This action might not be possible to undo. Are you sure you want to continue?

**Signal Representation and Baseband Processing
**

Version 2 ECE IIT, Kharagpur

Lesson 16 Representation of Signals Version 2 ECE IIT. Kharagpur .

• Consider a larger group of symbols in a sequence and design signals spread over long time duration [sequence based modulation .. for practical considerations. Usually. The duration of a symbol now is twice the duration of one information bit. this is not a difficult task. 1 ≤ i ≤ M can be expressed as linear combinations of N orthonormal basis functions. where N ≤ M. • When several bits make one symbol. Let us consider a systematic approach to identify M symbols from the input information sequence. …. If the rate of incoming information is Rb bits/sec.1 Version 2 ECE IIT. amplified and transmitted. If s1(t). you will learn about Representation of signals following the Gram-Schmidt orthogonalization procedure. N ⎧0 ≤ t ≤ T si (t ) = ∑ sij Φ j (t ) . The fundamental detection problem in a receiver. if we choose M = 4 for the same binary information sequence.After reading this lesson. s2(t). if the information sequence is binary and if we choose M = 2. As mentioned earlier in Module #1. ⎨ j =1 ⎩i = 1. {si(t)}. Signal space and signal constellation. sM(t) are real valued energy signals. the symbol rate is Rb/2 symbols per second. a very important issue in-between is to represent information symbols in suitable energy signals so that the signals can be modulated.. Use of signal space for signal detection. Kharagpur . any set of M energy signals. For a continuous stream of input of information sequence what kind of strategy should we take to represent them as signals? One may think of multiple alternatives including the following: • Consider one symbol at a time and design a signal for the symbol..16. we can identify ‘1’ as one symbol and ‘0’ as the other. we may consider a group of two bits at a time to define one symbol..demodulation strategy].. a digital modulator is supposed to accept stream of information-bearing symbols (usually bits) and represent them appropriately with or without the help of a carrier. So. Else. If the format of input information is known... where ‘m’ is a positive integer. For example. M is so chosen that M = 2m. This problem is addressed in general by a scheme known as Gram-Schmidt Orthogonalization Gram-Schmidt Orthogonalization The principle of Gram-Schmidt Orthogonalization (GSO) states that. 2. M ≥ N 4. each of duration ‘T’ sec. consider one bit at a time and design for the symbol. The next issue is to design ‘M’ energy signals for these M symbols such that the energy of each signal is limited within the symbol duration.

16... ⎨ ⎩ j = 1.. we know the corresponding scalar coefficients (Fig. T ∫ ϕi (t ). if i ≠ j 4... alternatively. if we know the signals and the basis functions. if the basis functions are known and the scalars are given.e... by following Fig..3 Note that each basis function has unit energy over the symbol duration ‘T’. i...1 Pictorial depiction of Equation 4.16. 2.where.2). T ⎧i = 1.16.16. if i = j ⎩0. Now..1.s which are orthonormal to each other... si 1 × ϕ1 (t) si 2 × ϕ2 (t) + + ∑ + si (t) 0≤t≤ T siN × ϕN (t) Fig. We will consider realvalued basis functions ϕj (t) .. Kharagpur . 2.. 4. Or. we can generate the energy signals. M sij = ∫ si (t )ϕ j (t )dt .ϕ j (t )dt = ⎨ 0 ⎧1. 4.16.2 The ϕj(t)-s are the basis functions and ‘sij’-s are scalar coefficients.1 Version 2 ECE IIT. N 0 4. 4.16.

.16.2 Pictorial depiction of Equation 4.Correlator Structure T × s2(t) s1(t) |←T→|←T→| ϕ1 (t) ∫ ( )dt 0 si1 T si ri (t) × ϕ2 (t) ∫ ( )dt 0 si2 T × ϕN (t) ∫ ( )dt 0 siN Scalar Fig. Then.g. aM ≠ 0.4 Verify that even if two coefficients are not zero. 1 < i ≤ M.16.. To start with.. such that.16. 1 sM (t ) = − [ a1s1 (t ) + a2 s2 (t ) + . {si (t)}.. can be completely described by a subset of energy signals whose elements are linearly independent. + aM −1sM −1 (t )] aM =− 1 aM M −1 i =1 ∑ a s (t ) i i 4..16. e. 4.5 Version 2 ECE IIT. 0≤t<T 4. then s1(t) and s3(t) are dependent signals..2 Justification for G-S-O procedure Part – I: We show that any given set of energy signals. a1 ≠ 0 and a3 ≠ 0. there must exist a set of coefficients {ai}. let us assume that all si(t) -s are not linearly independent. Let us arbitrarily set. a1s1 (t) + a2s2 (t) + …… + aM sM (t) = 0. Then. 1 ≤ i ≤ M over 0 ≤ t < T. not all of which are zero. Kharagpur ..

