3 Quick questions about information theory

AI Thread Summary
The discussion addresses three questions related to information theory. For part A, uniquely decodable codes identified are 1, 3, 4, and 5, while instantaneous codes are 1, 4, and 5, with code 2 being non-decodable. In part B, the probabilities for the second-order extensions are calculated as P(aa) = p1*p1, P(ab) = p1*q1, and P(bb) = q1*q1. Part C clarifies that "rate" can refer to information rate (average bits per symbol) or transmission rate (bits per time). The discussion concludes with a request for confirmation on the correctness of these points.
degs2k4
Messages
72
Reaction score
0

Homework Statement



A) Select uniquely decodable codes and instantaneous codes from Code1 to 5 of the image below:
2ut2fxe.png


B) Personal question about second-order extension probabilites. If we have:
probability of a symbol a P(a) = p1
probability of a symbol b P(b) = q1
Which is the probability of symbols aa, ab, bb of the second-order extension of the source?

C) What is "rate"?

The Attempt at a Solution



A)

Uniquely decodable codes: 1,3,4,5 (in code 2 the extension(AE)=00100=extension(BA) )
Instantaneous codes: 1,4,5 (code 2 was not uniquely decodable, and in code 3 A is prefix of other codes)

B)

P(aa) = p1*p1
P(ab) = p1*q1
P(bb) = q1*q1

c)

It depends on what it refers to:
" Information rate R " : avg bits/symbol, R< Channel Capacity implies theoretical error-free transmission (Shannon)

" transmission rate " : bits/time(sec)

Is everything ok?

Thanks in advance.
 
Physics news on Phys.org
Noone ?
 
Back
Top