L1 - YORK UNIVERSITY CSE4210 CSE4210 Architecture and...

Info iconThis preview shows pages 1–6. Sign up to view the full content.

View Full Document Right Arrow Icon
1 YORK UNIVERSITY CSE4210 CSE4210 Architecture and Hardware for DSP Lecture 1 Introduction & Number systems YORK UNIVERSITY CSE4210 Administrative Stuff • CSE4210 Architecture and Hardware for DSP • Text: VLSI Digital Signal Processing Systems: Design and Implementation. K. Parhi. Wiley Interscience • Posted articles
Background image of page 1

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

View Full DocumentRight Arrow Icon
2 YORK UNIVERSITY CSE4210 Administrative Stuff • Office hours: Monday 1-2pm TR 3-4pm • Room 2026 CSEB x40607 •HW 0% • Quizes 10% • Midterm 25% • Projects 25% • Final 40% YORK UNIVERSITY CSE4210 Topics • Number systems • Fast arithmetic • Algorithm representation • Transformation (retiming, unfolding, folding) • Systolic arrays and mapping algorithms into hardware • Low power design
Background image of page 2
3 YORK UNIVERSITY CSE4210 Introduction • Introduction to DSP algorithms • Non-terminating programs in real time. • Speed depends on applications (audio, video, 2-D, 3-D, …) • Need to design families of architectures for specified algorithm complexity and speed constraints YORK UNIVERSITY CSE4210 Typical DSP Programs DSP System nT …3T 2T T 0 Input Output 3-Dimensional optimization: Area, Speed, Power) Usually, speed is a requirement, area-power tradeoff P=C V 2 F
Background image of page 3

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

View Full DocumentRight Arrow Icon
4 YORK UNIVERSITY CSE4210 Examples • FIR filter, x(n) is the input, y(n) output • IIR filter 1 0 () ()( ) J j y nh j x n j = = 10 ) ) Q P ik yn aiyn i bkxn k == =− +− ∑∑ YORK UNIVERSITY CSE4210 Examples • Convolution 11 00 ) ) MN ij xihn i h jxn j −− = −= For n=1 to M+N-2, y(n)=0, For i=0:M-1, y(n)=y(n)+x(i)*h(n-i) end end MAC operation
Background image of page 4
5 YORK UNIVERSITY CSE4210 More Complex Examples Motion Estimation • Image (frame) is divided into macroblocks • Each macroblock is compared to a macroblock in the reference frame using some error measure. • The search is conducted over a predetermined search area. • A vector denoting the displacement of the (motion) is sent. YORK UNIVERSITY CSE4210 More Complex Examples Motion Estimation • Many measures of errors could be used. • The displaced block difference s(m,n) using MAD (Mean Absolute Difference) is defined as ∑∑ + + = = = 1 0 1 0 ) , ( ) , ( ) , ( N i N j n j m i y j i x n m s m,n are in is the search area, N is the macroblock size.
Background image of page 5

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

View Full DocumentRight Arrow Icon
Image of page 6
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 22

L1 - YORK UNIVERSITY CSE4210 CSE4210 Architecture and...

This preview shows document pages 1 - 6. Sign up to view the full document.

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