so I came across M-ary Modulation technique in my ADC subject. While I understood the basic principle behind it, I still cant get my head around the amplitude levels taken . Like lets say i am using 16QAM, here i get that each signal point in the constellation corresponds to a specific quadbit, which is made up of 4 bits, and we have 16 different symbols possible here, but how do we set the amplitude levels here. All those constellation diagrams present in books and on internet, how are the phase and amplitude mapping being done. Like i know i am missing some basics here, Anyone could clear this doubt please.
1
-
\$\begingroup\$ See this info.support.huawei.com/info-finder/encyclopedia/en/QAM.html \$\endgroup\$– Antonio51Commented Apr 15 at 15:25
Add a comment
|
2 Answers
\$\begingroup\$
\$\endgroup\$
2
The constellation diagram shows the phase and amplitude mapping directly: If you draw a line from a point in the constellation to the origin (0/0), the amplitude is equal to the length of that line, and the phase is the angle of the line relative to the real axis.
Alternatively, in more physical terms: The real part of a point's coordinate determines the amplitude of the in-phase carrier signal, while the imaginary part of the coordinate determines the amplitude of the quadrature (90° phase-shifted) carrier.
When you create such a constellation diagram, the goal is to maximize the euclidian distance between the individual symbol points. This distance is quite literally a margin for error - when noise disturbs the transmission, it "pushes" the symbol point around in the constellation. The larger the distances between the individual symbols, the more noise power is needed to actually cause a reception error as the noise needs to push the signal farther.
A rectangular grid (as in QAM) achieves equal distance between all adjacent symbols, which means that all transmitted bits are protected against errors equally. By varying the distances, you can protect some bits more than others.
Additionally, you need to choose a way to map bit patterns onto the symbol points. The usual approach is to choose a gray-code, which assigns patterns to adjacent symbols that only differ in a single bit. That way, even when noise pushes a received symbol so far from its correct position that it gets mis-detected as a different symbol, chances are that only a single bit is affected by this error. In general, though, the mapping of bit patterns to symbols can be arbitrary. Some mappings are just more advantageous than others.
-
\$\begingroup\$ thanks for the explanation. I had one small doubt though. Suppose we are talking about 8QAM, how do i know the amplitude levels here, like can it be 2 or even 4 ? how do we determine that? by drawing constellation diagram? like how do i proceed with it? I know it may sound novice but any info u can give? \$\endgroup\$– blitzzzCommented Apr 16 at 4:31
-
\$\begingroup\$ @blitzzz Yes, you do need to know how the symbols are arranged. \$\endgroup\$ Commented Apr 16 at 15:47
\$\begingroup\$
\$\endgroup\$
QAM signals must be normalized to a known amplitude before the slicing to identify discrete symbols can occur. In order to do this, the receiver must be able to identify the peak values of the signal and adjust its gain accordingly.
QAM communication systems usually include a pseudorandom component (a "scrambler") that is used to reduce the probability that a pathological data pattern would cause the automatic gain control to fail.
