what's a symbol?

10 posts by 5 authors in: Forums > CWNA - Enterprise Wi-Fi Admin
Last Post: March 5, 2007:
  • I see references to symbols in the cwna book, but can't get a clear idea of what it is. I used the search function on this site as well as did a google, but haven't found an explanation yet. I vaguely remember something about different symbol rates or types allowing fast ethernet to get the speeds it was able to get over vanilla ethernet (10Mbps).


  • Let's see if I can answer my own question. A symbol period is the reference time that a device senses the signal and determines whether that signal is a 0 or 1. Can I assume that a symbol is therefore this sensed bit, and is a symbol rate the same as saying bit rate?


  • By (Deleted User)

    The symbol rate or baud rate is basically the rate at which the state of the circuit changes. If a circuit can carry two bits per second, then it's symbol rate is two. In other words, if the circuit can only carry two bits (0 and 1) such as in BPSK (which only uses two phases to represent 0 and 1), then its symbol rate is equal to its bit rate (1 bit per symbol). Different modulation methods carry multiple bits per symbol. In modulation methods such as QPSK for example where the circuit can carry four different bit patterns (00,01,10,11) using 4 differnet phase shifts, the encoding rate doubles or the bit rate becomes twice the symbol rate (2 bits per symbol). In more complex modulation methods, the symbol rate gets higher and bit rate even higher. So to answer your question, symbol rate is not always the same as bit rate.

  • By (Deleted User)

    Welcome Bazzouqa,

    Very nice 1st post of many I am sure to come. Nyquist and Shannon would be proud to see this. ;-)

  • Bazz,

    Thanks for the reply. If I understand you correctly, a symbol is sort of like a sliding size byte in that the underlying encoding technique can change the equivalent byte size from 1 to N bits, which is un-like a fixed byte size of 8 bits for computers.

    Is that correct, or am I off track?

  • I think that most of the replies on this thread are basically correct. I'll add my answer in case my particular way of wording it makes things clearer to some readers.

    Information transmission involves changing some quality of a medium. When we transmit information via sound, we change the pressure (that's the quality) of the air (that's the medium). When we transmit information via electricity, we change the amplitude, phase, or frequency (that's the quality) of the electrical current in the copper wire (that's the medium).

    In a digital transmission system, it is implicit that time is divided up in to discrete "slices", or sampling times. The receiver of the digital signal measures the signal for a single sampling time and then makes a determination of what value that signal should have. The length of a single "slice" of time determines the baud rate of the system. The baud rate is the number of samples per second that the system uses. The shorter the sample, the higher the baud rate.

    In a digital transmission system, there are a pre-defined number of possible values that the signal can be interpreted to have. In a binary system, the signal can only have two values: 0 and 1. In a more complex system, the signal can be interpreted as having four different values: 00, 01, 10, and 11. As the complexity of the system increases, the number of possible values that the signal can be interpreted to have increases.

    Combine the two concepts described above, and you have an important ratio: bits per baud. The bits per baud is determined by the complexity of the encoding method. Higher data rates can be achieved by increasing the baud rate, by increasing the bits per baud, or some combination of the two. If you increase the baud rate, you increase the frequency of the signal. This places a practical upper limit on the baud rate of the system, since higher-frequency systems present certain engineering challenges. If you increase the bits per baud, you increase the susceptibility of the system to signal corruption. In other words, for higher bits-per-baud, you need higher signal-to-noise ratio to avoid corruption. This also eventually approaches an engineering upper limit, although engineers seem to have made more progress at pushing this limit than the frequency-based one.

    When the CWNA study guide refers to "symbols," it is referring to "bauds" from the example above. In 802.11b, the symbol rate is 11 million per second. As the data rate changes (1, 2, 5.5, 11 Mega bits per second), the bits-per-baud rate is changing based on different encoding and modulation methods. We don't say "bauds" when talking about 802.11 because that term is more associated with modems and because 802.11's designers say "symbols".

    Also, 802.11 has a complex encoding and modulation system in which bits are passed through a spread spectrum encoding and then modulated via BPSK, QPSK, etc... Because there are so many steps in the data transmission path, we need different names for the data at each stage of the path. So, before the data is encoded or modulated, we call it bits. After the data goes through spread-spectrum modulation, we call it chips. After the data goes through BPSK, QPSK, etc... encoding, we call it symbols.

  • By (Deleted User)

    In case the concept of symbol is still not clear, consider this: When a carrier is modulated whether by changing its frequency, phase, amplitude, or a combination of phase and amplitude as in QAM for example, each of these state changes is assigned a unique pattern of binary bits (the number of bits in the pattern depends on the modulation scheme). The pattern of bits being represented essentially forms the symbol. You can think of a symbol as an analog sample in time that conveys the pattern of bits through the pattern of phase, frequency, or amplitude shifts. Technically, I wouldn't call it a "byte" or bits for that matter, although many do, because that can be misleading. You don't have actual 1's and 0's transmitted on an analog carrier, that's only in the digital world. State changes on an analog carrier represent the bits. I think that?¡é?€??s where the term symbol comes from since it?¡é?€??s a representation of the bits being transmitted. Hope this helps.

  • Hi,

    In a simple way can we state that Symbols are the chip bits when they are encoded ?

    Also when they are encoded they use either barker code or CCK, and the type of modulation is either BPSK,DBPSK, QBPSK, DQBPSK, QAM and the modulation method depends upon how many bits they will modulate.

    Is this statement correct ?


  • First of all, I want to thank everyone that has responded. Each of you has added a piece to the puzzle and it helps. The varying ways that people have tried to answer the question, quantifies to me just how hard of a concept this is to understand.


  • batjedi Escribi?3:

    In a simple way can we state that Symbols are the chip bits when they are encoded ?

    Also when they are encoded they use either barker code or CCK, and the type of modulation is either BPSK,DBPSK, QBPSK, DQBPSK, QAM and the modulation method depends upon how many bits they will modulate.

    I think you're on the right track. I would agree with your first statement. The data packet is represented by binary pieces of information called bits. After the data packet is encoded by spread spectrum, the binary pieces of information are called chips. After the resulting sequence of chips is encoded onto an analog waveform, the analog representations of the binary chips are called symbols.

    I think that I also agree with your second paragraph, although I refer to spread-spectrum (that's Barker Coding or CCK) as modulation and PSK, QAM, etc... as encoding, whereas you seem to reverse the terms modulation and encoding. I guess that's okay, because in my experience even the RF engineers who design these systems don't seem to use those two terms consistently, so who can blame us poor regular folks!

Page 1 of 1
  • 1