Semi-Persistent Scheduling in NR Sidelink Mode 2: MAC Packet Reception Ratio Model and ns-3 Validation

TL;DR

This work tackles the challenge of predicting NR SL mode 2 PRR under Semi-Persistent Scheduling by developing a closed-form MAC collision model that captures SPS specifics such as the reselection counter, sensing windows, and half-duplex constraints. The authors derive a two-event collision framework with $P_{COL}=P_{COL,1}+P_{COL,2}$ and link PRR to MAC collisions and HD errors, then extend the model to account for multiple Layer-2 transmissions per RRI ($N_{Se}$) and a minimum reselection threshold ($X$). They validate the analytical model against ns-3 simulations, showing good agreement in under-saturated regimes and providing insights on when duplicating transmissions helps and whether increasing $X$ benefits PRR. The results offer practical guidelines for NR SL design and SPS parameter tuning, highlighting that PDCP duplication improves reliability mainly in under-saturated networks and that the minimum reselection-proportion $X$ should remain at the lower bound, with future work needed to incorporate PHY effects for saturated scenarios.

Abstract

5G New Radio (NR) Sidelink (SL) has demonstrated the promising capability for infrastructure-less cellular coverage. Understanding the fundamentals of the NR SL channel access mechanism, Semi-Persistent Scheduling (SPS), which is specified by the 3rd Generation Partnership Project (3GPP), is a necessity to enhance the NR SL Packet Reception Ratio (PRR). However, most existing works fail to account for the new SPS features introduced in NR SL, which might be out-of-date for comprehensively describing the NR SL PRR. The existing models ignore the relationships between SPS parameters and, therefore, do not provide sufficient insights into the PRR of SPS. This work proposes a novel SPS PRR model incorporating MAC collisions based on new features in NR SL. We extend our model by loosening several simplifying assumptions made in our initial modeling. The extended models illustrate how the PRR is affected by various SPS parameters. The computed results are validated via simulations using the network simulator (ns-3), which provides important guidelines for future NR SL enhancement work.

Figures (21)

• Figure 1: SPS MAC Procedure.
• Figure 2: SPS Stage 1: Reference UE Case.
• Figure 3: SPS Stage 2: Reference UE Case.
• Figure 4: Sensing Window and Selection Window.
• Figure 5: Collision Event 1.