Flexible Backhauling with Massive MIMO for Ultra-Dense Networks

Petteri Kela*, Mário Costa, Jussi Turkka, Kari Leppänen, Riku Jäntti

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

11 Citations (Scopus)
182 Downloads (Pure)


One of the main challenges for wide-scale deployment and timely adoption of ultra-dense networks (UDNs) in future 5G is the backhaul. Typically, mmW technologies for backhaul require line-of-sight conditions while high-capacity wired based solutions need a significant investment in infrastructure. Such limitations pose practical constrains on the scalability of UDNs and increase the deployment cost of dense networks. In this paper, we consider in-band backhaul for UDNs based on massive MIMO systems in sub-6GHz. In particular, we propose a scheme for allowing simultaneous downlink transmissions in backhaul and access network on a single frequency-band that exploits a novel combination of state-of-the-art practical transmit and receive beamforming techniques. A novel frame structure for allowing a co-existence between massive MIMO based backhaul and UDNs is also proposed. Moreover, a solution for in-band uplink transmissions that exploits time-division-duplex (TDD) and spatial multiple-access is also provided. Extensive numerical results using a realistic system-level simulator are given. Results show that the performance of a UDN with the proposed inband backhaul scheme reaches ~58% of the throughput of a similar access network with ideal (e.g., wired) backhaul. Our results also show that the proposed scheme provides an increase in throughput of ~30% compared to a time division duplex (TDD) scheme for in-band backhaul. Further advantages of the proposed massive MIMO based in-band backhaul scheme for UDNs include reusing both the (scarce) spectrum in sub-6GHz and acquired macro-sites, thus providing a seamless transition from LTE to 5G networks.
Original languageEnglish
Article number7762799
Pages (from-to)9625-9634
Number of pages10
JournalIEEE Access
Early online date1 Dec 2016
Publication statusPublished - 27 Jan 2017
MoE publication typeA1 Journal article-refereed


  • 5G
  • massive MIMO
  • Ultra-dense networks
  • vehicular users

Fingerprint Dive into the research topics of 'Flexible Backhauling with Massive MIMO for Ultra-Dense Networks'. Together they form a unique fingerprint.

Cite this