Communication over Wireless Fading Channels

The most fundamental and unique characteristic of wireless channels is the random time-variation of the channel strengths. Communication over fading channels has been a topic of study since the '60s. A very different view of the problem, however, emerges from recent research. The traditional view of fading is that it is a source of unreliability that has to be compensated for by various diversity and power control techniques. The modern view considers fading as a source of randomization that can be exploited to get significant capacity boost, even beyond that of a non-faded channel. Two prime examples are:
Opportunistic Communication. Dynamic rate and power allocation can be performed over the dimensions of time, frequency, antennas, and users in a wireless system. In a fading environment, the channel will be strong sometime, somewhere, and opportunistic schemes can choose to transmit in only those channel states.

Multi-antenna Communication. In systems with multiple transmit and multiple receive antennas, random fading increases the number of degrees of freedom available for communication by ensuring that the channel matrix is well conditioned. The phenomenon is also called spatial multiplexing.

We propose to look at two sets of problems in the context of this modern view of fading: Channel Uncertainty in Opportunistic Communication. The fundamental bottleneck limiting performance of opportunistic communication schemes is the channel uncertainty at the transmitter due to variations in the channel and delayed feedback. The main question is how best to perform opportunistic communication in face of such channel uncertainty. Building on the experience of implementing these ideas in Qualcomm's HDR system, some specific problems we propose are: Analysis of the capacity of multi-user fading channels with noisy delayed feedback

> The problem of optimal channel probing for the strong channel states using a limited amount of power

> Performance scaling of opportunistic schemes in wideband systems with many users

> TDD (time-division duplex) versus FDD(frequency-division duplex) systems in the role of reducing noise and delay in the feedback
Diversity versus Spatial Multiplexing in Multi-Antenna Systems. While fading provides the potential for spatial multiplexing gain, many of the existing coding schemes are designed instead to maximize the diversity advantage, a traditional notion framework.We put forth the viewpoint that there is a fundamental tradeoff between diversity
advantage and spatial multiplex gain.We propose to analyze this tradeoff and design schemes that can perform close to this optimal tradeoff over a wide range. One also can view this tradeoff as that between the traditional use and the modern use of multiple antennas.