Millimeter-Wave (mmWave) WiFi can provide very low latency and multi-Gbps throughput, but real-world deployments usually do not achieve the theoretically feasible performance. One main source of inefficiency is the contention-based random channel access, as it requires omni-directional reception which limits performance. Additionally, carrier sensing at mmWave frequencies is highly unreliable, leading to reduced channel usage. In this work, we present SIGNalling in the PHY Preamble (SIGNiPHY) for efficient directional communications, a solution that allows to embed user identity in the preamble of data packets. It allows for true early user identification and then immediately steering the beam towards the transmitter while receiving the physical layer preamble. SIGNiPHY enables directional reception in random access mmWave networks, and additionally helps to quickly filter unwanted packets. It does not affect any preamble functions and is backward-compatible with legacy stations. We implement SIGNiPHY on an FPGA-based mmWave testbed and show that it achieves 99.6% decoding accuracy even under very low SINR conditions. We also implement SIGNiPHY in ns-3 to evaluate large networks and show that it achieves throughput gains between 13% and 230% compared to different baseline schemes, due to the lower packet loss rate and improved spatial sharing. SIGNiPHY has been accepted for publication at MobiSys 2023.

About Nina Grosheva

Nina Grosheva is a PhD student in the Wireless Networking Group at IMDEA Networks and in the Telematics program at the University Carlos III de Madrid. Previously, she completed an MSc in Communications Engineering at RWTH Aachen University and a BSc in Electrical Engineering at Saint Cyril and Methodius University in Skopje. Her research interest is in mmWave networks, with a particular focus on MAC and network layer design and analysis.

This event will be conducted in English