ENHANCED ZP-OFDM RECEIVER IN MULTI-RELAY COOPERATIVE NETWORKS ∗ Homa Eghbali, ∗∗ Sami Muhaidat, and ∗∗ Ibrahim Abualhaol ∗ School of Engineering Science, Simon Fraser University, Burnaby, Canada, V5A1S6. ∗∗ Khalifa University of Science, Technology, and Research, United Arab Emirates, Emails: {homa.eghbali, muhaidat, ibrahimee}@ieee.org. Abstract—sub-optimal Zero-Padding Orthogonal Fre- quency Division Multiplexing (ZP-OFDM) receivers have been developed in the literature to trade-off performance with complexity. In this paper, we propose a novel coop- erative low-complexity minimum mean square error (LC MMSE)-ZP-OFDM receiver that is able to bring complex- ity reduction in the receiver design, while outperforming conventional cooperative MMSE-ZP-OFDM. I. I NTRODUCTION Zero-Padding (ZP) of multicarrier transmission has been proposed to ensure symbol recovery with- out the need for error control techniques. Re- cently, there have been intensive research efforts on Orthogonal Frequency Division Multiplexing (OFDM)-based multihop relaying [1], [2]. In [3] the information rate of OFDM and OFDMA networks consisting of one source/destination pair and mul- tiple relays is examined.In [4], Ding and Uysal investigate the performance of relay selection in a precoded?? cooperative OFDM system with the AF protocol. In this letter, we propose a novel iterative detec- tion scheme for multi-relay cooperative ZP-OFDM transmissions with the amplify-and-forward (AF) relaying. Not only the proposed receiver outper- forms cooperative MMSE-ZP-OFDM, but also uses low-complexity computational methods by avoiding inversion of channel dependent matrices. Notation: − (.), (.) T , (.) † and(.) H denote conjugate, transpose, pseudo inverse and Hermitian transpose operations, respectively. ∗ denotes linear convolu- tion, |.| denotes the absolute value, and ‖.‖ denotes the Euclidean norm of a vector. [.] k,l denotes the (k,l)th entry of a matrix, [.] k denotes the kth entry of a vector, I M denotes the identity matrix of size M , and 0 M×M denotes all-zero matrix of size M × M . Q = Q −1 M = Q H M represents the M ×M Inverse Fast Fourier Transform (IFFT) matrix whose (l,k) ele- ment is given by Q(l, k)=1/ √ M exp(j 2πlk/M ) where 0 ≤ l, k ≤ M − 1. Bold upper-case letters denote matrices and bold lower-case letters denote vectors. II. S YSTEM MODEL We consider a multiple-relay assisted coopera- tive wireless communication system with a single source (S ), N R half-duplex relay terminals (R i ), i =1, 2, ..., N R , and a single destination (D). The source, destination, and all relays are equipped with single transmit and receive antennas. Any linear modulation technique such as QAM or PSK can be used. We assume the AF relaying and adopt the user cooperation protocol proposed by Nabar et al. [5]. Specifically, in the broadcasting phase, the source node transmits to the relay nodes and the destination node. In the relaying phase, the relay nodes forward a scaled noisy version of the received signal to the destination node. The channel impulse responses (CIRs) for S → R i , S → D, and R i → D links for the i th relay terminal and j th transmission block are given by h j SR i =  h j SR i [0], ..., h j SR [L SR i ]  T , h j SD =  h j SD [0], ..., h j SD [L SD ]  T and h j R i D =  h j R i D [0], ..., h j R i D [L R i D ]  T , respectively, where L SR i , L SD , and L R i D denote the corresponding channel memory lengths. All S → R i , S → D, and R i → D links are assumed to experience frequency selective Rayleigh fading. The random vectors h SR i , h SD , and h R i D are assumed to be independent zero-mean complex 2012 25th IEEE Canadian Conference on Electrical and Computer Engineering (CCECE) 978-1-4673-1433-6/12/$31.00 ©2012 IEEE