Towards Integrating True Random Number Generation in Coherent Optical Transceivers

TL;DR

This paper explores integrating quantum random number generation into existing coherent optical transceivers using vacuum fluctuations, demonstrating proof-of-principle experiments with potential for seamless quantum-classical data integration.

Contribution

It introduces two architectures for quantum entropy sources in coherent transceivers and demonstrates their feasibility through experimental validation.

Findings

Achieved >2 dB optical-electrical noise clearance over 11 GHz bandwidth.

Demonstrated time-interleaved random number generation at 10 Gbaud.

Validated randomness extraction using a two-universal strong extractor.

Abstract

The integration of quantum communication functions often requires dedicated opto-electronic components that do not bode well with the technology roadmaps of telecom systems. We investigate the capability of commercial coherent transceiver sub-systems to support quantum random number generation next to classical data transmission, and demonstrate how the quantum entropy source based on vacuum fluctuations can be potentially converted into a true random number generator for this purpose. We discuss two possible implementations, building on a receiver- and a transmitter-centric architecture. In the first scheme, balanced homodyne broadband detection in a coherent intradyne receiver is exploited to measure the vacuum state at the input of a 90-degree hybrid. In our proof-of-principle demonstration, a clearance of >2 dB between optical and electrical noise is obtained over a wide bandwidth of more than 11 GHz. In the second scheme, we propose and evaluate the re-use of monitoring photodiodes of a polarization-multiplexed inphase/quadrature modulator for the same purpose. Time-interleaved random number generation is demonstrated for 10 Gbaud polarization-multiplexed quadrature phase shift keyed data transmission. The availability of detailed models will allow to calculate the extractable entropy and we accordingly show randomness extraction for our two proof-of-principle experiments, employing a two-universal strong extractor.