On 12/2/17, Samuel Hunt <s...@maxxwave.co.uk> wrote: > Hi Adrian, > > This I understand, but then to double the air rate would then lead to a net > reduction in sensitivity due to the wider bandwidth. >
Hi Sam, The gain obtained by using FEC will not only compensate but increase your sensitivity despite the wider noise bandwidth due to increased symbol rate, but only if you perform soft symbol decoding, typically with the Viterbi algorithm. Let me explain what I mean by using a binary FSK system. Hard decision decoding means you look at the levels and you take everything above 0 as a 1 and everything below 0 as a 0 and pass that to the decoder. This method discards a lot of information, especially when you have slight errors with symbol interference crossing the threshhold a little . With soft decision decoding, you pass to the FEC decoder a range of say 1...255, with 1 being a definite 0 and 255 a definite one. Then you let the Viterbi decoder choose the correct path through the trellis. This means that small errors, which cross the hard threshhold slightly, will be corrected using FEC information. A convolutional code with K=7, N=2 will provide ~3 dB gain in the 1e-3 to 1e-2 BER zone compared to uncoded channel. There are plots available on the internet for different FEC implementations. This was just an example for understanding FEC. There are other techniques which work even better, for example using an inner code with Viterbi decoding and an outer code like Reed-Solomon. Some encoders work best at high symbol rate and introduce high latency, others can deal with low rates and very little latency. Best regards, Adrian ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Freetel-codec2 mailing list Freetel-codec2@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/freetel-codec2
