Proceedings of the First Young Vietnamese Scientists Meeting (YVSM’05), Nha Trang, June 12-16, 2005 MUSIC For OFDM By Using An Uniform Antenna Array With Two Elements TRAN CAO QUYEN Research center for Electronics and Telecommunications, Coltec 203, G2, 144 Xuan Thuy Road, Cau Giay District, Hanoi, Vietnam Email: tcquyen@yahoo.com , Tel: 04-7547728 Multicarrier communication (MC) has capability of high speed transmission, fading combating, simplixity of equalization leading to its widely application. Among many MC application, OFDM(Othorgonal Frequency Division Multiplex) is a very promising candidate. In addition, problem of DOA( Direction of Arrival ) estimation using MUSIC algorithm is solved for point source and futher for sources with pertubation.This paper presents the application of MUSIC algorithm for DOA estimation for OFDM communication. The simulation results in AWGN channel are shown. I. INTRODUCTION MUSIC is short of Multiple Signal Classification proposed in [1].This is one of famous subspace based array processing to DOA estimation (Direction of Arrival) and even strength, polarization characteristic of the incoming signal. In MUSIC, for point source the autocovariance matrix of incoming signals on antenna array is considered. The rank of that matrix depends on number of array elements corresponding to number of independent uncorrelated inputs. In previous research of DOA determination, it is proved that MUSIC is the most efficiently compared to Maximum Likelihood Estimation [2], ESPRIT [3], and others having the ideas based on them. For OFDM signal, it consists of components having its amplitude independently and its frequency orthorgonality [4],[5],[6]. Especially, in OFDM all subcarriers are transmitted simultaneously in the channel so transmission speed is increased and there is a capability of application variety modulation scheme for individual subcarriers. In this paper, the problem of DOA estimation for OFDM using an uniform antenna array with two elements will be presented. II. MUSIC ALGORITHM MUSIC is short of multiple signal classification proposed in[1]. In this paper, the correlation matrix of output signal is given by { } ) 1 ( x x R H xx E = Therefore, the relation between incoming signal and output signal can be expressed as ) 2 ( 2 I A A R R ss xx δ + =