Reliability analysis of Uplink Grant-Free transmission over Shared Resources

Gilberto Berardinelli, Nurul Huda Mahmood, Renato Barbosa Abreu, Thomas Jacobsen, Klaus I. Pedersen, Istvan Z. Kovacs, Preben Elgaard Mogensen

Research output: Contribution to journalJournal articleResearchpeer-review

37 Citations (Scopus)
268 Downloads (Pure)


Uplink grant-free schemes have the promise of reducing the latency of a user-equipment-initiated transmission by avoiding the handshaking procedure for acquiring a dedicated scheduling grant. However, the possibility of successfully delivering a payload within a latency constraint may be severely compromised in case of grant-free operations over shared radio resources. In this paper, we study the performance of two different uplink grant-free schemes over shared resources recently discussed within the fifth generation new radio standardization, namely, a solution based on a stop-and-wait (SAW) protocol and a blind retransmission approach. Performance is evaluated assuming Rayleigh fading channels with a maximum ratio combining (MRC) multi-antenna receiver. Analytical results show the benefits of grant-free transmission with respect to the traditional grant-based approach for a tight latency constraint. A high-order receive diversity is beneficial to leverage the MRC gain and enables the possibility of achieving the 10 -5 outage probability target set for ultra-reliable low-latency communication services. The blind retransmission approach is significantly penalized by identification and signaling errors, while a SAW solution with potentially scheduled retransmissions out of the shared bandwidth leads to the lowest outage probability, at least for frequent packet arrivals.

Original languageEnglish
JournalIEEE Access
Pages (from-to)23602-23611
Number of pages10
Publication statusPublished - 2018


  • Multiple access
  • grant-free transmission


Dive into the research topics of 'Reliability analysis of Uplink Grant-Free transmission over Shared Resources'. Together they form a unique fingerprint.

Cite this