Communication Systems

EE-351
Fourier Series for a periodic signal
• Example: Consider a periodic stream of pulse with width d= T/4 with
amplitude A, and time period T. ∞
𝑥 𝑡 = 𝑐𝑘 𝑒 𝑗2𝜋𝑘𝐹𝑜 𝑡
𝑘=−∞
Also,
1 𝑇/2
𝑐𝑙 = 𝑥(𝑡)𝑒 −𝑗2𝜋𝑙𝐹𝑜 𝑡 𝑑𝑡
𝑇 −𝑇/2
Therefore, for the above particular signal,
1 𝑇/4 −𝑗2𝜋𝑙𝐹 𝑡
𝑐𝑙 = 𝐴𝑒 𝑜 𝑑𝑡
𝑇 0
Fourier Series for a periodic signal
𝐴 𝑇/4 𝐴𝑇 𝐴
If 𝑙 = 0, 𝑐𝑙 = 1. 𝑑𝑡 = =
𝑇 0 𝑇4 4
• This Fourier coefficient corresponding to 𝑙 = 0, also known as DC
coefficient corresponds to the 0 freq. Rest of them corresponds to 𝑙 ≠
0 that are AC coefficients.
1 𝑇/4 𝐴 𝑒 −𝑗2𝜋𝑙𝐹𝑜 𝑇/4 −1 𝐴
If 𝑙 ≠ 0, 𝑐𝑙 = 0 𝐴𝑒 −𝑗2𝜋𝑙𝐹𝑜 𝑡 𝑑𝑡 = = 1 − 𝑒 −𝑗𝜋𝑙/2
𝑇 𝑇 −𝑗2𝜋𝑙𝐹𝑜 𝑗2𝜋𝑙
Taking 𝑒 −𝑗𝜋𝑙/4 common,
𝑗𝜋𝑙
𝐴 𝐴 − 4 𝜋𝑙 𝐴 −𝑗𝜋𝑙 𝜋𝑙
= 𝑒 −𝑗𝜋𝑙/4 𝑒 𝑗𝜋𝑙/4 − 𝑒 −𝑗𝜋𝑙/4 = 𝑒 (2𝑗𝑠𝑖𝑛 )= 𝑒 4 𝑠𝑖𝑛
𝑗2𝜋𝑙 𝑗2𝜋𝑙 4 𝜋𝑙 4
Fourier Series for a periodic signal
𝐴 𝜋𝑙 −𝑗𝜋𝑙
𝑐𝑙 = sin( )𝑒 4
𝜋𝑙 4
Hence,
𝐴
, 𝑙=0
4
𝑐𝑙 = 𝐴 𝜋𝑙 −𝑗𝜋𝑙
sin 𝑒 4 ,𝑙 ≠ 0
𝜋𝑙 4
For magnitude only,
𝑗𝜋𝑙
𝐴 𝜋𝑙 − 4 𝐴 𝜋𝑙
𝑐𝑙 = sin 𝑒 = sin magnitude spectrum
𝜋𝑙 4 𝜋𝑙 4
Power of a periodic signal
For a periodic signal x(t), the power is defined as:
1 𝑇/2
𝑃= 𝑥(𝑡) 2 𝑑𝑡
𝑇 −𝑇/2
∞

𝑥 𝑡 = 𝑐𝑘 𝑒 𝑗2𝜋𝑘𝐹𝑜 𝑡
∞ 𝑘=−∞ ∗ ∞

𝑥∗ 𝑡 = 𝑐𝑘 𝑒 𝑗2𝜋𝑚𝐹𝑜 𝑡 = 𝑐𝑚 ∗ 𝑒 −𝑗2𝜋𝑘𝑚𝑡
𝑚=−∞ 𝑚=−∞
1 𝑇/2 1 𝑇/2 ∞
P= 𝑥 𝑡 . 𝑥 ∗ (𝑡) 𝑑𝑡 = −𝑇/2 𝑘=−∞ 𝑐𝑘 𝑒 𝑗2𝜋𝑘𝐹𝑜 𝑡 ∞ 𝑐
𝑚=−∞ 𝑚
∗ 𝑒 −𝑗2𝜋𝑘𝑚𝑡 𝑑𝑡
𝑇 −𝑇/2 𝑇
Power of a periodic signal
𝑇/2 ∞ ∞
1
= 𝑐𝑘 𝑐𝑚 ∗ 𝑒 𝑗2𝜋(𝑘−𝑚)𝐹𝑜 𝑡 𝑑𝑡
𝑇
∞ ∞ −𝑇/2 𝑘=−∞ 𝑚=−∞ ∞ ∞
𝑇/2
∗
1
= 𝑐𝑘 𝑐𝑚 𝑒 𝑗2𝜋(𝑘−𝑚)𝐹𝑜 𝑡 𝑑𝑡 = 𝑐𝑘 𝑐𝑚 ∗ 𝛿(𝑘 − 𝑚)
𝑇 −𝑇/2
𝑘=−∞ 𝑚=−∞ 𝑘=−∞ 𝑚=−∞
The only term in these double summation that will survive corresponding to
𝑘=𝑚
Therefore, 𝑐𝑘 𝑐𝑚 ∗ will be 𝑐𝑘 𝑐𝑘 ∗ which is 𝑐𝑘 2
𝑃= ∞ 𝑐
𝑘=−∞ 𝑘 𝑘 𝑐 ∗ = ∞ 𝑐
𝑘=−∞ 𝑘
2
Power of a periodic signal
𝑇/2 ∞
1
𝑥(𝑡) 2 𝑑𝑡 = 𝑐𝑘 2
𝑇 −𝑇/2 𝑘=−∞
1 𝑇/2
 −𝑇/2 𝑥(𝑡) 2 𝑑𝑡 power in time domain
𝑇
 ∞ 𝑐
𝑘=−∞ 𝑘
2 power in frequency domain, also known as the
Parseval’s theorem