25.08.2021 Views

082-Engineering-Mathematics-Anthony-Croft-Robert-Davison-Martin-Hargreaves-James-Flint-Edisi-5-2017

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Eliminating the unwanted values inthe first column, wefind

R 1

R 2

→R 2

+R 1

⎝ 2 −1 1 2

0 0 2 6 ⎠

R 3

→R 3

−3R 1 0 0 −5 −15

8.10 Gaussian elimination 291

Entriesunderthesteparenowzero.Toreducethematrixtoechelonformwemustensure

each successive row has more leading zeros than the row before.We continue:

R 1

R 2

⎝ 2−112

0 026⎠

R 3

→ 2R 3

+5R 2 0 0 0 0

which isnow inechelon form.

In this form there is a row which starts off with a non-zeroxvalue, that is the first

row, there is a row which starts off with a non-zerozvalue, but no row which starts off

withanon-zeroyvalue.Therefore,wechooseytobethefreevariable,y = λsay.From

thesecondrowwehavez = 3andfromthefirst2x −y+z = 2,sothat2x = λ−1,that

isx = (λ −1)/2.

Engineeringapplication8.2

TheVandermondeMatrix

Many data storage devices, for example the Blu-ray disc , and data transmission

standards, for example WiMAX ® , have built-in techniques to reduce the effects of

errors which occur during normal use. These errors originate from noise and interference

and they can result in the loss of data. One such technique is the use of an

error-correctingcodetoprovideameasureofprotectionagainstdataerrors.Errorcorrecting

codes are used to encode the data when it is stored or transmitted. The

data is then decoded when it is read or received. In the case of a transmission line,

theerror-correctingcodemakesitpossibletodetecterrorsinthereceivedsignaland

tomake corrections,so thatthe errors arenotsubsequently retransmitted.

Oneimportantclassoferror-correctingcodesarebasedonReed-Solomoncodes,

details ofwhicharebeyond the scopeofthistext.However animportantmathematical

concept usedinReed-Solomoncodes is thatofaVandermondematrix.

The Vandermonde matrix can be illustrated by considering the problem of representing

a signal by a polynomial. For example, to approximate a signal f (t) by a

second-degree polynomialwewrite

f(t)≈a 0

+a 1

t+a 2

t 2

wherea 0

,a 1

anda 2

are coefficients of the polynomial which must be found. These

are found by forcing the original signal f (t) and its polynomial approximation to

agree at three different values oft, sayt 0

,t 1

andt 2

. This gives rise to the following

systemofequations:

1t 0

t 2 ⎞ ⎛ ⎞ ⎛ ⎞

0 a

⎝1t 1

t1

2 ⎟ 0

f(t 0

)

⎠ ⎝a 1

⎠ = ⎝f(t 1

) ⎠

1t 2

t2

2 a 2

f(t 2

)

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!