A Predictive Semi-Persistent Scheduling Scheme for Low-Latency Applications in LTE and NR Networks

Abstract

The Long Term Evolution (LTE) technology is ubiquitously implemented today and it achieves noticeable success in fulfilling the data transmission demand instigated by the unprecedented growth of mobile services. However, the LTE might fail to provide qualified services for Machine Type Communications (MTC) and the emerging Tactile Internet, as both require data being transmitted with extremely low latencies (0.5∼5 ms). In this paper, we propose a novel predictive Semi-Persistent Scheduling (SPS) scheme with the corresponding SPS scheduler that takes advantage of transmission history, to effectively reduce the uplink latency of LTE systems. Moreover, the proposed SPS design is compatible with new radio (NR) technology developed by Third Generation Partnership Project (3GPP), since both systems share a similar medium access control (MAC) layer skeleton. Results from extensive simulations show that the proposed SPS scheme outperforms the conventional dynamic scheduling scheme and existing SPS schemes in terms of latency and scheduling accuracy. In particular, with a short SPS periodicity, the proposed SPS scheme achieves a sub-4 ms uplink latency under various traffic loads.