Analysis of Mid-Haul Characteristics for LTE-NR Multi-Connectivity in Heterogeneous Cloud RAN

Abstract

To address the capacity requirements resulting from huge growth in mobile data traffic, the mobile network operators (MNOs) are deploying heterogeneous Cloud based Radio Access Network (C-RAN) with a mix of Base stations viz. 4G Macro Base stations (offering better coverage) and 5G Small cells (offering better radio capacities). With LTE-NR Multi-Connectivity feature in 5G network, the User Equipments (UEs) can connect with both 4G and 5G simultaneously to take advantage of better coverage and capacity, so that its QoS is met appropriately. To maintain such simultaneous connections (with UE) across different BTS, the network must support stringent high bandwidth and low-latency links (mid-haul links) between the Base station nodes. In a C-RAN environment, the inter Base station mid-haul links can be viewed as inter Virtual Machine (VM) communication link. In this paper, we analyze the characteristics of the mid-haul link for LTE-NR Multi-Connectivity feature (specifically E-UTRAN-NR Dual Connectivity EN-DC configuration) in a heterogeneous C-RAN deployment. We simulate the C-RAN scenario using NS3 network simulator with heterogeneous mix of 4G and 5G-NR Base stations with appropriate cloud architecture split. We use the BTS power model to derive the Computational Requirement (CR) of the Base station for vBBU (virtual Baseband Unit) placement algorithm in C-RAN. Through simulation, we review the bandwidth requirement of the mid-haul link and assess variation in the end-to-end delay when latency of mid-haul link is varied. From available cloud infrastructure hardware bench-mark results, we observe that the latency of inter VM communication links, varies depending on whether the two communicating VMs are on same or different Cloud servers. Thus, we propose “Neighbour Association-aware Placement” (NAP) algorithm for placement of the vBBUs in the same Cloud server and assess the benefits in the case of EN-DC configuration.