Conservative Link Adaptation for Ultra Reliable Low Latency Communications

TL;DR

URLLC in 5G imposes strict latency ($<$ $1-10$ ms) and reliability ($\text{PLR} \approx 10^{-5}$) requirements. The paper introduces a conservative link adaptation that estimates worst-case channel degradation over the CQI-to-transmission interval using ΔCQI(Δt)= $\max\left[ CQI(t'-\Delta t) - CQI(t') \right]$ for $t' \in (t-W, t)$ and updates CQI as $CQI(t_{SCH}) = \max(0, CQI(t_{last\_CQI}) - \Delta CQI(\Delta t))$, combined with subband CQI estimation and a resource-aware MCS/RB allocation strategy. This method aims to preserve reliability while reducing resource usage, by selecting robust MCS when degradation is worst-case anticipated. NS-3 simulations show the approach meets URLLC PLR targets and achieves up to ~6x reductions in RB usage compared with always selecting MCS0, outperforming last-CQI baselines in highly variable channels. The proposed approach thus offers robust URLLC performance with efficient resource utilization across varying mobility conditions.

Abstract

Ultra reliable low latency communications (URLLC) is one of the most promising and demanding services in 5G systems. This service requires very low latency of less than $1-10$ ms and very high transmission reliability: the acceptable packet loss ratio is about $10^{-5}$. To satisfy such strict requirements, many issues shall be solved. This paper focuses on the link adaptation problem, i.e., the selection of a modulation and coding scheme (MCS) for transmission based on the received channel quality indicator (CQI) reports. On the one hand, link adaptation should select a robust MCS to provide high reliability. On the other hand, it should select the highest possible MCS to reduce channel resource consumption. The paper shows that even for one URLLC user, link adaptation is still a challenging problem, especially in highly-variant channels. To solve this problem, a conservative link adaptation algorithm is designed. The algorithm estimates the strongest channel degradation at the time moment of the actual packet transmission and selects an MCS taking into account the worst degradation. The obtained results show that the proposed algorithm is efficient in terms of both the packet loss ratio and the channel resource consumption.

Figures (5)

• Figure 1: Rayleigh fading.
• Figure 2: CQI estimation.
• Figure 3: Influence of WND for vehicle UE (60 kmph).
• Figure 4: Influence of CQI period for pedestrian UE (3 kmph).
• Figure 5: Influence of CQI period for vehicle UE (60 kmph).