I have given this some thought but I can't really seem to make sense of the best way to benchmark the receiver side of transceivers in my link budgets. The basic problem comes from the receiver sensitivity only being given at certain BER levels and bit rates. I would like to compare apples to apples and do a theoretical comparison between different vendors by just using their data sheets.
I would like to use the receiver datasheet to calculate the system sensitivity at any bit rate, BER, deviation (where applicable) and bandwidth.
Assuming FSK, I have played around with using the minimum theoretical Eb/No and estimating the receiver NF as the difference between the Eb/No at sensitivity limit and the theoretical value. The problem here is that I get no real correlation to the datasheet when changing for example bit rate, I could even end up with negative Eb/No when going to extremes.
My assumption is (and we all know what they about those) that either I have a problem with the bandwidth I use in the calculations or the receiver NF can not be assumed to be constant for different bit rates.
Are there any receiver gurus out there that can help shed some light on my problem? Is there for instance a reasonable accurate way to estimate the receiver NF from one or several sensitivity measurements? Are there any literature that anyone could recommend where I can find some better answers?
I would like to use the receiver datasheet to calculate the system sensitivity at any bit rate, BER, deviation (where applicable) and bandwidth.
Assuming FSK, I have played around with using the minimum theoretical Eb/No and estimating the receiver NF as the difference between the Eb/No at sensitivity limit and the theoretical value. The problem here is that I get no real correlation to the datasheet when changing for example bit rate, I could even end up with negative Eb/No when going to extremes.
My assumption is (and we all know what they about those) that either I have a problem with the bandwidth I use in the calculations or the receiver NF can not be assumed to be constant for different bit rates.
Are there any receiver gurus out there that can help shed some light on my problem? Is there for instance a reasonable accurate way to estimate the receiver NF from one or several sensitivity measurements? Are there any literature that anyone could recommend where I can find some better answers?
