Search for Low Autocorrelation Binary Sequences

Abstract

Binary sequences with low autocorrelation are very important for many applications in communication and radar systems, while the construction of such sequences is one of the most challenging problems in signal design theory. In this letter we propose a Local Optimum Shift (LOS) algorithm to search for low autocorrelation binary sequences. The LOS uses an ingenious operation to exploit and explore the search space efficiently. Some new binary sequences with the best known Peak Sidelobe Level (PSL) are found out by the LOS. For binary sequences of lengths up to 100, the LOS can generate promising sequences with higher merit factor and lower PSL quickly.