Dynamics of Interference in Wireless Networks

The temporal and spatial dynamics of multiple-access interference in wireless networks has significant impact on diversity techniques and protocols. The goal of this project is to gain a deep understanding of interference correlation in the time and space domain and to rigorously analyze its impact on system performance. Taking into account various sources of correlation - node locations, channel, and traffic - we aim at deriving expressions for the correlation coefficient of the overall interference power received at a certain point in space. Based on such expressions, we analyze the overall outage rate and throughput of cooperative diversity schemes.

Devices in wireless communication systems may suffer from interference, which is the superposition of a signal with unwanted signals that disturbs a receiver from receiving and decoding the desired signal. For example, if several devices transmit their data packets simultaneously on the same frequency channel to a base station, packet collisions occur at this base station. Interference management is therefore an essential building block in wireless systems. It performs signal processing of interference and avoids or reduces interference by means of multiple access protocols or scheduling. It is important to note that a receiver is generally not exposed to a constant interference power, but this value varies widely over time and the location of the receiver. The reason for this dynamic behavior is, for example, the mobility of devices and obstacles as well as the varying transmission characteristics. This temporal and spatial dynamics of interference was fundamentally investigated in the project Dynamics of Interference in Wireless Networks funded by the Austrian Science Fund (FWF). We developed stochastic interference models and derived expressions for the reception quality. A deep understanding of interference dynamics is essential for many types of interference management. In order to improve the reliability of wireless transmissions, a signal is often transmitted over several antennas (spatial diversity) or several times in succession (temporal diversity). The performance of such diversity techniques depends on the difference of the interference in the different transmission paths or time instances. If the interference values of two channels are highly correlated, it is useless to send the same signal over both channels, since either both transmissions succeed or both fail. In contrast, diversity methods work well if the transmission paths are independent of each other. The project contributed in several ways toward a theory of interference dynamics: First, we characterized interference signals in various stochastic networks in terms of its correlation and joint probabilities of successful transmissions at different time instants or locations. Second, we developed mathematical tools, so-called interference functionals, which can be employed to calculate a wide range of performance measures in wireless networks. Third, we analyzed diversity techniques and showed that interference dynamics has significant impact on their performance. Our follow-up work exploits the attained knowledge for the purpose of interference prediction in wireless networks.