Floodlight Quantum Key Distribution: A New Framework for High-Rate Secure Communication

Dr. Zheshen Zhang
Massachusetts Institute of Technology
2016-07-15 (周五) 10:00
Bldg. 4, Shanghai branch

The advent of quantum computers will doom public-key cryptography’s RSA encryption standard. Quantum key distribution (QKD), however, offers a solution for the post-RSA era: quantum physics’ no-cloning theorem can ensure safe creation of one-time pads that permit communication with full information-theoretic security. However, propagation loss suffered in long-distance transmission sets an ultimate limit on the producible secret-key bits per optical mode –– known as the Takeoka-Guha-Wilde rate-loss trade-off. Because prevailing QKD systems do not employ many optical modes per channel-use, their secret-key efficiencies in bit/channel-use are vastly less than unity, rendering their secret-key rates in bit/s far less than what will be needed for one-time-pad encryption of large files. In this talk, I introduce floodlight quantum key distribution (FL-QKD), a radically different paradigm that thrives by employing many photons distributed over a huge number of low-brightness optical modes per channel-use to boost both QKD’s secret-key efficiencies and secret-key rates at long distances. We show that the use of multiple optical modes per channel-use, plus photon-coincidence channel monitoring, makes the new protocol capable of a 2 Gb/s secret-key rate over a 50-km fiber link. Our initial proof-of-concept experiment, done with 10-dB propagation loss (equivalent to a 50-km fiber link), demonstrated a secret-key efficiency of 0.55 bit/channel-use, beating the ultimate limit for all single-mode one-way QKD protocols. With a 100 Mb/s modulation rate, we demonstrate a 55 Mb/s secret-key rate against the optimum collective attack. This rate is already a 50-fold improvement over all existing QKD demonstrations. Moreover, FL-QKD could be pushed to the long-sought Gb/s-class secret-key rates at metropolitan distances with available equipment, i.e., no new technology need be developed.