System Properties Especially: Linear Time Invariant Systems

Slides:



Advertisements
Similar presentations
Signals and Systems – Chapter 2
Advertisements

Discrete-Time Linear Time-Invariant Systems Sections
Lecture 6: Linear Systems and Convolution
Signals and Systems Lecture #5
Discrete-time Systems Prof. Siripong Potisuk. Input-output Description A DT system transforms DT inputs into DT outputs.
Analysis of Discrete Linear Time Invariant Systems
Digital Signals and Systems
1 Signals & Systems Spring 2009 Week 3 Instructor: Mariam Shafqat UET Taxila.
Course Outline (Tentative)
Basic System Properties. Memory Invertibility Causality Stability Time Invariance Linearity.
Time Domain Representation of Linear Time Invariant (LTI).
Chapter 3 Convolution Representation
DISCRETE-TIME SIGNALS and SYSTEMS
Time-Domain Representations of LTI Systems
ECE 8443 – Pattern Recognition EE 3512 – Signals: Continuous and Discrete Objectives: Convolution Definition Graphical Convolution Examples Properties.
Discrete-time Systems Prof. Siripong Potisuk. Input-output Description A DT system transforms DT inputs into DT outputs.
Signal and Systems Prof. H. Sameti Chapter #2: 1) Representation of DT signals in terms of shifted unit samples System properties and examples 2) Convolution.
EEE 301 Signal Processing and Linear Systems Dr
BYST CPE200 - W2003: LTI System 79 CPE200 Signals and Systems Chapter 2: Linear Time-Invariant Systems.
Linear Time-Invariant Systems
COSC 3451: Signals and Systems Instructor: Dr. Amir Asif
Course Outline (Tentative) Fundamental Concepts of Signals and Systems Signals Systems Linear Time-Invariant (LTI) Systems Convolution integral and sum.
Chapter 2 Time Domain Analysis of CT System Basil Hamed
Hossein Sameti Department of Computer Engineering Sharif University of Technology.
ECE 8443 – Pattern Recognition EE 3512 – Signals: Continuous and Discrete Objectives: Causality Linearity Time Invariance Temporal Models Response to Periodic.
ENGG 330 Class 2 Concepts, Definitions, and Basic Properties.
EEE 503 Digital Signal Processing Lecture #2 : EEE 503 Digital Signal Processing Lecture #2 : Discrete-Time Signals & Systems Dr. Panuthat Boonpramuk Department.
Course Outline (Tentative) Fundamental Concepts of Signals and Systems Signals Systems Linear Time-Invariant (LTI) Systems Convolution integral and sum.
4. Introduction to Signal and Systems
Time Domain Representation of Linear Time Invariant (LTI).
Signals and Systems Analysis NET 351 Instructor: Dr. Amer El-Khairy د. عامر الخيري.
1 “Figures and images used in these lecture notes by permission, copyright 1997 by Alan V. Oppenheim and Alan S. Willsky” Signals and Systems Spring 2003.
Signals and Systems Lecture #6 EE3010_Lecture6Al-Dhaifallah_Term3321.
Description and Analysis of Systems Chapter 3. 03/06/06M. J. Roberts - All Rights Reserved2 Systems Systems have inputs and outputs Systems accept excitation.
Chapter 2. Signals and Linear Systems
Time Domain Representations of Linear Time-Invariant Systems
Time Domain Representation of Linear Time Invariant (LTI).
Chapter 2. Signals and Linear Systems
Properties of LTI Systems
What is System? Systems process input signals to produce output signals A system is combination of elements that manipulates one or more signals to accomplish.
CEN352 Dr. Nassim Ammour King Saud University
Discrete-time Systems
Linear Time Invariant Systems
EE 309 Signal and Linear System Analysis
Recap: Chapters 1-7: Signals and Systems
Chapter 1. -Signals and Systems
Description and Analysis of Systems
Signals and Systems Using MATLAB Luis F. Chaparro
Research Methods in Acoustics Lecture 9: Laplace Transform and z-Transform Jonas Braasch.
UNIT V Linear Time Invariant Discrete-Time Systems
Signal and Systems Chapter 2: LTI Systems
UNIT-I SIGNALS & SYSTEMS.
Lecture 5: Linear Systems and Convolution
Signals and Systems EE235 Leo Lam Leo Lam ©
CS3291: "Interrogation Surprise" on Section /10/04
Signals and Systems EE235 Leo Lam ©
2. Linear Time-Invariant Systems
Signals and Systems EE235 Leo Lam Leo Lam ©
Signals & Systems (CNET - 221) Chapter-2 Introduction to Systems
Signals & Systems (CNET - 221) Chapter-2
Signals & Systems (CNET - 221) Chapter-3 Linear Time Invariant System
Signals and Systems EE235 Leo Lam ©
LECTURE 05: CONVOLUTION OF DISCRETE-TIME SIGNALS
LECTURE 07: CONVOLUTION FOR CT SYSTEMS
Concept of frequency in Discrete Signals & Introduction to LTI Systems
Convolution sum.
Lecture 3: Signals & Systems Concepts
2.3 Properties of Linear Time-Invariant Systems
Lecture 3 Discrete time systems
SIGNALS & SYSTEMS (ENT 281)
Presentation transcript:

