Experimental verification of real‐valued orthogonal PN sequence applied to pulse compression sonar

Abstract

In the pulse compression sonar, the code with a sharp autocorrelation function is used as in spread spectrum communication, and they share many common technologies. In this paper, the real-valued orthogonal pseudo-noise (PN) sequence that can be used in both over a broad frequency range including dc is applied to the pulse compression sonar. By adapting digital correlation processing, a high performance close to the theory has been obtained. The codes are generated from a binary orthogonal PN sequence with a length of 127 and a multivalue orthogonal PN sequence with a length of 256 and are sent into an acoustic transmission path at a frequency of about 40 kHz. The correlation operation of the received code and the transmitted code is carried out on a personal computer. The multiple echo is detected at a low noise while the Gaussian noise and the locked sinusoidal wave can be suppressed with a processing gain close to the one predicted by the theory.

A real‐valued periodic orthogonal sequence derived from a real‐valued shift‐orthogonal finite‐length sequence and its fast periodic correlation algorithm

“…In addition, we noted when the end point correlation value is negative {a 2,0,i } and positive {a 2 Here, care must be taken that the end point correlation value ε m=1 g of the finite-length sequence {a m,0,i g } obtained by inserting a 0 value is equal to the end point correlation value ε 1 g of the finite-length sequence {a 2,0,i g } that has the original length 2. A proposal that forms a partial sequence using only a partial sequence from a finite-length sequence of length 2 in this manner is called a proposal by length 2.…”

“…Utilizing this convolution, we convoluted a finitelength sequence {a 2 The end point correlation value must satisfy 0 < |ε 1 | ≤ 1/2, however. In addition, we noted when the end point correlation value is negative {a 2,0,i } and positive {a 2 Here, care must be taken that the end point correlation value ε m=1 g of the finite-length sequence {a m,0,i g } obtained by inserting a 0 value is equal to the end point correlation value ε 1 g of the finite-length sequence {a 2,0,i g } that has the original length 2.…”

“…In order to avoid complexity when describing a procedure to obtain N sets of correlation output {y i′ } from N sets of input {x i }, real-valued orthogonal periodic sequence {â N,l ,i } and the sequence numbers of periodic partial sequences {b 2 processes. Since a convolution is a commutative operation, the final effect will not change even if the order of the convolution changes.…”

“…represent parametric variables used for correlation processing of periodic partial sequences {b 2,i }, {b 2…”

“…A periodic sequence, such as an impulsive periodic autocorrelation function, is suitable for repetitive type pulse compression radar that repeatedly emits encoded impulse trains [1,2]. A two-valued orthogonal periodic sequence of {+1, -1} exists.…”