Orthogonal Cyclic Prefix for Time Synchronization in MIMO-OFDM

Orthogonal Cyclic Prefix for Time Synchronization in MIMO-OFDM Gajanan R. Gaurshetti & Sanjay V. Khobragade Dr. Babasaheb Ambedkar Technological University, Lonere E-mail : gaurshetty@gmail.com, svk2305@gmail.com Abstract - Time synchronization is an important factor in Multiple-Input-Multiple-Output (MIMO) Orthogonal Frequency Division Multiplexing (OFDM) as it is an efficient and popular communication system. A new method is proposed in this paper for time synchronization using optimal size of orthogonal cyclic prefix in MIMO- OFDM. Simulation result of proposed method shows acquisition probability for time synchronization is 1 when SNR -5 db. Keywords - Cyclic Prefix (CP), MIMO-OFDM, Orthogonal codes, Synchronization. I. INTRODUCTION One of the fastest growing areas in consumer electronics is multimedia application based on wireless communication for Metropolitan Area Network (MAN) [1]. A rapidly increase in demand of high data rates to support new features, advanced functionality and better services for multimedia need continuously evolution in wireless MAN [3]. On account of its potential in supporting high data rate transmission over frequency selective fading channels and the simplicity it offers in channel equalization, Orthogonal Frequency Division Multiplexing (OFDM) has become an attractive technique for wideband communications [2]. With the introduction of multiple antennas in the realization of Multiple-Input Multiple Output (MIMO) OFDM, the transmission capacity is largely expanded as the number of spatial communication channels is multiplied. The MIMO (Multiple Input, Multiple Output) system is one of several forms of smart antenna technology for wireless communications in which multiple antennas are used at the both side of transceiver. MIMO system is used to increase link capacity by sending different data stream over different transmit antenna or to improve the link reliability by sending the same data stream over different antenna using Space Time Block Code (STBC) [4]. With help of STBC in MIMO-OFDM we can achieve high data rate but another concern is synchronization. Synchronization must be established between transmitter and receiver for better reception of massage symbols. The synchronization should be done in frequency as well as in time. The frequency synchronization is done by estimating carrier frequency offset of local oscillators between transmitter and receiver and then proper modulation demodulation process. To detect the beginning of the each frame, time synchronization is required. If there is fine time synchronization then beginning of useful data frame can be identified [5]. Orthogonal codes as preamble are used for time synchronization. These preambles with massage symbols forms a frame and these frames are transmitted. At the receiver side local sequences are correlated with these received frames and thus the beginning of the frame is identified where the correlation peak is detected. The rest of the paper is organized as follows, In Section-II we describe the MIMO-OFDM system with parameters of Wireless MAN standard, and Section-III is to represent different conventional orthogonal sequences for time synchronization. In Section-IV We describe the proposed synchronization technique. Finally, simulation result and conclusion are done in Section-V and Section-VI respectively. II. MIMO-OFDM SYSTEM A system model of MIMO-OFDM used for simulation is shown in Figure.1. The MIMO-OFDM system is consisting of N t X N r where N t is no. of transmit antennas and N r is no. of receive antennas. In this paper we used MIMO-OFDM of 2X1. There are different blocks are used in given system description of each block is given bellow. 81

