International Journal of Academic Scientific Research ISSN: 2272-6446 Volume 7, Issue 1 (February - March 2019), PP 33-46 www.ijasrjournal.org www.ijasrjournal.org 33 | Page Improving Performance of Direction of Arrival Estimation Using Sparse Arrays in Smart Antenna System Abed Alnaser Ramadan 1 , Abdel Fattah Fares 2 and Rami Khal 3 1 Postgraduate student (PhD), Faculty of Electrical and Electronics Engineering, University of Aleppo, Syria 2 Department of Electronics Engineering, Faculty of Electrical and Electronics Engineering, University of Aleppo, Syria 3 Department of Communications Engineering, Faculty of Electrical and Electronics Engineering, University of Aleppo, Syria ABSTRACT: This paper indicates importance of using sparse arrays (SA) in direction of arrival (DOA) estimation algorithms in smart antenna system (SAS). Analytical study of sparse arrays is introduced, which include coprime array, extended coprime array, nested array, coprime array with compressed interelement spacing (CACIS), and coprime array with displaced subarrays (CADiS).Paper evaluates these sparse arrays using their difference coarray equivalence and derives the analytical expressions of the coarray aperture, the achievable number of unique lags, the maximum number of consecutive lags and degree of freedom (DOF). Compared to uniform arrays with ( ) sensors, sparse arrays increase the degree of the freedom from ( )to ( ). For comparison of performance of these sparse arrays, numerical example is introduced, where the results indicate that nested array structure provides coarray with unique lags (that are all consecutive), which are larger than that of prototype and extended coprime. Results also indicate that the CACIS structure yields flexibility in trade-off between unique lags and consecutive lags, whereas the CADiS structure allows the minimum interelement spacing to be much larger than the typical half-wavelength requirement, but at the expense of a decrease in consecutive lags. Furthermore, the nested CADiS slightly outperform the nested CACIS due to the higher number of consecutive lags achieved. We propose the scheme for DOA estimation using suitable sparse arrays with SS-MUSIC or LASSO algorithms. According to results, we can choose suitable sparse array and DOA estimation algorithm in SAS depending to the radio situation and the purpose of this SAS. All mentioned arrays and algorithms are simulated using MATLAB. Results of simulations support the theoretical expressions. Keywords: Sparse Arrays (SA), Coprime Arrays, Nested Arrays, CACIS Structure, CADiS Structure. 1. INTRODUCTION The Smart Antenna System (SAS) embeds the antenna elements and the digital signal processing unit which enables it to form a beam for a desired direction taking into account the multipath signal components. Hence, signal to interference and noise ratio (SINR) can be improved due to the nulls produced towards the interferers in the direction of signal of noninterest (SonI) and the overall spectrum efficiency can be increased [1]. It is common in practice that the number of sources to be estimated is larger than the number of sensors in the array. However, the degree of freedom (DOF) of the conventional source estimation algorithms is limited by the number of sensors. In general, an array antenna with physical sensors can identify up to −1 sources. To detect more sources, additional sensors are required to increase the achievable DOF, which leads to an increase in complexity. Therefore, an active research topic has been focused on how to increase the DOF for source estimation [2, 3]. Sparse arrays (SAs) open a new approach to sensor array processing due to the higher degree of freedom offered in the difference-coarray domain. Coprime arrays and nested arrays are examples of sparse arrays obtained from a union of two uniform linear arrays (ULAs) with different interelement spacing. The increased degree of freedom has been used to identify ( ) sources from only sensors [4, 5, 6]. In addition to coprime and nested arrays, generalized configurations of the coprime array concept are considered, which comprise two operations. The first operation is the compression of the interelement spacing of one subarray in the coprime array by a positive integer. The resulting coarray structure is referred to as coprime array with compressed interelement spacing (CACIS). The second operation introduces a displacement between the two subarrays, yielding a coprime array with displaced subarrays (CADiS) [2].