensure equitable transmission opportunities for all tags. We prototype the Spray tag using COTS components and implement the excitation and receiver with software-defined radio (SDR) platform. Our evaluation shows that the system supports 30 parallel tags transmitting in the bandwidth of 600 kHz, and the throughput can reach more than 280 kbps.