Fronthaul data compression for Uplink CoMP in cloud radio access network (C-RAN)

Yinan Qi, Muhammad Zeeshan Shakir, Muhammad Ali Imran, Khalid A. Qaraqe, Atta Quddus, Rahim Tafazolli

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

The design of efficient wireless fronthaul connections for future heterogeneous networks incorporating emerging paradigms such as cloud radio access network has become a challenging task that requires the most effective utilisation of fronthaul network resources. In this paper, we propose to use distributed compression to reduce the fronthaul traffic in uplink Coordinated Multi-Point for cloud radio access network. Unlike the conventional approach where each coordinating point quantises and forwards its own observation to the processing centre, these observations are compressed before forwarding. At the processing centre, the decompression of the observations and the decoding of the user message are conducted in a successive manner. The essence of this approach is the optimisation of the distributed compression using an iterative algorithm to achieve maximal user rate with a given fronthaul rate. In other words, for a target user rate the generated fronthaul traffic is minimised. Moreover, joint decompression and decoding is studied and an iterative optimisation algorithm is devised accordingly. Finally, the analysis is extended to multi-user case and our results reveal that, in both dense and ultra-dense urban deployment scenarios, the usage of distributed compression can efficiently reduce the required fronthaul rate and a further reduction is obtained with joint operation.
Original languageEnglish
Pages (from-to)1409-1425
JournalTransactions on Emerging Telecommunications Technologies
Volume27
Issue number10
Early online date4 Aug 2016
DOIs
Publication statusPublished - Oct 2016
Externally publishedYes

Fingerprint Dive into the research topics of 'Fronthaul data compression for Uplink CoMP in cloud radio access network (C-RAN)'. Together they form a unique fingerprint.

  • Cite this