System Properties Especially: Linear Time Invariant Systems

Scaling Homogeneity: Scaling the input is the same as scaling the output signal. CT: ℋ 𝑘𝑥 𝑡 =𝑘ℋ 𝑥 𝑡 DT: ℋ 𝑘𝑥 𝑛 =𝑘ℋ 𝑥 𝑛

Adding Additivity: Adding input signals is the same as adding individual output signals. CT: ℋ 𝑖=1 𝑀 𝑥 𝑖 𝑡 = 𝑖=1 𝑀 ℋ 𝑥 𝑖 𝑡 DT: ℋ 𝑖=1 𝑀 𝑥 𝑖 𝑛 = 𝑖=1 𝑀 ℋ 𝑥 𝑖 𝑛

Linear: Scaling and Adding Linearity: Having both Homogeneity and Additivity. Only need to show for 𝑀= 2 since induction can then be used to show it holds for all positive 𝑀. CT: ℋ 𝑖=1 𝑀 𝑘 𝑖 𝑥 𝑖 𝑡 = 𝑖=1 𝑀 𝑘 𝑖 ℋ 𝑥 𝑖 𝑡 DT: ℋ 𝑖=1 𝑀 𝑘 𝑖 𝑥 𝑖 𝑛 = 𝑖=1 𝑀 𝑘 𝑖 ℋ 𝑥 𝑖 𝑛

Time Invariance Time shifting the input is the same as time shifting the output signal. CT: ℋ 𝑥 𝑡−𝜏 = ℋ 𝑥 𝜆 𝜆=𝑡−𝜏 DT: ℋ 𝑥 𝑛−𝑁 = ℋ 𝑥 𝑝 𝑝=𝑛−𝑁 On the left H is acting on x as a function of t. Since we are using a substitution of variables on the right, H is acting on x as a function of t. This distinction will become more obvious in the examples.

LTI Having both Linearity and Time Invariance. CT: ℋ 𝑖=1 𝑀 𝑘 𝑖 𝑥 𝑖 𝑡− 𝜏 𝑖 = 𝑖=1 𝑀 𝑘 𝑖 ℋ 𝑥 𝑖 𝜆 𝜆=𝑡− 𝜏 𝑖 DT: ℋ 𝑖=1 𝑀 𝑘 𝑖 𝑥 𝑖 𝑛− 𝑁 𝑖 = 𝑖=1 𝑀 𝑘 𝑖 ℋ 𝑥 𝑖 𝑝 𝑝=𝑛− 𝑁 𝑖 Fundament requirement for system analysis discussed in following chapters.

Other Properties Bounded-Input Bounded-Output Stability Causality If the input 𝑥 𝑛 ∞ <∞, then the output 𝐻 𝑛 ∞ <∞. ⋅ ∞ means the maximum absolute value Causality The output does not occur before the input. Said another way: the output at time 𝑡 does not depend on future input values (e.g. 𝑡+1). Invertibility There is a unique output for every unique input. This means that if you know the output, then you can figure out what the input is. E.g. 𝑦=2𝑥 is invertible, but 𝑦= 𝑥 2 is not. Static vs. Dynamic Static: the output depends on the input at the same time. Dynamic: the output depends on input values at other times.

Example: H[x(t)] = 3x(t) + 1 Linearity: ℋ 𝑘 1 𝑥 1 𝑡 +𝑘 2 𝑥 2 𝑡 = 𝑘 1 ℋ 𝑥 1 𝑡 + 𝑘 2 ℋ 𝑥 2 𝑡 3k1x1(t) + 1 + 3k2x2(t) + 1 = k1(3x1(t) + 1) + k2(3x2(t) + 1)? 3k1x1(t) + 1 + 3k2x2(t) + 1 = k13x1(t) + k1 + k23x2(t) + k2? NO does 2 = k1 + k2? This equality does not hold for all values of k1,2. Time Inv.: ℋ 𝑥 𝑡−𝜏 = ℋ 𝑥 𝜆 𝜆=𝑡−𝜏 3x(t – τ) + 1 = (3x(λ) + 1)|λ=t – τ? YES 3x(t – τ) + 1 = 3x(t – τ) + 1? This holds for all values of τ.

Example: y(t) = H[x(t)] = 3x(t) + 1 BIBO Yes, If the input is bounded the output is bounded to 3 times the input plus 1. Causality Yes, does not depend on future values. Invertible Yes, x(t) = (1/3)(y(t) – 1). Dynamic or Static Static, only depends on the input at the same time.

