Coherent wave propagation in disordered media gives rise to many fascinating phenomena as diverse as universal conductance fluctuations in mesoscopic metals and speckle patterns in light scattering. Here, the theory of electromagnetic wave propagation in diffusive media is combined with information theory to show how interference affects the information transmission rate between antenna arrays. Nontrivial dependencies of the information capacity on the nature of the antenna arrays are found, such as the dimensionality of the arrays and their direction with respect to the local scattering medium. This approach provides a physical picture for understanding the importance of scattering in the transfer of information through wireless communications.The ongoing communications revolution has motivated researchers to look for novel ways to transmit information (1, 2). One recent development (3, 4) is the suggestion that, contrary to long-held beliefs, random scattering of microwave or radio signals may enhance the amount of information that can be transmitted on a particular channel. Prompted by this suggestion, we introduce a realistic physical model for a scattering environment and analytically evaluate the amount of information that can be transmitted between two antenna arrays for a number of example cases. On the one hand, this lays a new foundation for complex microwave signal modeling, an important task in a world with ever-increasing demand for wireless communication, and, on the other, introduces a new arena for physicists to test ideas concerning disordered media.From information theory (5), the capacity of a channel between a transmitter and a receiver, that is, the maximum rate of information transfer at a given frequency, can be described in terms of the average power of the signal S and the noise N at the receiver: C = log 2 (1+S/N ). More generally (2), the communication channel connecting several transmitters and receivers is described by a matrix G iα giving the amplitude of the received signal α due to transmitter i. The information carried by the channel can be characterized using several quantities, such as the capacity or mutual information, which are typically functionals of the matrix G, which must be known in order to predict these quantities. Often G cannot be predicted for actual systems, such as wireless communication networks or optical fibers, because of the complicated scattering and interference of waves that is involved. It is crucial, therefore, to develop physical models for the signal propagation, as it is only through such models that one can understand the real effects of scattering and interference on the amount of information that can be communicated.In many cases, only partial information is available for prediction; in these situations, one only has a statistical description of G. Instead of making assumptions about G directly, the usual procedure in information theory, we introduce statistical models for the physical environment from which we derive the properties of G. The adv...