Capacity of distributed PHY-layer sensor networks

Sensor networks are comprised of nodes with minimal baseband and RF functionalities. In such networks, it is assumed that a source sensor communicates with a target sensor over a number of relaying sensors by utilizing distributed low-complexity space-time encoding techniques, hence the resulting communication scenario is a generalized form of orthogonalized multiple-input multiple-output (MIMO) channels. The contributions of this paper are the derivation of the Shannon capacity in terms of natural units per second per Hertz for such space-time encoded distributed communication scenarios. Closed-form capacity expressions are derived for ergodic flat-fading Rayleigh and Nakagami channels, as well as the communication-rate outage probabilities for aforementioned channels. It is shown that the distributed Alamouti scheme yields the best performance over ergodic channels. In the case of nonergodic channels, the 3/4-rate sporadic space-time block code (STBC) is shown to give optimum performance. Finally, Monte Carlo simulations are used to assess the performance of distributed multistage sensor networks. It is shown that notable power savings can be achieved, compared to the traditional single-link sensor networks.