Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

On the Coexistence of eMBB and URLLC in the Cell-Free Massive MIMO Downlink

Giovanni Interdonato, Stefano Buzzi

TL;DR

This paper tackles the challenge of coexisting eMBB and URLLC in the downlink of cell-free massive MIMO by formulating a unified information-theoretic framework that couples infinite-blocklength eMBB performance with finite-blocklength URLLC reliability using a mismatched decoding approach. It introduces four downlink coexistence strategies—superposition coding (SPC) and three puncturing schemes (local, cluster-based, network-wide)—under imperfect CSI and random URLLC activation patterns, and analyzes precoding (LP-MMSE, MR) and power control (weighted fractional power allocation). The results reveal that spatial degrees of freedom at the access points are crucial for mitigating SPC interference, while puncturing is needed to meet strict URLLC reliability, with local puncturing offering the best eMBB-URLLC trade-off in the scenarios studied. The framework extends prior unified approaches to CF-mMIMO and provides design insights for enabling flexible, heterogeneous service coexistence in next-generation RANs, including potential extensions to uplink scenarios and massive machine-type communications.

Abstract

We investigate the non-orthogonal coexistence between the ultra-reliable low-latency communication (URLLC) and the enhanced mobile broadband (eMBB) in the downlink of a cell-free massive multiple-input multiple-output (MIMO) system. We provide a unified information-theoretic framework that combines a finite-blocklength analysis of the URLLC error probability based on the use of mismatched decoding with an infinite-blocklength analysis of the eMBB spectral efficiency. Superposition coding and three levels of puncturing are considered as alternative downlink coexistence strategies to cope with the inter-service interference and the URLLC random activation pattern, under the assumption of imperfect pilot-based channel state information acquisition at the access points and statistical channel knowledge at the users. Numerical results shed light into the trade-off between eMBB and URLLC performances considering different precoding and power control strategies.

On the Coexistence of eMBB and URLLC in the Cell-Free Massive MIMO Downlink

TL;DR

This paper tackles the challenge of coexisting eMBB and URLLC in the downlink of cell-free massive MIMO by formulating a unified information-theoretic framework that couples infinite-blocklength eMBB performance with finite-blocklength URLLC reliability using a mismatched decoding approach. It introduces four downlink coexistence strategies—superposition coding (SPC) and three puncturing schemes (local, cluster-based, network-wide)—under imperfect CSI and random URLLC activation patterns, and analyzes precoding (LP-MMSE, MR) and power control (weighted fractional power allocation). The results reveal that spatial degrees of freedom at the access points are crucial for mitigating SPC interference, while puncturing is needed to meet strict URLLC reliability, with local puncturing offering the best eMBB-URLLC trade-off in the scenarios studied. The framework extends prior unified approaches to CF-mMIMO and provides design insights for enabling flexible, heterogeneous service coexistence in next-generation RANs, including potential extensions to uplink scenarios and massive machine-type communications.

Abstract

We investigate the non-orthogonal coexistence between the ultra-reliable low-latency communication (URLLC) and the enhanced mobile broadband (eMBB) in the downlink of a cell-free massive multiple-input multiple-output (MIMO) system. We provide a unified information-theoretic framework that combines a finite-blocklength analysis of the URLLC error probability based on the use of mismatched decoding with an infinite-blocklength analysis of the eMBB spectral efficiency. Superposition coding and three levels of puncturing are considered as alternative downlink coexistence strategies to cope with the inter-service interference and the URLLC random activation pattern, under the assumption of imperfect pilot-based channel state information acquisition at the access points and statistical channel knowledge at the users. Numerical results shed light into the trade-off between eMBB and URLLC performances considering different precoding and power control strategies.
Paper Structure (7 sections, 14 equations, 3 figures)

This paper contains 7 sections, 14 equations, 3 figures.

Table of Contents

  1. Introduction
  2. System Model
  3. eMBB-URLLC Coexistence Strategies
  4. Precoding and Power Allocation
  5. Performance Analysis
  6. Simulation Results
  7. Conclusion

Figures (3)

  • Figure 1: CDFs of the DL sum SE achieved by LP-MMSE and MR with FPA ($\nu\!=\!0.5$, $\omega\!=\!0.2$) and EPA. $K\!=\!40$, $\alpha\!=\!0.2$, $a_{\mathrm{u}}\!=\!10^{-0.5}$, $T\!=\!5$, $n_{\mathrm{d}}\!=\!114$.
  • Figure 2: DL per-user error probability achieved by LP-MMSE. Settings are those reported in Fig. \ref{['fig:fig1']}. The error probability target is $\epsilon^{\mathsf{dl}}_{\text{{}trg}}\!=\!10^{-5}$.
  • Figure 3: Network availability achieved by LP-MMSE and MR with FPA, $\nu\!=\!0.5$, $\omega\!=\!0.8$. Here, $K\!=\!40$, $M\!=\!16$, $\alpha\!=\!0.2$, $T\!=\!2$, $n_{\mathrm{d}}\!=\!285$.