MAC Privacy protection ---------------------- The 802.1 Security Task Group is proposing an amendment to IEE Std 802.1AE-2018 to specify MAC Privacy protection. We know that adversaries can correlate frame sizes and frame transmission timing with user identities, specific applications, and even (in some cases) detailed communication content. This exposure is in addition to any information available from the frame's MAC addresses. MAC Privacy protection would work in conjunction with MACsec (as currently specified/implemented/deployed) by allowing user data frames and padding to be encapsulated within larger frames that are MACsec confidentiality protected and transmitted between privacy protecting peers. This is clearly not a free lunch: packing and padding adds delay and consumes bandwidth. Privacy protection is most attractive on links with an end to end delays that are multiples of on-the-wire transmission times. That characteristic holds for some important use cases, where frames are most exposed. A privacy protecting transmitter can balance reducing the correlation and fingerprinting opportunity offered to an unauthorized observer with its own desire to minimize the quality of service impact. At one extreme, all encapsulating frames can be sent at fixed intervals, with a fixed size, packing small frames into the largest frame size, and sending frames comprising nothing but padding when there is nothing to transmit. That should remove an adversary's ability to make deductions based simply on the level of traffic, and may be appropriate when unused bandwidth would just be wasted. At the other extreme, padding frames to some multiple octet size may be effective in concealing important content such as the sizes of specific application fields. A privacy protecting receiver can recover the encapsulated frames without having to know the transmission strategy, however simple or complicated. In the recently concluded Salt Lake City interim we produced a first cut at a PAR, to summarize the proposed work for 802.1 and other potential participants. In July we plan to ask 802.1 for permission to develop the PAR in the September interim and to pre-circulate that to 802.0 in time for approval in the November plenary. Our first cut can be found here: http://www.ieee802.org/1/files/public/docs2019/dk-seaman-mac-privacy-protection-draft-par-0519-v00.pdf MAC Privacy protection should fit within 802.1AE (technically and editorially) as an additional ISS shim, allowing us to take advantage of existing 802.1Q [just as we did for EDEs (Ethernet Data Encryption devices)] and existing MACsec functionality. Mick Seaman, Chair, 802.1 Security Task Group Background information can be found in: http://www.ieee802.org/1/files/public/docs2019/new-fedyk-traffic-flow-security-0219.pdf http://www.ieee802.org/1/files/public/docs2019/new-fedyk-traffic-flow-security-slides-0519-v00.pdf http://www.ieee802.org/1/files/public/docs2019/new-seaman-mac-privacy-0519-v00.pdf http://www.ieee802.org/1/files/public/docs2018/e-seaman-privacy-in-bridged-networks-1018-v01.pdf