Resource Allocation Policies For Short-packet Communication With Imperfect CSI

The optimization of minimum power allocation strategy in communication system to meet the user service quality requirements is studied in this paper.In the scenarios of ultra-reliable and low-latency communications(URLLC),assuming that the channel state information(CSI)is not known at the receiver and transmitter,the point-to-point communication with a single-antenna under the Rayleigh block fading channel model is studied.The finite blocklength(FBL)coding rate is adopted to ensure the delay constraint,and the Hybrid Automatic Repeat request with Incremental Redundancy(IR-HARQ)with one retransmission is employed to improve the transmission reliability.Effective capacity is employed to characterize the throughput with renewal service processes under statistical delay constraints.First,the power allocation policy is considered under the noncoherent transmission,in which we establish the minimum transmission power framework under the constraint of the user's communication quality of service requirements and propose the corresponding power allocation algorithm.By comparing the retransmission scheme with single transmission,numerical results show that the retransmission scheme brings higher energy efficiency.Secondly,due to the unknown CSI,the power allocation strategy is also considered under the pilot signal assisted transmission.Similarly,under the constraints of URLLC and minimum throughput,the power allocation algorithm is proposed and compared with the noncoherent transmission scheme.The results show that when the coherence interval is small,the noncoherent transmission scheme achieves better performance.With the increase of the coherence interval,the pilot signal assisted transmission has higher energy efficiency.