Frequency Hopping Sequences with Optimal Partial Autocorrelation ...

4 downloads 0 Views 105KB Size Report
Jul 1, 2004 - Optimal criteria on partial Hamming autocorrelation. • Partial Hamming correlation function for a period N and a correlation window length.
Coding and Information Theory Lab.

1

Frequency Hopping Sequences with Optimal Partial Autocorrelation Properties

July 1, 2004.

Yu-Chang Eun, Seok-Yong Jin, Yun-Pyo Hong, and Hong-Yeop Song Yonsei University Seoul, Korea

Yonsei University Electrical and Electronic Eng.

Coding and Information Theory Lab.

1

Outline



Motives



FH Systems



Strictly-Optimal FH Sequences



Summary & Remarks

Yonsei University Electrical and Electronic Eng.

Coding and Information Theory Lab.

2

Motives

• Most of FH sequences so far have been designed so that – their maximum periodic Hamming correlation is minimized – with the number of hopping slots (frequencies) that is a power of a prime.

• Usually, the correlation window is shorter than the period of the FH sequence. ⇒ A sequence having good partial Hamming autocorrelation ?

Yonsei University Electrical and Electronic Eng.

Coding and Information Theory Lab.

3

Tx & Rx structure of a FH system J (t )

Tx Binary Data

Binary to M-ary Mapper

M-ary Data

d (t )

MFSK Modulator

Frequency Hopping

c (t ) PN Code Generator

Rx

Frequency Synthesizer s (i )

Frequency Dehopping

d' (t ) to noncoherent data detector

PN Code Aquisition & Tracking

c' (t )

PN Code Generator

Frequency Synthesizer s (i )

• Correlation window length – usually shorter than the period of the FH sequence due to the limited synchronization time or hardware complexity – may vary depending on the channel condition

Yonsei University Electrical and Electronic Eng.

Coding and Information Theory Lab.

4

FH sequences with optimal partial autocorrelation properties ¦ Optimal criteria on partial Hamming autocorrelation • Partial Hamming correlation function for a period N and a correlation window length L starting at t,

HXY (τ ; t | L) =

t+L−1 X

h[x(j), y(j + τ )], 0 ≤ τ < N

(1)

j=t

where h[x, y] = 1 if x = y and h[x, y] = 0 if x 6= y. 0

t

t+

t + +L 1

t

t+L 1

Xˆ X 0

Yonsei University Electrical and Electronic Eng.

Coding and Information Theory Lab.

5

• The maximum of the partial Hamming autocorrelation function (p-HAF) H(X | L) =

max

0

Suggest Documents