Research Output
Hybridization of modified sine Cosine and Cuckoo search algorithms for binary sequences in radar communication system
1 - 1 of 1
Publication
Finding binary sequences with high Merit Factor (MF) is a main problem in Low Autocorrelation Binary Sequences (LABS) for range compression radar to provide information concerning the target's location, motion, size and other parameters. MF is a measure of the quality of the sequence with low Energy level (E) and flexible length of binary sequences. The largest binary sequence of high MF is known as a supply for a good initial bound for exact search methods. This study attempted to solve the constraints of Sine Cosine Algorithm (SCA) that may have redundancy values of search agent and tendency to trap in local optima for LABS in achieving high MF. The main contribution of this work is the Hybridization of Modified Sine Cosine and Cuckoo Search Algorithms (HMSCACSA) to develop an improved identification capability compared to the conventional SCA. Second contribution is the HMSCACSA featuring the Inverse Filtering (IF) for LABS. In this case, SCA was modified using Latin Hypercube Sampling (LHS) to identify SCA random solution. Therefore, Modified SCA (MSCA) serves to
guide SCA in obtaining a better local optimum in the exploitation phase with a fast convergence achievement through an optimum value of the solution. HMSCACSA lead to an excellent combination to determine the ideal nest location and a new position to achieve better results in the search process. The HMSCACSA optimizer was tested over fourteen (14) benchmark experimental functions within a classical category of 24 significant benchmark issues, a current variety of benchmark issues IEEE CEC 2017. The effectiveness of HMSCACSA was validated in terms of performance and statistical analysis, which included hybrid-to-hybrid metaheuristic, Particle Swarm Optimization- Grey Wolf Optimizer (PSOGWO), particle swarm optimization-artificial bee colony (PSOABC), and particle swarm optimization-gravitational search algorithm (PSOGSA) to ensure that the comparison between the metaheuristics was fair. Compared to other hybrid metaheuristics, the majority of the used benchmark mathematical functions indicated that the simulation of the HMSCACSA achieved faster convergence curves by 63.89%. In LABS, HMSCACSA using IF method was proposed to achieve better results with two large MF is 12.12 (LABS-IF) and 12.6678 (HMSCACSA-IF) were obtained for both lengths 231 and 237 where both of the optimal solutions belong to the skewsymmetric sequences. Now, the MF was outperformed up to 24.335% and 2.708% against the state-of-the art benchmark algorithms, xLastovka and Golay. These results indicated that the simulation of the proposed algorithm had quality solutions in a fast convergence curve, better optimal means and standard deviation. The HMSCACSA-IF has a potential to increase the target detection capability with low energy and good range resolution that has low autocorrelation function in LABS.