Unified Capacity Limit of Non-coherent Wideband Fading Channels

TL;DR

This paper unifies the understanding of capacity limits in non-coherent wideband fading channels by introducing the concept of bandwidth occupancy, showing that different signaling schemes' capacities converge when considering this measure.

Contribution

It introduces bandwidth occupancy as a key metric, unifies capacity analysis for peaky and non-peaky schemes, and provides closed-form expressions for optimal occupancy.

Findings

Achievable rates depend on bandwidth occupancy, not just signaling type.

Capacity approaches wideband AWGN capacity as occupancy nears a critical value.

Trade-off identified between approximation accuracy and capacity bounds.

Abstract

In non-coherent wideband fading channels where energy rather than spectrum is the limiting resource, peaky and non-peaky signaling schemes have long been considered species apart, as the first approaches asymptotically the capacity of a wideband AWGN channel with the same average SNR, whereas the second reaches a peak rate at some finite critical bandwidth and then falls to zero as bandwidth grows to infinity. In this paper it is shown that this distinction is in fact an artifact of the limited attention paid in the past to the product between the bandwidth and the fraction of time it is in use. This fundamental quantity, called bandwidth occupancy, measures average bandwidth usage over time. For all signaling schemes with the same bandwidth occupancy, achievable rates approach to the wideband AWGN capacity within the same gap as the bandwidth occupancy approaches its critical value, and decrease to zero as the occupancy goes to infinity. This unified analysis produces quantitative closed-form expressions for the ideal bandwidth occupancy, recovers the existing capacity results for (non-)peaky signaling schemes, and unveils a trade-off between the accuracy of approximating capacity with a generalized Taylor polynomial and the accuracy with which the optimal bandwidth occupancy can be bounded.