lecture_13-15 - 58 Lecture 13-15 Subspaces: We have...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 58 Lecture 13-15 Subspaces: We have introduced the notion of angle, length and distance into the space of signals. The next concept we look at is related to the notion of a plane or line in the space of signals. Note that in 3-dimensions to specify a plane (passing through the origin) we need 2 coefficients, i.e., equation of a plane in 3D is: c 1 x 1 + c 2 x 2 + x 3 = 0 To specify a line in 3-dimensions we need 4 coefficients (intersection of 2 planes): c 1 x 1 + c 2 x 2 + x 3 = 0 c 3 x 1 + c 4 x 2 + x 3 = 0 This is called algebraic characterization. It turns out that it is difficult to extend this idea to the general Hilbert space. So we look at a geometric characterization. The key observation we want to make here is that: 1. If vectors vectorx and vector y belong to a plane or a line, then the vector vectorx + vectory also belongs to that plane or that line, respectively. 2. If vector vectorx belongs to a plane or a line, then αvectorx belongs to the same plane or line, respectively. Using these ideas we can define a subspace of a Hilbert space. Definition: A set of vectors V is said to be a subspace of a Hilbert space S if the following conditions are satisfied. 1. If vectorx ∈ V , vectory ∈ V then vectorx + vectory ∈ V . 2. If vectorx ∈ V then αvectorx ∈ V for all α . Examples: 1. A line in 2D euclidean space is a subspace. 2. Let S be the set of all CT signals. Let V represents a set of CT signals that are frequency limited to the range [ − 1 / 2 , 1 / 2] Hz. V is a subspace of S . 3. Let S → Set of all DT signals. Let V be a set of all DT signals that are frequency limited to {− π/ 2 ,π/ 2 } . V is a subspace of S . Before we go further, let us look at two concrete applications where different inner products come into play. 59 Digital Communication: Consider a 2-input-2-output communication channel modeled as: Y = X + N where X is the input vector, N is the noise vector and Y is the output vector. All these vectors belong to R 2 . Let N be modeled as a zero mean jointly Gaussian random vector with probability density function: f N ( n ) = 1 2 πK 1 2 exp bracketleftbig − 1 2 n T K − 1 n bracketrightbig where K is the covariance matrix of N K = bracketleftbigg E [( N 1 − E [ N 1 ]) 2 ] E [( N 1 − E [ N 1 ])( N 2 − E [ N 2 ])] E [( N 1 − E [ N 1 ])( N 2 − E [ N 2 ])] E [( N 2 − E [ N 2 ]) 2 ] bracketrightbigg where N = bracketleftbigg N 1 N 2 bracketrightbigg . We want to transmit a binary digit of information over the channel. If is to be transmitted, send X = s and if 1 is to be transmitted, send X = s 1 where s and s 1 are 2 vectors which are deterministic quantities. The decoder is given by a partition of the output space R 2 into 2 regions D and D 1 . If Y ∈ D then is decoded and if Y ∈ D 1 then 1 is decoded....
View Full Document

This note was uploaded on 01/10/2012 for the course EECS eecs551 taught by Professor J during the Spring '11 term at University of Michigan.

Page1 / 9

lecture_13-15 - 58 Lecture 13-15 Subspaces: We have...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online