A. Random Integer Generator The Random Integer Generator block generates uniformly distributed random integers in the range [0, M-1], where M is the M-ary number. Here M=2 is used. B. Interleaver The Interleaver block rearranges the elements of its input vector without repeating or omitting any elements. Parameters for interleaver are no. of sub-channels are 16 and N cbps = 192, N cpc = 1.. C. OSTBC Encoder OSTBC is an Alamouti s scheme can be applied to the system with two transmit antennas and its spacetime encoding can be described as follow, (1). Where S1 and S2 are complex signals to be transmitted and * denotes a conjugate operation. Rows indicate the time domain and columns represent the space domain. A. Pilot and Guard Band Insert Pilot and Guard Band are used for channel estimation at receiver side. B. IFFT The IFFT block computes the inverse fast Fourier transform (IFFT) of each row of a sample-based 1-by-P input vector, or across the first dimension (P) of an N-D input array. IFFT size is 256. C. Add Cyclic Prefix In MIMO-OFDM cyclic prefix is used to remove inter symbol interference. The cyclic prefix is nothing but tail symbols of massage frame and it prepended to Fig.1 : MIMO-OFDM System that message frame. Here length of CP is one fourth of the FFT/IFFT size and it is 256/4 = 64. D. Rayleigh Fading Channel Rayleigh Fading channel represents wireless fading channel. We have considered 5 multipath and their delay between different multipath in microsecond and power gain of each multipath is (0, 0.3, 0.15, 0.31, 0.37) and (0, -1.5, -1.4, -3.6, -0.6) respectively. Rayleigh channel coefficient between transmitter N t and receiver N r is given by (2) Where M rt is the number of multipath between N t transmitter and N r receiver α Lrt, δ Lrt are the gain and delay of the path L rt respectively. And output of transmitter is given by 82

(3) Where N rt is the AWGN signal noise and S r (t) is given by (4) Where at (t) is output of OSTBC, f is carrier frequency and θ is carrier phase. III. ORTHOGONAL CODES In order to get fine time synchronization we must chose such orthogonal code sequence whose autocorrelation and cross-correlation is very good. Most popular orthogonal codes used for time synchronization are stated bellow. A. Gold Sequence The Gold sequences are defined using a specified pair of sequences u and v, of period N = 2 n - 1, the set G (u, v) of Gold sequences is defined by Where T represents the operator that shifts vectors cyclically to the left by one place and represents addition modulo 2. Note that G (u, v) contains N + 2 sequences of period N [7]. Gold sequences have the property that the cross-correlation between any two, or between shifted versions of them, takes on one of three values: -t(n), -1, or t(n) - 2, where t(n) is (6) (5) The disadvantage of Gold Sequence is inter-code interference due to high cross-correlation. C. CAZAC Sequence CAZAC sequence is nothing but constant amplitude zero autocorrelation sequence. CAZAC sequence is represented as follow When L = even When L = odd (8) Where L is length of CAZAC sequence and M ϵ N which is smaller than L and range of n is 0 < n < L-1 [8]. Autocorrelation and cross-correlation function of both Hadamard sequence and CAZAC sequence are good but the acquisition probability for time synchronization is less than the proposed orthogonal code. A. Orthogonal Code IV. PROPOSED TECHNIQUE The Orthogonal code that is being used in given system model is orthogonal variable spreading factor (OVSF) code. OVSF code is defined as the rows of an NXN matrix and is given by (9) Where C N is defined only for N = 2 r at depth r for C 1 = [1]. C i+1 can be constructed with the help of C i where i<r B. Hadamard Sequence Hadamard code is generated from hadamard matrix whose entries are +1 or -1 so called bipolar and the rows of matrix are orthogonal. This matrix is given as (7) The N-by-N Hadamard matrix has the property that H N H N T = NI N where I N is the N-by-N identity matrix [8]. and is given by Fig. 2 : Cyclic Prefix structure (10) Orthogonal Variable spreading factor has very good autocorrelation and cross-correlation function and very 83