Example: H[x(t)] = 5x(t+2) Linearity: ℋ 𝑘 1 𝑥 1 𝑡 +𝑘 2 𝑥 2 𝑡 = 𝑘 1 ℋ 𝑥 1 𝑡 + 𝑘 2 ℋ 𝑥 2 𝑡 5k1x1(t+2) + 5k2x2(t+2) = k15x1(t+2) + k25x2(t+2)? YES This equality holds for all values of k1,2. Time Inv.: ℋ 𝑥 𝑡−𝜏 = ℋ 𝑥 𝜆 𝜆=𝑡−𝜏 5x(t+2–τ) = (5x(λ+2))|λ=t–τ? YES 5x(t+2–τ) = 5x(t–τ+2)? This holds for all value of τ.

Example: y(t) = H[x(t)] = 5x(t+2) BIBO Yes, If the input is bounded the output is bounded to 5 times the input. Causality No, it does depend on future values. Invertible Yes, x(t) = (1/5)y(t-2). Dynamic or Static Dynamic, depends on the input at different time(s).

Example: H[x[n]] = nx[n] Linearity: ℋ 𝑘 1 𝑥 1 𝑛 +𝑘 2 𝑥 2 𝑛 = 𝑘 1 ℋ 𝑥 1 𝑛 + 𝑘 2 ℋ 𝑥 2 𝑛 nk1x1[n] + nk2x2[n] = k1nx1[n] + k2nx2[n]? YES This equality holds for all values of k1,2. Time Inv.: ℋ 𝑥 𝑛−𝑁 = ℋ 𝑥 𝑝 𝑝=𝑛−𝑁 nx[n–N] = px[p]|p=n–N? NO nx[n–N] = (n-N)x[n–N]? This does not hold for all value of N.

Example: H[x[n]] = nx[n] BIBO No, the 𝐻 𝑥 𝑛 →∞ as 𝑛→∞ for bounded nonzero 𝑥 𝑛 Causality No, it does depend on future values. Invertible No, x[n] = (1/n)y[n] works except for n=0. At n=0, the output will be zero regardless of the input. Can’t determine x at n=0. Dynamic or Static Static, only depends on the input at the same time.

Example: H[x(t)] = sin(x(t)) Linearity: ℋ 𝑘 1 𝑥 1 𝑡 +𝑘 2 𝑥 2 𝑡 = 𝑘 1 ℋ 𝑥 1 𝑡 + 𝑘 2 ℋ 𝑥 2 𝑡 No why? Time Inv.: ℋ 𝑥 𝑡−𝜏 = ℋ 𝑥 𝜆 𝜆=𝑡−𝜏 Yes why?

Example: y(t) = H[x(t)] = sin(x(t)) BIBO Yes, why? Causality Yes, why? Invertible No, why? Dynamic or Static Static, why? How do any of these change with y(t) = tan(x(t))

Common LTI system operators Summation of two signals (additivity): as long as x1 and x2 are both inputs or they are produced through LTI combinations of the same signal.

Common LTI system operators Scaling (homogeneity) by a constant

Common LTI system operators Time Shift (Time Invariance): predominantly used in DT systems. Figure shows delays, but can also have lead or negative delay, 𝑧 𝑁

Common LTI system operators Differentiation: CT systems only.

Common LTI system operators Integration: CT systems only.

Linear Combinations of LTI operators are also LTI. Differences and Accumulation are two examples for DT LTI systems.

Why is LTI so important How can we use the properties of LTI to find y[n] = H[x[n]]? given we know h[n]= H[δ[n]] (impulse response). Any DT signal can be expressed as a sum (additivity) of time-shifted (time invariance) and scaled (homogeneity) unit impulses: 𝑥 𝑛 = 𝑖=−∞ ∞ 𝑘 𝑖 𝛿 𝑛− 𝑁 𝑖 = 𝑖=−∞ ∞ 𝑥 𝑁 𝑖 𝛿 𝑛− 𝑁 𝑖 The output is in a form similar to the LTI equation on the previous slide. 𝑦 𝑛 =ℋ 𝑥 𝑛 =ℋ 𝑖=−∞ ∞ 𝑥 𝑁 𝑖 𝛿 𝑛− 𝑁 𝑖 And using the property of LTI, the output can be expressed in terms of h[n]. 𝑦 𝑛 = 𝑖=−∞ ∞ 𝑥 𝑁 𝑖 ℋ 𝑥 𝑖 𝑝 𝑝=𝑛− 𝑁 𝑖 = 𝑖=−∞ ∞ 𝑥 𝑁 𝑖 ℎ 𝑛− 𝑁 𝑖 More on this in the next chapter … Chapter 5.