Packet fronthaul –design choices towards versatile RAN deployments
The resulting cluster architecture has several advantages over the original large CRAN. It creates smaller stand-alone units that are easier to resize. Clusters can be replicated to build small to large CRANs.
The transport switching platform uses the eCPRI aggregation properties, creating a transport network with fewer connections and allowing flexible connectivity to everyone. The connection between the clusters is reduced as they communicate through the spinal switches in the hub. This leads to a simpler implementation with minimal wiring. Using a resilience principle that combines redundancy with load sharing reduces system size and effectively uses physically deployed features and connectivity.
The architecture works equally well for Cloud-RAN deployment, as it follows the general principles of physical deployment in the cloud and data center. It complements the deployment of 3PP cloud infrastructure, but requires integration into the orchestration system.
Packet management transport network architecture
The RAN application for eCPRI uses primarily L2 Ethernet for its connectivity with stringent performance requirements. This can be done in plain L2 Ethernet fabric, but when the fabric introduces additional pathways, it becomes difficult to manage and control using L2 management protocols. This can happen when a redundancy link is added to the network and can lead to cell dropout or poor performance due to a violation of eCPRI’s specific requirements.
The backbone network architecture makes it possible to separate the physical network structure from the needs of the application services. The IP pad can handle networks from simple to complete networks using protocols that have been used and well understood for years to share network topology and specific information. ECPRI L2 Ethernet traffic is processed as a virtual overlay Ethernet service with strict service requirements, operating through the underlying network structure. This can also be combined with additional virtual Ethernet or IP overlay services with different features. These services can be anything from point to point to multipoint to multipoint depending on the needs of the application.
The pad-to-pad network structure further enhances deployment ease and connection flexibility by being able to detect software-defined overlays with optional traffic paths designed on the transport switch pad, as shown in Figure 6. The Distributed IP-Based Protocol the infrastructure makes the pad independent and gives a quick response in case of failures. The routing protocol extensions allow the dissemination of MPLS forwarding information, including traffic engineering with ISIS / OSPF based segmental routing (SR-MPLS). This also eliminates the need for multiple management planes, but still uses IP and MPLS in the forwarding plane. There are also proposals to remove the MPLS redirection plane and replace it with IPv6 based segmental routing (SR) 4.
Overlay is the virtual network that works on the pad to create connectivity for the L2 EVPN and L3 IP-VPN services. The multithreaded border gateway protocol and the SDN controller are used to simplify the creation and control of virtual networks and services. This includes controlling the traffic engineering for the virtual service in the network by using the capabilities of segmented traffic routing (SR-TE) in the underlying infrastructure.
The topology of the fronthaul network and the flexibility in the 5G NR deployment architecture allow the allocation of functions to different sites for different services and this leads to the need to simplify the deployment and operation of both RAN and related transport. Achieving cost-effective deployment and operation requires an SDN control package and management / orchestration with close interaction between the two.
Packet interface synchronization architecture
When switching from CPRI to eCPRI in the fronthaul, the time allocation between the baseband and the RU is shifted from being transferred from the CPRI frame itself to being transferred from a packet-based precision time protocol (PTP). The ITU-T has specified several PTP profiles with the relevant telecommunications properties, where eCPRI has selected the profile described in G.8275.1 for packaging, . The requirements for timing and frequency stability in the front operator’s network remain the same compared to CPRI-based solutions.
In order for RAN to support full functionality, an effective synchronization solution must be available. For packet switches on the front network, special attention must be paid to maintaining synchronization to or between radio stations and sufficient availability of synchronization. Switches can distribute synchronization and hold functionality located in RAN nodes (mainband, radio) or outside (switches) to mitigate synchronization interruptions from GNSS-derived sources. As a backup, remote GNSS receivers with PTP distribution are optional.