Not Quite....
Quote:
Originally Posted by Tom Bombadil View Post
I'm no communications expert on 8PSK vs QPSK, but in some brief readings on them, what I've seen is that 8PSK has demonstratedly better performance under poor weather conditions but only a very minor advantage in bit rate efficiency.
If this is true, then E* might have a very small advantage if they were both running 4 HD channels per transponder, but at 6 vs 4, D* is going to have a BIG bit rate advantage.
As a HD PQ fanatic, if I ever switch to D*, I will be hoping that they never switchover to 8PSK as being limited to 4 channels per transponder is great for quality. Unquote
Quote:
This surprises me. QPSK requires less power to achieve the same bit error rate as 8PSK. I would expect QPSK to work better in poor weather conditions because of this. Unquote
The amount of bandwidth a signal uses on a transponder is related to the symbol rate (modulation rate). Transponder bandwidth is the amount frequency spectrum used. For a given bit rate, 8PSK (8 modulation states) reduces the BW by ~67% compared to QPSK's reduction of 50% as each modulation state (symbol) carries 3 bits of information to QPSK's 2 bits of information. (The reference is to 2-state BPSK). To reduce transponder BW utilization by 75%, one could use 16-QAM where each symbol carries 4 bits of information. The downside of increasing the modulation states is that for a given bit error rate, the amount of power in the carrier most increase and transponder power is limited. This explanation is an over simplification because forward error correction is another factor to consider when calculating a link budget. Here's a satellite modem performance table example:
BER ? 10-8
QPSK 3/4 Turbo 3.9 dB 2048 kbps 1365.333 ksps
8-PSK 3/4 Turbo 6.3 dB 2048 kbps 910.222 ksps
16-QAM 3/4 Turbo 7.7 dB 2048 kbps 682.667 ksps
The Bit Error Rate (BER) is the performance objective. The 3/4 Turbo refers to the Forward Error Correction. 3/4 means for every 3 payload bits, 4 bits are transmitted or 1 FEC bit is added on average. The 2048 is the payload or user bit rate and the last figure is the symbol rate. The figures in dB are EB/NO which is related to carrier-to-noise ratio. Essentially the higher figures require more power other factors being equal. The last column is symbol rate which is directly related to transponder bandwidth utilization. I don't know what BER performance objectives E* uses. As you can see, weather has nothing to do with the preferred modulation scheme. BTW, a BER of 10-8 means 1 X 10 exponent -8 which means 1 error is allowed in 100 million received bits.