easy to generate these sequences. We can easily change the length of these sequences. B. Cyclic Prefix structure In conventional method large length of preamble are used for time synchronization and because of large length preamble data rate of useful symbol is decreased. Instead of separate preamble we use part of CP for time synchronization. In MIMO-OFDM cyclic prefix is used to remove inter symbol interference. The cyclic prefix is nothing but tail symbols of massage frame and it prepended to that massage frame. Length of cyclic prefix depends on length of FFT, in given system model 256 length FFT is used so length of CP is given as L = (FFT Length)/4 = 64. Out of 64 first 32 symbols of OVSF sequence are inserted and remaining 32 are tail symbols of massage frame i.e. actual CP as shown in Figure.2. by Et.al. Geo and Zhang [6]. During simulation threshold detection set to 80%. Figure.4 shows the correlation peak at receiver side for OVSF sequence with threshold detection set to 80% and SNR = -5dB. The acquisition probability for time synchronization for proposed method i.e. OVSF sequence and conventional Hadamard sequence, CAZAC sequence are shown in Figure.5 and it is observed that Hadamard and CAZAC sequence have less acquisition probability than proposed sequence when threshold detection set to 80%. Figure.5 shows for SNR -5dB that the acquisition probability is 1 for proposed sequence. Addition of 8 symbol CP before OVSF sequence gives good correlation peak and keeps their orthogonally at the receiver side. The correlation function R Yi,Zi of the received signal Y i and locally generated sequence Z i at received antenna N r is (11) Where L is the index of subcarrier. The correlation algorithm block and threshold detection block are placed at each receiver N r as shown in Figure.3. The correlation block correlates the received signal Y i with local sequence Z i and sends output to the threshold detection block to detect the correlation peak beyond an expected threshold. Fig. 4 : Correlation peak value of OVSF sequence Fig. 3: Time synchronization block Once the correlation peak reaches to the expected threshold that index is the time synchronization point. V. SIMULATION The proposed system is simulated according to Wireless MAN standards as mentioned in section-ii. The acquisition probability method used is mentioned Fig. 5: Acquisition probability for time synchronization for OVSF sequence, Hadamard sequence and CAZAC sequence 84

Fig. 6: BER performance without and with proposed method We have plotted BER vs. E b /N o for two scenario with and without proposed method as shown in figure.6 and it is clear that with proposed method BER is improved. VI. CONCLUSION The proposed method for time synchronization improves the system performance. The proposed orthogonal sequence has very good acquisition probability for time synchronization at SNR -5dB compared to conventional orthogonal sequences. As the time synchronization is very good which results in an improved BER performance. We can implement the proposed method for time synchronization for 4X4 and 8X8 MIMO-OFDM system as a future work. VII. REFERENCES [1] Lin Chin-Hsien, Lin Yi-Hsien Wu Chih-Feng, Wang Chorng-Kuang Kuang, A MIMO-OFDM digital baseband receiver design with adaptive equalization technique for IEEE 802.16 WMAN IEEE International Conference on Acoustics, Speech and Signal Processing, pp. 617 620, April 2009. [2] G. J. Foschini and M. J. Gans, On limits of wireless communications in a fading environment when using multiple antennas Wireless Personal Communications, vol. 6, pp. 311 335, 1998. [3] IEEE Standard for Local and Metropolitan Area Networks Part 16: Air Interface for Broadband Wireless Access Systems, IEEE Std. 802.16-2009, May 2009. [4] V. Tarokh, A. Naguib, N. Seshadri, and A. Calderbank, Space-time codes for high data rate wireless communication: performance criteria in the presence of channel estimation errors, mobility, and multiple paths IEEE Transactions on Communications, vol. 47, no. 2, pp. 199 207, feb 1999. [5] J.-J. van de Beek, M. Sandell, M. Isaksson, and P. Ola Borjesson. Low-complex frame synchronization in ofdm systems Fourth IEEE International Conference on Universal Personal Communications, pp. 982 986, Nov 1995. [6] Z. Gao, J. Xu, and Z. Zhang, A synchronization scheme for mimo ofdm system International Conference on Communications, Circuits and Systems (ICCCAS), pp. 15 18, July 2010. [7] Najjar, Leìla Lab. TECHTRA, Ecole Super. des Commun. Ariana, Tunisia Siala, Mohammed, A new scheme for preamble detection and frequency acquisition in OFDM systems 16th IEEE International Conference on Electronics, Circuits, and Systems, pp. 1008 1011, Dec 2009. [8] Madhukumar, Chen, Yang, Kai, Chin, Francois-P Po Shin, Comparison of signature sequences for synchronization of UWB systems Vehicular Technology Conference, 2004. VTC 2004- Spring, Vol.5, pp. 2585 2589, May 2004. 85