# RP-HW4-sol - Kyung Hee University Department of Electronics...

This preview shows pages 1–4. Sign up to view the full content.

1 Kyung Hee University Department of Electronics and Radio Engineering C1002900 Random Processing Homework 4 Solutions Spring 2010 Professor Hyundong Shin Issued: May 12, 2010 Due: May 26, 2010 (No acceptance of overdue submission) Reading: Course textbook Chapter 8, Lecture Notes VIII–IX HW 4.1 (a) Suppose that  ˆ X was unbiased. Then from the quality  ˆ X  , we have   / / // / / ˆ for all ˆ for all ˆˆ for all , ˆ for all , ˆ for all . xd x xdx xx      12 2 1 1 0 1 0 11 00 1 1 0 10 0 (4.1) The final equality implies that ˆ X cannot be unbiased, so we have reached a contradiction. (b) Yes, we can find a MVU as follows. Let ˆ Xg X  be any unbiased estimator. Then    ; . X gx f x dx gx dx gxdx  0 0 2 0 1 (4.2) This last equality holds for all  0 , so that we can differentiate it with respect to to ob- tain that   gx 2 , which must hold for any unbiased estimator. Therefore, we conclude that ˆ XX 2 is the only unbiased estimator, and so must be the MVU. HW 4.2 Let

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

View Full Document
2 ,, . N NN XW r r r      0 00 1 11 1  Xr W (4.3) Then, we wish to estimate A given A  W . By inspection,   , N A 2 I . This prob- lem falls exactly into the canonical scalar/vector. We can find the Cramer-Rao bound (CRB) as follows:  / ;e x p N N fA A     2 22 2 2 X x xr (4.4)   ln ; / ln /l n N A A NA A    2 2 2 1 2 1 2 2 X xx r x r x r (4.5)   ln ; A A  2 1 X x rx rr (4.6)   ln ; A  2 1 X x rr (4.7)   ln ; , i f i f . N n n N IA A r r r r N r       2 2 2 1 2 2 0 2 2 1 1 1 1 1 X X X (4.8) The CRB:  i f ˆ Var i f . N r r r A N r  2 2 1 1 1 1 X X (4.9) Now we can easily check for the existence of an efficient estimator:   eff ln ; ˆ . AA A 2 2 1 X X X X rX rr rX (4.10)
3 Since the right-hand side of (4.10) does not depend on the parameter A , this estimator is valid and must be the efficient estimator for the problem. Since it is efficient, its MSE must be given by the CRB:   eff , i f ˆ Var i f . N r r r A N r   2 22 2 11 1 1 1 X (4.11) The estimator is consistent if eff ˆ lim Var N A  0 X . We see that   eff i f ˆ lim Var i f . N rr A r 01 X (4.12) Hence, our efficient estimator is consistent for r 1 .

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

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

## This note was uploaded on 06/10/2010 for the course ELECTRONIC C1002900 taught by Professor Hyungdongshin during the Spring '10 term at Kyung Hee.

### Page1 / 14

RP-HW4-sol - Kyung Hee University Department of Electronics...

This preview shows document pages 1 - 4. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online