HI version is available. Content is displayed in original English for accuracy.
Advertisement
Advertisement
⚡ Community Insights
Discussion Sentiment
67% Positive
Analyzed from 245 words in the discussion.
Trending Topics
#lower#book#got#noise#concepts#ofdm#digital#communications#goldsmith#cover

Discussion (3 Comments)Read Original on HackerNews
Early versions had a fairly fatal flaw. They responded to bad reception (reception but failure of checksum etc due to interference) by 'downshifting' e.g. reducing their transmission rate from 10Mbit to lower rates.
This seems rational, but in practice it usually shifted all the way to the minimum (1Mbps!) and got no relief. In fact, packets were 10 to 100 times 'longer' in airtime, congesting the channel so nobody got anything through.
Why did they get no relief? Because frequently the noise was bursty e.g. microwave oven pulse or auto spark plug noise. Going slower meant larger (longer to transmit) packets, which provided a 'bigger target' for the noise to poke a hole in.
A more useful response was to try the fastest rate, a smaller packet which got hit with fewer 'darts' impacting any particular packet.