by Gary Croke
On July 10th, I participated in a panel at IEEE 5G World Forum with colleagues from Vodafone, Bell Labs/Nokia, Facebook and Collinear.
FORUM: 5G vRAN Over Less than Ideal Backhaul
Through a lively discussion and exchange of ideas we explored backhaul/fronthaul considerations in a virtualized RAN in 5G network. We discussed CRAN, vRAN and their benefits, challenges and issues. We also delved into the proposed new fronthaul interface, functional splits, and tradeoffs leading into a wide range of topics including how to leverage installed base of LTE transport for vRAN and 5G, dealing with QoS challenges including delay, jitter and transient micro-bursts, network slicing, and industry changes with respect to transport layer.
Below I summarize my personal top takeaways from this week’s discussion regarding fronthaul for 5G and vRAN:
1. Today’s CRAN Not Viable for 5G
There was agreement (no surprise) that despite its benefits for lowering costs, and enabling features like coordinated multi-point (COMP) and massive MIMO, CRAN is not a viable architecture for 5G – primarily because it does little to address RAN “vendor lock-in” issues and extraordinary capacity requirements of the CPRI-fronthaul interface becomes unworkable with 5G – leaving costly dark fiber as the operators’ only transport option. In addition to these things, lack of performance monitoring and OAM capabilities makes today’s fronthaul interface a challenge to manage in practical sense.
2. Open RAN Has Momentum, Open Fronthaul Interface is Key
Consensus from the team was that that open RAN (as a superset of vRAN) has strong industry momentum – driven mostly by operators and a desire to lower costs. Open RAN requires open and standard fronthaul interface that solves some of the issues with today’s CPRI-based fronthaul which is why this discussion on fronthaul is critical to the larger industry direction. It should be said that this topic sparked a lively debate on the merits of an open RAN – whether or not this will actually enable or stifle innovation. Proponents advocated potential for lower costs and increased opportunity for more players and innovation. Detractors pointed out the standard fronthaul interface has the possibility to actually prevent or slow advances. A debate for the ages I guess… and we left it as an open question. We’ll just have to see how this plays out.
3. Progress on Open Fronthaul Interface Looks Promising
Recent progress on standard fronthaul interface looks promising. Functional splits provide viable options to enable significant reduction in fronthaul capacity requirements (10-12 times) over traditional CPRI. In addition, today’s CPRI interface lacks performance monitoring and OAM capabilities, leaving fronthaul networks unmonitored and difficult to troubleshoot. The progress on standards such as 802.1cm time sensitive Ethernet, ieee 1914.1/3 packet based fronthaul standards which will help in this area is also encouraging. That being said, the standards continue to evolve and we should expect more change to the functional split definitions for the new fronthaul interfaces in the months and years to come
4. Wireless Looks to be an Entirely Viable Option for vRAN Transport
Based on capacity and latency requirements of the new fronthaul interfaces and some of the testing results described by panelists, non-ideal backhaul solutions like wireless technologies look promising as 5G fronthaul technology (we have wireless transport solutions available today at 10Gbps capacity and 10-40us latency). Street level deployments are still a major challenge and it’s essential that we as an industry continue to innovate on the radio and antenna technology to deliver practical solutions that overcome some of the space, power, mounting, and regulatory challenges that come with street level deployments. antenna regulations need to be more flexible to allow small, integrated antennas to be used in current and future bands that allow mass deployment at street level. The industry is still hampered by antenna regulations drawn up decades ago, particularly for E-Band.
Beyond this, given the challenges with designing, deploying, and operating backhaul networks and the expected increase in numbers of cell sites with 5G, it’s imperative that the transport network drive more operational simplicity through advanced automation capabilities and tying these capabilities into the larger operator automation strategies.
With the exceedingly high capacity requirements of CRAN/CPRI, wireless backhaul is generally not a viable option, leaving operators with limited backhaul possibilities – which means more cost. vRAN is promising because it enables operators to utilize wireless backhaul solutions which are necessary for building low cost backhaul networks.
Many of these takeaways confirm our recent analysis on vRAN implications on 5G fronthaul/backhaul which you can see here:
- BLOG: V-RAN Technology: A Promising Alternative for C-RAN with More Flexible Backhaul Options
- BLOG: A Key to More Cost-effective Backhaul and Fronthaul Networks – The Case for vRAN
The industry activity on open RAN and vRAN along with associated open, standard fronthaul interfaces is tremendous and promises to provide operators with lower cost, more flexible network options. Let’s hope the progress continues