Rateless Coded Multi-User Uplink Transmission With Distributed Fronthaul Compression in Cloud RAN

Yu Zhang,Zhehao Fan,Limin Meng

doi:10.1109/access.2020.3030132

Abstract

In this paper, we design the rateless coded uplink transmission in cloud radio access network (C-RAN) with two users, two remote radio heads (RRHs), and one baseband processing unit (BBU) pool under block fading channel. Each user sends the Raptor coded signals continuously until receiving ACK from the BBU pool. A distributed fronthaul compression scheme with LDPC code is proposed for the RRHs which can reduce the compression loss without increasing the fronthaul traffic, through leveraging the correlation between the received signals at both RRHs. To guarantee the compression performance, the LDPC code profile is optimized based on extrinsic information transfer (EXIT) analysis. Furthermore, the Raptor code profiles applied at the two users are also jointly optimized to improve the average system throughput. Simulation results show that the proposed transmission scheme with the optimized LDPC compression code and Raptor code can achieve good BER and throughput performance.

Highlights

In the future communication networks, massive users and traffic demands under limited available communication resource impose a stringent requirement on the spectrum efficiency and energy efficiency
We investigate the extrinsic information (EI) passing on the decoding graph during the decoding iteration, where EI is defined as mutual information of the message bit and its corresponding LLR, and derive the EI update rules
In this work, we designed the Raptor coded transmission for two-user uplink in cloud radio access network (C-RAN) with two remote radio heads (RRHs) under block fading channel, including the distributed fronthaul compression with LDPC code at RRHs and joint decompression and decoding at the BBU pool

Summary

Introduction

In the future communication networks, massive users and traffic demands under limited available communication resource impose a stringent requirement on the spectrum efficiency and energy efficiency. In C-RAN, all the BBUs are backward centralized into a cloud computing resource pool and each RRH is located closer to the users. BBU pool and RRHs are connected through high-speed fronthaul links. The main advantage of C-RAN lies in that the coordinated multiple points transmission (CoMP) can be realized inherently which can greatly improve the system spectrum efficiency [1], and the Capital Expenditure (CAPEX) and Operating Expense (OPEX) are reduced compared with the conventional RAN [2]. In C-RAN, the RRHs and BBU pool communicate through the fronthaul with restricted capacity. The signals conveyed through the fronthaul should be compressed first to meet the capacity limit.

Objectives

Results

Conclusion