AboutLuis Emiliani Expertise Strengths
- Satellite link design and analysis
- Vsat network design and analysis
- Propagation aspects for terrestrial and satellite links
- ITU-R P recommendations
Can not help with:
- Questions related to TV receivers, TV kits, or specific networks such as DirecTV or DISH
- Antenna pattern synthesis, RF MoM analysis, S-parameters.
Experience My experience is in satellite communication networks and terrestrial microwave links.
Expert: Luis Emiliani Date: 7/14/2008 Subject: Regarding Eb/No and FEC coding
Question QUESTION: Dear Luis,
i want to know that, what is the relationship between EbNo and BER and what is the effect of changing Modulation and FEC on these two parameters.
ANSWER: Hi
thanks for asking
the relationship between BER and Eb/no depends on the type of modulation selected. Actually, what you get from the analysis of the constellation diagram of each modulation is the probability of decoding a symbol in error, which is a function of the energy of the symbol. from there you can have the probability of errors on the bit stream, related to the Eb/No.
in wikipedia you will find a description of the expressions used for various modulation types
it all depends on the distance between symbols in the constellation diagram. if you increase the number of symbols (from QPSK to 8PSK you change from 4 to 8 symbols), they are closer to each other and the probability of error is higher for the same Eb/No
so you need more power to make sure the errors in the channel do not affect the decoding of the symbols.
if you look at 256-TCM things are really really bad!
one way of improving performance is to protect the signals by adding redundancy. this is called coding, and you can build codes that detect errors, and codes that detect and correct errors. parity check is a form of error detecting, and block codes are a means of detecting and correcting errors. the amount of errors that can be corrected depends on how you create the code and on the number of bits that differ from coded word to coded word (the distance of the code)
since the signal is protected, you can transmit with less power and still decode everything correctly, at the expense of an increase in bandwidth.
this improvement is called the coding gain.
so the effect of changing codes results in a reduction in the Eb/No required for a given BER threshold, compared to the uncoded case.
if you change modulation and reduce the number of symbols, you will get a similar effect. if you do both at the same time (say, from QPSK 3/4 coded to BPSK 1/2 coded) the improvement is the sum of both (one is the change in required Eb/No because of the increase in symbol distance, and the other is the change due to coding gain).
the performance of codes is evaluated via simulations so normally you do not find a closed expression for the probability of error of a coded modulation. normally you see graphs displaying the improvement.
i hope this helps
best regards
luis
---------- FOLLOW-UP ----------
QUESTION: Dear Luis,
i want to know that how the FEC coding technique detect and correct errors in the channel.Can you please explain it in the form of packet level in which how the codes are embbed to the system are mentioned.
Answer Hi,
there are some tutorials on the web that can help you understand what it is all about.
