Professional Documents
Culture Documents
Week 4
Week 4
Digital Communications
Week # 4
1
Contents
3
Random Variables
CDF & PDF
4
Random Variables
Statistical Averages
5
Random Variables
Example of Tossing 2 Fair Coins
6
Random Variables
Example of Tossing 2 Fair Coins
7
Random Variables
Important PDFs
8
Random Variables
Gaussian or Normal PDF
9
Random Variables
Q(.) Table
10
Binary Communication Channel
Example: Threshold Detection
• Over a binary channel, messages m= 0 and 1 are transmitted
with equal probability using a positive and a negative pulse,
respectively.
• The received pulse corresponding to 1 is p(t), and the received
pulse corresponding to 0 is −p(t).
• Let the peak amplitude of p(t) be Ap at 𝑡 = 𝑇𝑝 .
• The pulses are transmitted in the presence of Noise.
11
Transmitted pulse;
Threshold Detection of Binary
Communication Channel
Example
• Because of the channel noise n(t), the received pulses will be ±p(t)+n(t)
• To detect the pulses at the receiver, each pulse is sampled at its peak amplitude.
• Because of the channel noise, the sampler output is ±Ap + n, where n, the noise
amplitude at the sampling instant, is an RV. 12
Threshold Detection of Binary
Communication Channel
Example
• Due to the symmetry of the situation, the optimum detection threshold is zero.
• Because noise amplitudes range from −∞ to ∞, the sample value −Ap + n can
occasionally be positive, causing the received 0 to be read as 1, and vice versa.
• If 0 is transmitted, it will be detected as 1 if −Ap +n > 0, that is, if n > Ap 13
Threshold Detection of Binary
Communication Channel
Example
If 𝑃(𝜖|𝟎) is the error probability given that 0 is transmitted, then
𝑃(𝜖|𝟎) = P(n > Ap)
Here, the noise PDF is zero mean, hence,
∞ 𝑛2
1
𝑃 𝜖𝟎 = න 𝑒 2𝜎2 𝑑𝑛
2𝜋𝜎 𝐴𝑝
𝑛 1
𝑙𝑒𝑡 𝜇 = ; 𝑑𝜇 = 𝑑𝑛
𝜎 𝜎
1 ∞ 𝜇2 Total error probability
𝑃 𝜖𝟎 = න𝐴 𝑒 2 𝑑𝜇
2𝜋𝜎 𝜎𝑝
𝐴𝑝
=𝑄
𝜎𝑛
Also, 𝑃(𝜖|𝟏) is the error probability given that 1 is transmitted
𝐴𝑝
𝑃 𝜖 𝟎 =𝑄 14
𝜎𝑛
Two-Dimensional Random Variables
CDF and PDF
15
Two-Dimensional Random Variables
Dependencies
16
Two-Dimensional Random Variables
Bayes’ Theorem
17
Two-Dimensional Random Variables
Correlated Gaussian RVs
18
Two-Dimensional Random Variables
Correlated Gaussian RVs
19
Multi-Dimensional Random Variables
Dependencies
20
N-Dimensional Random Variables
Jointly Gaussian RVs
21