16.8 s 21 = ∫ s2(t)φ 1(t )dt 0 T 4. 2.9 4. we can write 4.16.. Let {si(t)}. Part – II : We now show that it is possible to construct a set of ‘N’ orthonormal basis functions ϕ1(t). i = 1.. ….16. Now. . ϕ1 (t ) = 1 . eventually we will end up with a subset of linearly independent signals. following the above procedure for testing linear independence of the remaining signals.16. This set may be either linearly independent or not. energy of g2(t) = ∫ g 2 2 (t )dt 0 T Version 2 ECE IIT. i = 1.16. ϕN(t) from {si(t)}. This gives us a clue to determine the second basis function... there exists a set of {bi}. s11 = E1 Now.e. So.16.5 shows that sM(t) could be expressed as a linear combination of other si(t) – s.2. (M – 1). 0 ≤ t < T i i 4. T ∫ g 2 (t )ϕ1 (t )dt = ∫ s2 (t )ϕ1 (t )dt − s21 ∫ ϕ1 (t )ϕ1 (t )dt 0 0 0 T T = s21 − s21 = 0 → g 2 (t ) Orthogonal to ϕ1(t).. ϕ2(t). ….10 Let us now define an intermediate function: g2(t) = s2(t) – s21ϕ1(t). ….2.….. Kharagpur . (M – 1). we may express sM-1(t) as: 1 M −2 4. not all equal to zero such that. we verified that the function g2(t) is orthogonal to the first basis function. M −1 i =1 ∑ b s (t ) = 0.. where E1 denotes the energy of the E1 first signal s1(t). arbitrarily assuming that bM-1 ≠ 0. N. (M – 1)... from Eq.6 Again. i = 1.Eq.4.. If not. Next.ϕ1 (t ) = s11ϕ1 (t ) Where.7 sM −1 (t ) = − ∑ bi si (t ) bM −1 i =1 Now. 2. 0 ≤ t < T Note that.2…. i = 1. E1 = ∫ s12 (t )dt : 0 T ∴ s1 (t ) = E1 .16. we consider a reduced set with (M-1) signals {si(t)}. i = 1. 4. N ≤ M denote this subset. i. 2. s (t ) Let us choose the first basis function as.

16.13 ∫g 0 2 3 (t )dt gi (t ) Egi Indeed.15 Version 2 ECE IIT.s21 + s21 = E2 − s 2 2 21 ( Say ) So. we now set. j −1 2 0 ≤ t <T = s3 (t ) − [ s31ϕ1 (t ) + s32ϕ 2 (t )] where.ϕ 0 2 (t )dt = 0 . g3 (t ) = s3 (t ) − ∑ s3 jϕ J (t ) . i. Kharagpur .16.…. s31 = ∫ s3 (t )ϕ1 (t )dt and s32 = ∫ s3 (t )ϕ 2 (t )dt 0 0 T T It is now easy to identify that. For i=3.11 ϕ 2 (t ) = g 2 (t ) T = s2 (t ) − s21ϕ1 (t ) E2 − s212 4.s21.16.12 ∫g 0 2 2 (t )dt and Verify that: E2 = ∫ s2 2 (t )dt : Energy of s2(t) 0 T T ∫ϕ 0 1 2 2 (t )dt = 1 . g3 (t ) ϕ3 (t ) = T 4.e. ϕi (t ) = gi (t ) T = 4..16..16. ϕ3(t).e. N. i. 2. where gi (t ) = si (t ) − ∑ sijϕ j (t ) j =1 4.= ∫ [ s2 (t ) − s21ϕ1 (t )] dt 2 T = ∫ s2 2 (t )dt − 2.14 ∫g 0 2 i (t )dt i −1 for i = 1. we can determine the third basis function. 4. φ1(t) and φ2(t) are orthonormal to each other. Proceeding in a similar manner. φ2(t) is a time limited energy signal of unit energy. in general. T and ∫ ϕ (t ).s21 ∫ s2 (t )ϕ1 (t )dt + s212 ∫ ϕ12 (t )dt 0 0 0 0 T T T = E2 − 2.

