Abstract
Modern acoustic wearables with microphone arrays are promising to offer rich experience (e.g., 360° sound and acoustic imaging) to consumers. Realtime multi-track audio streaming with precise synchronization however poses significant challenges to the existing wireless microphone array designs that depend on complex digital synchronization as well as bulky and power-hungry hardware.
This paper presents a novel microphone array sensor architecture that enables synchronous concurrent transmission of multitrack audio signals using analog backscatter communication. We develop novel Pulse Position Modulation (PPM) and Differential Pulse Position Modulation (DPPM) baseband circuits that can generate a spectral-efficient, time-multiplexing, and multi-track-synchronous baseband signal for backscattering. Its lightweight analog synchronization supports parallel multimedia signals without using any ADCs, DSPs, codecs and RF transceivers, hence largely reducing the complexity, latency, and power consumption. To further enhance self-sustainability, we also design an energy harvester that can extract energy from both sound and RF. We have built a microphone array backscatter sensor prototype using an FPGA, discrete components, and analog devices. Our experiments demonstrate a communication range (sensor-to-reader) of up to 28 meters for 8 audio tracks, and an equivalent throughput of up to 6.4 Mbps with a sample rate over 48KHz. Our sensor achieves 87.4μs of streaming latency for 4 tracks, which is 650x improvement as compared with digital solutions. ASIC design results show that it consumes as low as 175.2μW of power. Three sample applications including an acoustic imaging system, a beamform filter, and a voice control system, all built with our phased-array microphone, further demonstrate the applicability of our design.
Authors
Jia Zhao, Wei Gong, Jiangchuan Liu
Publication:
ACM MobiCom (CCF-A) [Link] [PDF]
Keywords:
Microphone Array, Backscatter, Spatial Sound, Internet of Things
