Bounding Queue Length Violation Probability of Joint Channel and Buffer Aware Transmission

TL;DR

This paper analyzes the tail distribution of queue length violation probability in wireless communications, revealing conditions under which it can be zero or decay at different rates, with implications for QoS in TSN and URLLC.

Contribution

It provides a cross-layer analysis of queue length tail distributions under power constraints, extending understanding to Rayleigh and Nakagami-m fading channels.

Findings

Queue violation probability can be zero with diversity gains under power constraints.

The tail distribution can have a linear decay rate in limited receiver sensitivity scenarios.

The analysis applies to Rayleigh and Nakagami-m fading channels, validated by numerical results.

Abstract

Queue length violation probability, i.e., the tail distribution of the queue length, is a widely used statistical quality-of-service (QoS) metric in wireless communications. Characterizing and optimizing the queue length violation probability have great significance in time sensitive networking (TSN) and ultra reliable and low-latency communications (URLLC). However, it still remains an open problem. In this paper, we put our focus on the analysis of the tail distribution of the queue length from the perspective of cross-layer design in wireless link transmission. We find that, under the finite average power consumption constraint, the queue length violation probability can achieve zero with diversity gains, while it can have a linear-decay-rate exponent according to large deviation theory (LDT) with limited receiver sensitivity. Besides, we find that the arbitrary-decay-rate queue length tail distribution with the finite average power consumption exists in the Rayleigh fading channel. Then, we generalize the sufficient conditions for the communication system belonging to these three scenarios, respectively. Moreover, we apply the above results to analyze the wireless link transmission in the Nakagami-m fading channel. Numerical results with approximation validate our analysis.