On a Class of Time-Varying Gaussian ISI Channels

TL;DR

This paper investigates the capacity limits of a class of stochastic, time-varying Gaussian ISI channels with uncertain tap distributions, providing bounds and conditions under which capacity saturates or fails to scale with power.

Contribution

It introduces a lower bound on channel capacity for time-varying Gaussian ISI channels with unknown joint distributions, and offers a partial converse result under worst-case variation scenarios.

Findings

Capacity saturates at high power levels depending on tap radii

Lower bounds on capacity are derived for the class of channels studied

Rate scaling is limited under certain codebook constraints

Abstract

This paper studies a class of stochastic and time-varying Gaussian intersymbol interference~(ISI) channels. The probability law for the~$i^{th}$ channel tap during time slot~$t$ is supported over an interval of centre $c_i$ and radius~$ r_{i}$. The transmitter and the receiver only know the centres $c_i$ and the radii $r_i$. The joint distribution for the array of channel taps and their realizations are unknown to both the transmitter and the receiver. A lower bound (achievability result) is presented on the channel capacity which results in an upper bound on the capacity loss compared to when all radii are zeros. The lower bound on the channel capacity saturates at a positive value as the maximum average input power $P$ increases beyond what is referred to as the saturation power $P_{sat}$. Roughly speaking, $P_{sat}$ is inversely proportional to the sum of the squares of the radii $r_i$. A partial converse result is provided in the worst-case scenario where the array of channel taps varies independently along both indices $t$ and $i$ with uniform marginals. It is shown that for every sequence of codebooks with vanishing probability of error, if the size of each symbol in every codeword is bounded away from zero by an amount proportional to $\sqrt{P}$, then the rate of that sequence of codebooks does not scale with~$P$. Tools in matrix analysis such as matrix norms and Weyl's inequality on perturbation of eigenvalues of symmetric matrices are used in order to analyze the probability of error.