Normal approximations for fading channels
Tobias Koch (Universidad Carlos III de Madrid)
Capacity and outage capacity characterize the largest transmission rate at which reliable communication is feasible when there are no constraints on the packet length. Evaluated for fading channels, they are important performance benchmarks for wireless communication systems. However, the latency of a communication system is proportional to the length of the packets it exchanges, so assuming that there are no constraints on the packet length may be overly optimistic for communication systems with stringent latency constraints. Recently, there has been great interest within the information theory community in characterizing the largest transmission rate for short packet lengths. Research on this topic is often concerned with asymptotic expansions of the transmission rate with respect to the packet length, which then give rise to normal approximations. In this talk, I present a high-SNR normal approximation for noncoherent, single-antenna, Rayleigh block-fading channels that becomes accurate as the packet length and the signal-to-noise ratio (SNR) tend to infinity. By comparing our approximation to nonasymptotic bounds, I further illustrate its accuracy at finite packet length and SNR values.