IJRECE VOL. 2 ISSUE 4 OCT-DEC 2014 ISSN: 2393-9028 (PRINT) | ISSN: 2348-2281 (ONLINE) INTERNATIONAL JOURNAL OF RESEARCH IN ELECTRONICS AND COMPUTER ENGINEERING A UNIT OF I2OR 71 | Page Generation of Sequences with Good Correlation Properties Using Modified Bat Algorithm P. Ramya Raju 1 , Dr.P.Siddaiah 2 , S. Srinivasa Rao 3 1,3 Department of ECE, MGIT,Hyderabad, AP, India 2 Department of ECE, ANUCT, Guntur, AP, India Abstract: Radar applications require sequences with individually peaky autocorrelation. It is a combinatorial problem to obtaining such sequences. So designing a signal above referred is a challenging problem for which many global optimization algorithms like genetic algorithm, particle swarm optimization algorithm, simulated annealing, tunnelling algorithm were reported in the literature. The paper aims at the design of optimal set of Binary Sequences using Modified Bat Algorithm which makes use of Hamming Scan Algorithm for Mutation. The main advantage of Bat algorithm with HSA Algorithm is it enhances the search space of MBA, thereby preventing the local optimum trapping of MBA. The synthesized Binary Sequences using Modified Bat algorithm have better autocorrelation properties than Bat algorithm. Keywords—Auto-Correlation Sidelobe Peaks (ASP), Bat Algorithm (BA), Optimization, Modified Bat Algorithm (MBA), Discrimination Factor (DF), Hamming Scan Algorithm (HSA). I. INTRODUCTION Pulse compression is a method of breaking the unwanted constraint between resolution and range [4]. Pulse compression radar transmits a phase or frequency modulated pulse, which have both long (having good range characteristics) and wideband (having good range resolution criterion). The received echo signal is allowed to be processed in a matched filter that becomes a short pulse by compressing the long pulse which separates targets which are very close. However, it hides weak targets as separation is achieved at the cost of sidelobes. Therefore, in radar applications high discrimination factor of the pulse compression sequences are desired to achieve high range resolution. Range resolution is the ability of identifying closely spaced targets by radar receiver. The term Discrimination (D) is defined as the ratio of main peak in the auto-correlation to absolute maximum amplitude among side lobes, Moharir [2]. In this context, Barker [3] proposed binary sequences with sidelobe levels either unity or zero. In Barker codes the discrimination is equal to the length of the code. Unfortunately, lengths greater than 13 have not been found in Barker sequences. In fact the discrimination is the figure of merit of the code; many researchers have started working to design binary sequences with discrimination greater than 13. II. BINARY PHASE CODES If the sequence gets peaky main lobe energy for zero shifts and zero sidelobe energy for other shifts then the performance of a sequence is optimal. Different methods such as Merit factor, Discrimination factor, Energy efficiency and Quality factor are used to evaluate the goodness of pulse compression sequences. Let S = [x0, x1, x2, x3 ………..xN-1] be a real sequence of length N, with values of 1 and -1.  k i k n i i x x k r       1 0 (1) Where k=0, 1, 2… N -1 is its aperiodic autocorrelation. The Discrimination factor is defined as the ratio of amplitude of main peak of the auto correlation to the absolute maximum amplitude in the side lobes [7]. DF= ሺ଴ሻ  ≠0 |ሺሻ| (2) Discrimination is used to know whether a coded signal is good or poor. This means that a code with low discrimination is a poor code while a code with high discrimination is good code. DF is used as cost function in the paper. To measure how the main lobe signal is different from the peak side lobe level, discrimination is used and it gives the measure of goodness of the given coded waveform. III. BAT ALGORITHM From several decades optimization has been an active area of research because many real-world optimization problems become more complex, better optimization algorithms were needed. BA, bat algorithm is a population based algorithm. BA has been applied to several real-world problems; it is an optimization technique which is well adapted to the optimization of nonlinear functions in multidimensional space. BA, bat algorithm is mainly based on hunting behavior of bats; it is a relatively new population based metaheuristic approach. In this algorithm possible solution of the problem is represented by bat positions. Quality of the solution is indicated by the best position of a bat to find its prey. Bats are fascinating animals; mainly micro bats use a type of sonar, called, echo location, to detect prey, avoid obstacles, and locate their roosting crevices in the dark. They can emit a very loud and short sound pulse and it hits an object and, after some time it listens for the echo that bounces back from the surrounding objects. Most bats use short, frequency-