2. Version 2 ECE IIT. N and j = 1. 2.. • Find the energy of s1(t) as this energy helps in determining the first basis function ϕ1(t).…….. Signal Constellation is the collection of M signals points (or messages) on the signal space. Kharagpur .. for a convenient set of {ϕj (t)}. which is a normalized form of the first signal. known as the ‘Signal Space’ (Fig. we can represent a signal si(t) as a column vector whose elements are the scalar coefficients sij. j = 1.17 These M energy signals or vectors can be viewed as a set of M points in an N – dimensional Euclidean space. energy of the second signal (E 2).ϕ j (t )dt 0 T 4.2. i= 1. M ⎢ ⎥ ⎢ ⎥ ⎢ ⎣ siN ⎥ ⎦1× N 4. {si(t)}. 2. • Find the scalar ‘s21’. …. M. Concept of signal space Let... • Derive a subset of linearly independent energy signals. 2. 2. si (t ) = ∑ sijϕ j (t ). M Let us summarize the steps to determine the orthonormal basis functions following the Gram-Schmidt Orthogonalization procedure: If the signal set {sj(t)} is known for j = 1. Note that the choice of this ‘first’ signal is arbitrary.16.3).16. such that.….16. 2. i= 1.…. j =1 N sij = ∫ si (t )ϕ j (t )dt 0 T Now.……. j = 1.. 0 ≤ t <T.and sij = ∫ si (t ).4.. N : ⎡ si1 ⎤ ⎢s ⎥ si = ⎢ i 2 ⎥ . a special function ‘g2(t)’ which is orthogonal to the first basis function and then finally the second orthonormal basis function ϕ2(t) • Follow the same procedure as that of finding the second basis function to obtain the other basis functions.N and 0 ≤ t <T.i = 1. N ≤ M.….….16 for i = 1.. …. M and 0 ≤ t < T. 2.

s j cos angle between si & s j = si s j ( ) ( ) 4. s j = 0 .19 ∴ si & s j are orthogonal to each other if si .3 Sketch of a 2-dimentional-signal space showing three signal vectors s1 .16. 4. Kharagpur . s 2 and s3 Now. ( ) If E i is the energy of the i-th signal vector. the length or norm of a vector is denoted as si .ϕ2 E2 E1 s2 0 s1 s11 s12 ϕ1 (t ) s3 E3 2D Signal Space Fig.16.16.18 The cosine of the angle between two vectors is defined as: si . The squared norm is the inner product of the vector: 2 si = si . si = ∑ s 2ij j =1 ( ) N 4. Version 2 ECE IIT.

