• Technical Conference:  30 March – 03 April 2025
  • Exhibition: 01 – 03 April 2025
  • Moscone Center, San Francisco, California, USA

Mobile fronthaul in the 5G era

By Light Counting

4G fronthaul has converged on 10Gbps optics
In the past several years, LTE networks have been upgraded using additional frequency bands, carrier aggregation, and LTE-A, and small cells have been added within macrocell coverage areas, driving up fronthaul bandwidth requirements to the point where now many operators and equipment suppliers have standardized on 10Gb/s-capable multi-rate transceivers for all their fronthaul needs, since they can meet the majority of different transport speed requirements with one device, while reducing the complexity of specific site designs and spares inventory.  Many operators, especially those leasing their fronthaul fiber, have also deployed WDM systems in their fronthaul networks, as shown below. A typical fronthaul network incorporating WDM is shown below.


5G fronthaul will need faster optic, but how fast? 
Fronthaul requirements will change with the advent of the 5G mobile network. 5G has a target peak bandwidth of 20Gb/s, and to deliver it will utilize higher-frequency spectrum than LTE, in the millimeter wavelength range. The shorter wavelengths enable smaller antennas, which in turn permits the use of much higher-order MIMO antenna arrays. Whereas 4x8 and 8x8 MIMO are considered high-end for LTE, in 5G 64x64 MIMO is possible. And the higher the MIMO number, the greater the bandwidth needed on the corresponding fronthaul link. A second way bandwidth will be increased in 5G is via the use of 100 GHz frequency bands, vs. 20 Ghz in LTE, enabling a single radio to generate 5X more bandwidth that needs to be carried back to the core network from the cell site, everything else being equal.
In anticipation of the potential huge increase in fronthaul bandwidth that would be needed to support 5G radios, mobile equipment makers updated the CPRI specification to something called ‘eCPRI’, (published in August, 2017).  One of the key elements of eCPRI is to move some of the PHY-layer signal processing from the baseband unit to the remote radio head, which reduces fronthaul bandwidth by a factor of 10, in many cases.
When all the different factors affecting fronthaul bandwidth in 5G are added up (some driving it up, others reducing it), the expected bandwidth fall into the range of 14 to 30 Gb/s, depending on specifics of the eCPRI implementation, cell site, etc. In contrast, similar 5G network configuration using the older CPRI scheme which puts all the PHY layer processing in the baseband unit would require 236 Gb/s fronthaul bandwidth. The upshot is that while a 5G cell site would nominally generate 160 Gb/s or more of traffic, the real fronthaul bandwidth needed will be 14-30 Gbps, due to the use of eCPRI.
Similar to how 10Gb/s multi-rate optical modules have become the de facto standard for LTE fronthaul, we believe the next higher standard Ethernet speed will be used in 5G fronthaul. This means we expect to see 25GbE devices used in high volumes in 5G deployments, although some will be Industrial Temperature (I-Temp) and/or bi-directional versions made specifically for fronthaul applications.  

Higher-speed optics (above 25Gb/s) will also be needed in 5G networks
Over the past 12-18 months, LightCounting has heard repeatedly about mobile operators requesting optical transceivers with speeds of 50, 100, and even 400 Gb/s for ‘fronthaul’ or ‘backhaul’. Why such high speeds would be needed has been a real puzzle to figure out. ‘Fronthaul’ has been well-defined by the CPRI consortium for quite some time, but there is no industry consensus definition of wireless backhaul. LightCounting has narrowly defined backhaul as the first optical link originating at the BBU and carrying traffic towards/from the network core. Others expand the definition to include the access, aggregation, and core networks as well. Naturally as 25Gb/s data streams are aggregated going from BBU to the core, 50, 100, and even 400Gb/s transport might be needs. We think this greatly expanded definition of ‘backhaul’ is a possible explanation for the mystery reports of higher-speed backhaul optics. 

Stay Tuned
LightCounting will publish an updated summary forecast for wireless modules in its October 2017 Market Forecast Update, and will publish a complete detailed fronthaul and backhaul forecast as part of its report titled “Next Generation Optical Access” which will be published in November 2017. LightCounting clients can view more detailed and complete research note on 5G fronthaul here.

*Direct from the source, LIghtCounting Market Research

Posted: 5 October 2017 by Light Counting | with 0 comments

Blog post currently doesn't have any comments.
 Security code

The views expressed in this blog are those of the authors and do not necessarily reflect the views or policies of The Optical Fiber Communication Conference and Exposition (OFC)  or its sponsors.