Distributed space-time block coding

In this paper, a new class of distributed space-time block codes (DSTBCs) is introduced. These DSTBCs are designed for wireless networks which have a large set of single-antenna relay nodes /spl Nscr/, but at any given time only a small, a priori unknown subset of nodes S/spl sube//spl Nscr/ can be active. In the proposed scheme, the signal transmitted by an active relay node is the product of an information-carrying code matrix and a unique node signature vector of length N/sub c/. It is shown that existing STBCs designed for N/sub c/2 co-located antennas are favorable choices for the code matrix, guaranteeing a diversity order of d=min{N/sub S/,N/sub c/} if N/sub S/ nodes are active. For the most interesting case, N/sub S//spl ges/N/sub c/, the performance loss entailed by the distributed implementation is analytically characterized. Furthermore, efficient methods for the optimization of the set of signature vectors are provided. Depending on the chosen design, the proposed DSTBCs allow for low-complexity coherent, differential, and noncoherent detection, respectively. Possible applications include ad hoc and sensor networks employing decode-and-forward relaying.