@inproceedings{7fc5b9d940f644fdbdd6247aed011640,
title = "Calibrating a 4-channel Fully-Digital 60 GHz SDR",
abstract = "The Pi-Radio v1 software-defined radio (SDR) features a 4-channel fully-digital transceiver board operating in the 57-64 GHz band; when mated with the Xilinx RFSoC-based ZCU111 board, this forms a powerful SDR that can be used by the research community. This paper describes the calibration procedures for the SDR, with a special emphasis on not relying on expensive laboratory equipment / infrastructure like a spectrum analyzer, signal generator, or even an anechoic chamber. We hope this is an interesting read for those interested in the pipeline from hardware design schematic to fully functional SDR. ",
keywords = "SDR, beamforming, calibration, fully-digital, mmWave",
author = "Aditya Dhananjay and Kai Zheng and Jaakko Haarla and Lorenzo Iotti and Marco Mezzavilla and Dennis Shasha and Sundeep Rangan",
note = "Funding Information: rotated node trx-0002 counter clockwise by about 15 degrees, and repeated the experiment; the results are shown in Fig. 9, showing that the correct AoA and AoD have been detected. Finally, we returned trx-0002 to the original position, but rotated trx-0003 clockwise by about 25 degrees; the results are shown in Fig. 10; again, the correct AoA and AoD were detected. These experiments Getting the Pi-Radio v1 SDR to work correctly needed careful calibration. There are many ways in which the behavior of practical devices deviates from ideal, and these need to be calibrated out in order to get beams: a) crystal frequency offset correction; b) identification of linear operating ranges; c) timing offset corrections; d) LO phase offset corrections; e) magnitude corrections; f) IQ gain imbalance corrections; and g) IQ quadrature LO phase imbalance corrections. The goal of this paper is to be a tutorial in how to carefully calibrate an SDR node. Performing calibration has hitherto relied on expensive lab bench equipment like spectrum analyzers and signal synthesizers; in the mmWave frequency range, these are painfully expensive. This paper aims to demonstrate one set of techniques that can be used to calibrate SDRs in an affordable manner. We do not claim scientific novelty; the techniques described are well-known or simple enough to be obvious. All calibration code (as well as Pi-Radio{\textquoteright}s v1 SDR hardware design schematics) have been released on GitHub [3] using the free and highly permissive MIT license; anybody can use it in any way that they choose. It is our vision to democratize access to experimental wireless research not only through affordable SDRs that feature advanced transceiver technologies (like fully-digital), but also through open sourcing calibration code that can be used by the community either on the Pi-Radio or any other SDR platform. ACKNOWLEDGEMENTS The Pi-Radio v1 SDR hardware was designed, built, and tested (along with early calibration efforts) during the performance period of NSF STTR Phase-I Award #1821150, A Fully-Digital Transceiver Design for mmWave Communications (Phase-II pending). More advanced calibration techniques were developed and implemented during the performance period of the ARMY STTR Phase-I Award, Millimeter Waveforms For Tactical Networking #W911NF20P0038. An early version of the SDR (based on a different architecture) was developed at New York University, partly funded by a NIST grant, An End-to-End Research Platform for Public Safety Communications above 6 GHz (Award # 70NANB17H166). Publisher Copyright: {\textcopyright} 2020 ACM.; 14th ACM Workshop on Wireless Network, Testbeds, Experimental Evaluation and CHaracterization, WiNTECH 2020 ; Conference date: 25-09-2020 Through 25-09-2020",
year = "2020",
month = sep,
day = "21",
doi = "10.1145/3411276.3412195",
language = "English (US)",
series = "WiNTECH 2020 - Proceedings of the 14th ACM Workshop on Wireless Network Testbeds, Experimental evaluation and CHaracterization, Part of ACM MobiCom 2020",
publisher = "Association for Computing Machinery",
pages = "40--47",
booktitle = "WiNTECH 2020 - Proceedings of the 14th ACM Workshop on Wireless Network Testbeds, Experimental evaluation and CHaracterization, Part of ACM MobiCom 2020",
}