Casting our nets wider

Recent BACnet committee meetings, and this one especially, are highlighting the interactions across varied areas of technology and knowledge. The arena of IT and Internet Protocols (adding support for IPv6, for example) is one area potentially very broad as we search out our direction -- even our terminology is changing rapidly; just minutes ago BACnet-IT was jestingly renamed to BACnet "New Transport Binding" for the latest session's presentation on a demonstration project, based on the discussion & recommendations in the preceding session in the IT Working Group meeting. But this was to be expected.

A bigger surprise came in yesterday's MS/TP LAN working group meeting, on a discussion on expanding the maximum frame size for this low-cost LAN from 501 bytes to 4096 bytes. This technology has been around for some time and well understood - we thought. But IEEE's Kerry Lynn surprised me (at least) when he revealed that increasing the frame size 8-fold would increase the error rate a thousand-fold!

He referenced studies on CRC algorithms from Dr. Philip Koopman (online here), noting that there's a frequent assumption that "other have done their homework" in areas such as this, when in fact they might not have. Characteristics of the CRC-CCITT and other polynomials have been investigated; the following plot shows one result. Kerry will be providing guidance in this area.