. One of the M signals si(t).…. the received symbols are usually corrupted and once placed in the signal space.2. (i. then an obvious choice for maximum distance of separation is. much of the features of a modulation scheme.20 2 N T 2 2 For a pair of signals si (t) and sk(t). they may not match with the valid signal points in some respect or the other.M is transmitted in each time slot of duration ‘T’ sec.…. s1(t) = -s2(t). i=1. For example. r ( t ) =si ( t ) +w ( t ) . Kharagpur . Detection is on a symbol-by-symbol basis. Additive White Gaussian Noise (AWGN) processes W (t) is assumed with a noise N sample function w (t) having mean = 0 and power spectral density 0 [N0: 2 single sided power spectral density of additive white Gaussian noise. We map the received sample function r (t) on the signal space to include a ‘received vector’ or Version 2 ECE IIT. equivalently.e. si -sk is as much as possible to ensure that their detection is more robust even in presence of noise. The job of the receiver is to make “best estimate” of the transmitted signal si (t) (or. is very useful for designing a receiver as well. in absence of any noise. 2.e. Let us assume the following: 1. we may write. if R (t) denotes the received random process with a sample function r (t). Let us briefly consider the task of a good receiver in such a situation. Now. All symbols are equally probable.2. 3. i. So. si -sk = ∑ ( sij -s kj ) = ∫ ⎡ ⎣ si ( t ) -s k ( t ) ⎤ ⎦ dt j=1 0 It may now be guessed intuitively that we should choose si (t) and sk (t) such that the Euclidean distance between them.Ei = ∫ T 2 i 0 s ( t ) dt = N T ∫ T 0 ⎡N ⎤⎡ N ⎤ s φ t ( ) ⎢ ∑ ij j ⎥ ⎢ ∑ sik φ k ( t ) ⎥ dt ⎦ ⎣ j=1 ⎦ ⎣ k=1 = ∑∑ sij sik ∫ φ j ( t )φ k ( t ) dt as {φ j ( t )} forms an ortho-normal set j=1 k=1 0 N 2 = ∑ sij = si j=1 2 N 4. for all ‘i’. Noise is discussed more in next two lessons] 4. they are equidistance from the origin).. the corresponding message symbol mi) upon receiving r(t). Use of Signal Space for Signal Detection in a Receiver The signal space defined above. such as the number of symbols used and the energy carried by the symbols. the receiver should detect one of these valid symbols only.e. In a sense. 0 ≤ t < T and i = 1. However. is embedded in the description of its signal space. if s1 (t) and s2 (t) have same energy E. the probability of occurrence of si(t) = 1/M.M.16. i.

The detection problem can now be stated as: ‘Given an observation / received signal vector ( r ). We discuss this important detection scheme in Lesson #19. Kharagpur . Q4. Maximum Likelihood Detection scheme provides a general solution to this problem when the noise is additive and Gaussian. the receiver has to perform a mapping from r to an estimate m for the transmitted symbol mi in a way that would minimize the average probability of symbol error’. Version 2 ECE IIT. S2(t) = Cos (2πft + π/3) and S3(t) = Sin 2πft. S1(t) = Cos2πft.16.‘received signal point’. This helps us to identify a noise vector. ∧ Problems Q4.also. which are orthonormal to each other over 1 sec. Verify that Eq4.3) Consider a binary random sequence of 1 and 0.16.3 is valid.16.16. Draw a signal constellation for the same.1) Sketch two signals. Comment whether the three signals are linearly independent? Q4.2) Let. w(t) .

- Simple Network Management Protocol - Wikipedia, The Free Encyclopedia
- Parameter Based Anomaly Detection System With H-Parameter Estimation
- Bonafide IIITA Certificate
- Mobile data management PPt
- Reliablity Based Channel Assignment using Genetic Algorithm
- Smart Shopper
- Women s Employment
- Neural Based Face Detection
- Preserving national animal using wireless sensor network based hotspot algorithm
- Software Project Management
- Direct memory access
- Victimology New
- Hadith Sunnah
- FOSSination

This is a very good paper on the representation on of signals

This is a very good paper on the representation on of signals

- Quantum - Algebra Lineal.pdf
- Lab 9 Solns
- Stijn Lievens, Nedialka I. Stoilova and Joris Van der Jeugt- A class of unitary irreducible representations of the Lie superalgebra osp(1|2n)
- Hodge Dual
- lag8
- HW.pdf
- GDF
- The Identity Operator
- Encoding of Images Based on a Lapped Orthogonal Transform
- Entropy-Based Algorithms for Best Basis
- Barreira_pspm Lectures on Lyapunov Exponents Pesin
- Basis and Dimension
- qmassgn16173(quantujm 3)
- Vector Tutorial
- ÌÇÊ ËÈ
- vtf_0910.pdf
- MAT242_Final_2013F
- vector-cal=1
- Application of Zorn lemma
- Mscappmath2014 Solutions
- js_lect3.pdf
- 15UC01 2
- Wavelets and Subband Coding Solutions Manual
- nomentbl
- Kriya Reiki = Lesson 3
- CMPT1998-21
- 10.1.1.3
- Original Reiki Manual
- Configure Symbols for components.pdf
- Reiki Symbols - How to Draw Reiki Symbols
- 16_Representation of signals.pdf

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue reading from where you left off, or restart the preview.

scribd