WP 3. Single-User Optimal Transmit and Receive

This WP focuses on open problems, for single link transmission, which are mostly related to the space-time coding algorithms for single-user MIMO systems. Although most wireless systems support multiple users, the analysis of the single-link case is still meaningful to evaluate the system performance in cases where different users transmit over orthogonal channels. This initial simplification is useful to concentrate on one of the basic problems of multiantenna systems, that is how to find the best trade-off between three fundamental aspects: i) the potential rate increase, due to spatial multiplexing, ii) the improvement in terms of BER, which translates into a reduction of the power strictly necessary to achieve a required out-of-service probability, and iii) the receiver complexity.

We will provide a unified formulation to describe different modulation-multiplexing techniques, as a key step towards easy self-reconfigurability and possibility to merge space-time coding with multiplexing strategies. In particular, this formulation shows that OFDMA, DS-CDMA, and all kind of MC-CDMA schemes can be formulated using a general framework [Sca 99]. In this context each block of transmitted data is related to the block of information symbols by a linear precoder which might incorporate a cyclic prefix (CP).

We will then consider the best options for the linear precoder design, depending on the knowledge of the instantaneous channel and on the channel statistics, for both single antenna and multiple antenna transmitters. Alternative schemes will be studied in detail, searching for the best compromise between performance and receiver complexity.

Finally, possible combinations of linear precoding and forward error correction (FEC) coding schemes will be developed. In particular, we will analyse in detail a concatenation of space-time coding and FEC, with the goal of finding the best trade-off between diversity gain, coding gain, rate and complexity. Analysis will not be restricted to the case where the diversity gain is maximum, as in the majority of the available literature. In fact, in practical scenarios at moderate SNR, it is better to increase the coding gain, even at the price of a lower diversity gain. Therefore, the best compromise between coding and diversity gain, so as to minimise the SNR necessary to guarantee the target BER, will be established.