Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
Eb/N0 is important in dealing with binary data. it means when u want to send data as bits to a transmitter or recieve from reciever. SNR is dealt in analog frmat. e.g. when u have (base band or passband) modulated ur (e.g) binary data and have sent it in the C-T cannel.
furthermore, almost in every (base band or passband) modulation like PCM, PSK, QPSK, ... u can have expressions to relate Pe(probability of error) to Eb/N0.
as a weak conclusion, Eb/No is the equivalence to the SNR in digital comm. since SNR is more general and is dealt in any comm. system.
hope this helps......SNR and BW are exchangeable ie relation is exponential....if transmission requirement is (S/N)1 and BW1,the same can be transmiited via
(S/N)2 and BW2..
(S/N)2 =(S/N)1 BW1/ BW2
small increase in BW produce large advantage in REDUCED POWER
large increase in power will have less advantage in reduced BW ....
example can be PCM
in FM, abovesaid relation will false.In FM ,large BW will used for large SNR..
as a summary,digital system close to above expon.relation involving SNR and BW
Hi every one.. What is the need to relate SNR and Eb/No.....I wud like to answer this question in a different fashion.
Power Signal: finite average power, infinite energy, good model for analog signal
Energy Signal: zero average power, finite energy
Power signals are good for analog signals since they can be thought of as existing for a long time
Digital symbols exist over one symbol or bit interval, Tb, so this allows comparison between different M-ary signals ... from this we can say..
SNR, average signal power to average noise power is important for measuring performance in analog systems In digital communication, the ratio is the bit energy (Eb) divided by noise spectrum density (η), a normalized version of SNR Allows comparison when M-ary systems are used.
Noise power spectral density η/2 = noise power N
divided by bandwidth 2B.
Eb/η = (S*Tb)/(N/B)!
Conceptually SNR ad Eb/No are same. SNR is used to represent signal to oise ratio for analog signals. Eb is the Energy of one 'bit' and No is the noise spectral density ( i.e. noise power/ BW). Thus Eb/No is the signal to noise ratio for a digital sigal for a defined BW.
We must prior understand our respectively "language": if You define Eb how the amount of energy to transmit a binary symbol and No how the monolateral power spectrum density, then Eb/No is the SNR.
In general, the error probability depends by SNR, so You must specify at least the Modulation Format and if Your signal is Codified or not (where with codified I refer to protection code how convolutional code).
Suppose You work at RF, the costellation is a 4-QAM, works with zero-ISI (the complete channel response satisfy the Nyquist condition) and explain by complex envelops. Then, the P(e) is equal to 2Q(1/No) + Q^2(1/No). The quantity Q^2() can be trascurate respect Q() if You work at high SNR (for example in correspondency to P(e) = 10^-5). Now, You desire to eplain by the SNR: remember that the energy Es=Eb because the signal is uncodified. With a 4-QAM, the energy at RF is Es=1. You can write then Es/No = 1/No, where Es/No is Your SNR. Finally You have: P(e) = 2Q(Es/No).
If You work with a convolutionl code with Rate R=k/n and Free Distance Df, (by some pages of pssages!!!) You can approximate the P(e) how a function of R, SNR and Df. Note that in this case, with SNR I consider the quantiy Eb/No where in this case Eb is the energy request to transimit an uncodified symbol.
How exerices for You (gurpreet) try to explain qualitatively the advantages by Soft Decoder vs Hard Decoder. Good Study and keep in touch!