Reduced complexity near-optimal iterative receiver for Wimax full-rate space-time code

Abstract

IEEE802.16e standard proposes a full-rate spacetime block code (STBC) achieving the optimum compromise between diversity and multiplexing gains (scheme C). When applied on modulated and FEC encoded data, under the constraint of practical feasibility, we prove the necessity for an iterative turbolike receiver to outperform the spatial multiplexing structure (scheme B). Exploiting the particular definition of the code, we derive a reduced-complexity minimum mean square error (MMSE) turbo-equalizer and simulations show its efficiency.I. INTRODUCTION IEEE802.16e standard proposes a diversity mode, exploiting multiple antennas at the transmitter side (multiple input multiple output (MIMO) systems) to deliver high quality and/or high data rate transmissions. In the two transmit antenna case, three schemes are defined, among which the well-known Alamouti code [1] and the spatial multiplexing structure. While the first one fully exploits the available spatial diversity yielding a robust receiver and high quality transmissions, the second one aims at increasing the data rates without requiring higher order modulations nor expanded bandwidth. A third scheme is derived from the golden code [2] and optimizes the diversity-multiplexing gain trade-off. Compared to the spatial multiplexing structure, the data rate is the same but the time duration of the code is doubled. The price to pay for this additional diversity enabling higher quality transmissions is thus a time-delay in the data restitution as well as a complexity increase at both sides of the transmission link.In the IEEE802.16e standard, both single and multi-carrier modulations are available. Provided the OFDM frame is welldefined, flat fading assumption on each subchannel is true. We thus consider a non frequency selective MIMO channel, time-invariant over the full-rate STBC duration (i.e two symbol periods). In that case, classical STBC detection is well adapted. Maximum a posteriori (MAP) detection can be used provided the modulation order keeps low (BPSK, QPSK). For higher modulation orders, sphere decoders meet a good complexity/performance trade-off, as long as no soft output is required. In that case, sphere decoders generating a most probable candidate list can be used to provide soft output [3]. Their major drawback is the non bounded latency. In most cases, a forward error correcting (FEC) code is applied on the binary information sequence. In IEEE802.16e, different codes

Association and Joint Optimization of max-dmin Precoder with MIMO Turbo Equalization

“…The modern turbo equalization takes into account the iterative information exchanges not only between the MMSE IC and the FEC decoder, but also between the FEC decoder and the symbol-to-binary converter. Recent researches show that, by applying the intersymbol interference cancellation criterion, the turbo equalization significantly improves the system performance [5], [6].…”

Adaptive complexity MIMO turbo receiver applying turbo demodulation

“…We denote this mode as (TEq). On the other hand authors, in [5] [6] have used a MIMO turbo receiver with feedback to the demapper, we denote this mode as (TEq-TDem). In this latter, the demapper takes into account the a priori information coming from the decoder to improve its output.…”