Designing radar signals using complementary sequences
Srdjan Budisin, Branislav M. Popović, I. M. Indjin · 1987
S. Z. Budisin, B. M. Popovic, I. M. IndjinInstitute of Applied Physics, Novi Beograd, YugoslaviaINTRODUCTIONIn pulse compression radars with binary phase modulationBarker codes (1) and PN sequences (Golomb (2), McWilliams et al(3), Sarwate et al (4) ) are very often used and are wellunderstood. Much less known are complementary sequences, whoseaperiodic autocorrelation side-lobes are as good as for PNsequences as shown by Gutleber (5). They are much more numerousand have the complementary property. After the original paper fromGolay (6), not much theoretical work has been published on binarycomplementary (Golay) sequences, despite the obvious fact thatthey have a very regular structure. Anyhow, a lot of relatedpapers (Tseng et al (7), Sivaswamy (8), Sarwate (9), Popovic et al(10), Budisin (11)) witness the rising interest in complementarysequences. Complementary sequences are useful intelecommunications (5), as well as in radar applications (Skolnik(12)).Two sequences, a(i) and b(i), are called complementary iftheir autocorrelation functions R (i) and R (i) have the followinga bcomplementary property:R (i) + R (i) = O for i $ O (1)a bwhere R (i) and R (i) are the corresponding autocorrelationa bfunctions.The radar signal that exploits this property, as described in(8), consists of two pulses separated by the pulse repetition time(T ) and modulated by two complementary sequences as shown inprfigure 1.a..The autocorrelation function of this signal is shown infigure 1.b.Obviously, for ranges which correspond to t < T -T , this ispr san ideal autocorrelation function which provides ideal pulsecompression. Unfortunately, the ambiguity function for w $ O is