Distributed QAM-Based Space-Time Block Codes for Efficient Cooperative Multiple-Access Communication

摘要

Distributed space-time block coding schemes based on quadrature amplitude modulation (QAM) symbols are introduced that enable users in a wireless multiple-access relay channel to cooperate with each other in order to improve the reliability of their information in a fair, rate-efficient manner. The distributed coding schemes are designed for a system that consists of m users that need to send their information to a common destination. Each user is equipped with only one antenna and a half-duplex transceiver so that it can transmit to the destination or any other user at one time and also receive signals from any other user at another time. The destination is equipped with one or more receive antennas. The goal is to design cooperative schemes which achieve the maximum diversity order (which is equal to the diversity order of the outage probability) while being fair to the users in terms of rate allocation. Two cooperative coding schemes are proposed that meet these requirements-a two-phase cooperative scheme and a single-phase self-information canceling linear (SCL) scheme. The two-phase scheme is of lower decoding complexity and it requires the users to only transmit QAM symbols. The SCL scheme is of higher complexity with the users transmitting linear combination of QAM symbols but it also achieves better performance. Both schemes incorporate new classes of cooperation rules for deciding whether or not a user acts as a relay for another user. The usual "outage-based" cooperation rule, whereby a user cooperates with another user provided the mutual information of the channel between them is greater than the rate, is sufficient for deriving information-theoretic limits but it cannot be used directly for analysis of a particular coding scheme. Even though the decoder used by the destination in the proposed coding schemes assumes that the inter-user communication is always successful, our performance analysis does not make this assumption. In fact, it rigorously accou--nts for the decoding errors arising from the information exchange between users. Consequently, it sheds light on precisely what cooperation rules (among the class of rules analyzed) lead to maximal diversity.