New approaches to high-capacity multiple-antenna wireless

Summary form only given. Under Rayleigh flat-fading conditions, it is well-known by now that the capacity of a multiple-antenna wireless link increases linearly with the smaller of the number of transmit or receive antennas, with no increase in bandwidth or transmitted power. This phenomenon was verified experimentally with a scheme called BLAST (Bell Labs Layered Space Time). BLAST requires that the receiver learn the matrix of propagation coefficients. Our current research goal is to realize a high throughput in mobile environments where it may not be feasible to estimate the propagation matrix. A new signalling scheme, unitary space-time modulation, is nearly ideal for this task. We have developed some effective ways to design constellations of these signals. Where a user has extended access to the channel, a variation called differential unitary space-time modulation can be particularly convenient. Moreover space-time signals can function as their own channel codes, with spatial diversity replacing temporal diversity, an effect that we call space-time autocoding. This effect, which may obviate the conventional Shannon requirement to perform channel coding over many independent fades, can benefit differential unitary space-time modulation, as well as schemes where a user has access to the channel for only a short period. For example, within a single coherence interval of 16 symbols, for 7 transmit antennas, 4 receive antennas, and an 18 dB expected SNR, a total of 80 bits (rate 5) can theoretically be transmitted with a block probability of error less than 10/sup -9/, all without any training or knowledge of the propagation matrix.