I’ve heard from several sources the idea that when building out LTE Urban Small Cells, the backhaul design should be adequate to cope with future demands of 5G. For some that means dark fibre, for most it means high cost, delay and even makes it impractical. It can involve heavy machinery, trenching and even road closure for even the smallest of sites. Is this a sensible strategy to take from the outset?
What is 5G anyway?
The requirements, timeframe and scope of 5G still seem pretty vague to me. Whether this will be a single new RF technology that simultaneously meets the demand for low power wide area IoT devices alongside superfast high speed HD video remains to be seen. I think it more likely that LTE will continue to evolve and perhaps encompass an addition radio interface at some point, but the overall architecture and infrastructure will be based on today’s LTE design.
While some 5G solutions maximise spectral efficiency through techniques including massive MIMO (requiring large numbers of parallel antennae), general consensus is that reuse of spectrum through very large numbers of small cells/RF radio heads is the most effective solution for huge capacity and fast speeds.
This links to our view that the longer term small cell may not be high capacity (in terms of numbers of users) but high throughput (in terms of fast data rates). Think more of multiple firehoses than a distributed sprinkler system.
Urban Small Cells need to be flexible, short term investments
It seems to me there is a mindset that building an urban small cell site attracts similar timescales, costs and overheads to a larger macrocell installation. Operators would want a long term view that such investments will have a long lifetime to pay back the investment. Macrocell sites have a lifetime measured in decades, and current thinking seems to be similar for urban small cell investments.
This leads to predicting likely capacity requirements for many years ahead and catering for that growth from the outset. Mobile operators want dark fibre, sometimes multiple strands of it, to each site. This may be viable in a few countries where commercial and regulatory regimes support it, but it won’t be the general case.
A second approach is to beef up the capacity for existing urban rooftop macrocell sites, where 1Gbps backhaul capacity isn’t unusual, and use this as an aggregation point for clusters of nearby small cells. These can be connected via wireless backhaul – 60GHz point to point provide the highest link capacity alongside point-to-multipoint solutions which provide flexible distribution architectures.
Geert Standaert, CTO Proximus Belgium, believes that the macro network will reach a point where it can’t meet today’s exploding data demand. He’s thinking ahead of how best to deal with that, and this does include designing in more fibre to cope with future densification. For Urban small cells, the three key tenets of location, backhaul and power remain fundamental. He’s developed good partnerships with utilities such as the Electricity companies who can provide and enable new sites.
Cable companies (MSOs) may also get in on the act. US Cable operator Cox actively promotes the use of strand mounts (hanging on the wires above street level), pole mounts and ground cabinets to support small cell deployments.
A senior ATT representative told me that guidance for urban small cell design is to use fibre where possible, in order to prepare the ground for future 5G evolution. However that doesn’t rule out wireless backhaul. Sprint seems to be taking a more
Flexibility versus long term static investments
The large footprint of macrocells means that they are resilient to small changes within that space. This includes both daily migrations (eg to/from work, lunchstop, entertainment venues) and longer term (eg office moves, retail to residential change of use).
Small cells are much more targeted and may be needed to satisfy just one of those needs. A change of use from retail to residential might need additional nodes or relocation to better serve end-user demand.
This leads to a view that small cell deployments need to be much more flexible and adaptable. Significant use of wireless backhaul will be appropriate, so that new cells can be quickly deployed and existing ones moved. Architectures which demand dark fibre to every single urban small cell site will be slower to adapt and more costly.
Wireless backhaul enables adaptability and flexibility
While wireless backhaul solutions can scale to handle end user traffic, they would struggle to provide the multi-Gbps rates required for some network virtualisation architectures.
Cloud RAN that relies on dark fibre to every site may be of interest to a few network operators for use in some macrocell sites, especially where space is at a premium. But only where dark fibre costs and available are minimal and easily available.
Flexible urban wireless backhaul enables rapidly deployment and automatically adapts to network changes (in traffic from any site, link outages and network upgrades/densification). Combining this with backhaul aggregation points either at existing urban macrocell sites or additional dedicated nodes provides the ultimate in flexibility to meet rapidly changing evolution of mobile data traffic growth.