Mastodon Feed: Post

Mastodon Feed

Boosted by jsonstein@masto.deoan.org ("Jeff Sonstein"):
abraxas3d@mastodon.radio ("Abraxas3d W5NYV") wrote:

Here’s a figure from an upcoming article about optimizing Synchronization Words in radio signals.

Another solid contribution to open source digital radio from your friends at @OpenResearchIns 🙂

Article will discuss Barker codes, concatenated Barker codes, maximum length sequences and truncations, Zadoff-Chu sequences, metrics of sync words, and how we chose the sync words for Opulent Voice. See the draft article in ORI's next newsletter.

Sign up at https://www.openresearch.institute/newsletter-subscription/

Visualization of three results. First, when you use a maximum length sequence, produced by a linear feedback shift register, in a situation where it's a one-shot synchronization attempt, you don't get a good result. This is because there are lots of matches for the pattern throughout all the offsets. The offsets come from when you have a pattern you are looking for, and you are monitoring an incoming received data stream. These are false matches, which are equivalent to having high side lobes to a main lobe in an antenna. You might be tempted to point your antenna towards signals in the high side lobes. The second graph shows how a concatenated Barker Code works. It's got a solid high main lobe when the stored pattern matches up exactly with the pattern in the received data stream. But there are a couple of sidelobes that peak up. If there was also a lot of noise, then one of those might give you a false alarm. At the bottom of the graph is the brute-force search sequence. 24 bits that only give a solid 24 bits match at the exact center, when the alignment is perfect. Otherwise, only 3 bits match. This is a great sync word. It's the best we can get, because we did an exhaustive search. Use this one to find the edges of your frame!