Building a Backhaul Mix: Key to optimising costs and capacity

Backhaul , May 23, 2018

Atirek Gupta, Consultant, Capitel


The two primary ingredients of a mobile broadband business case are spectrum and backhaul. Once operators have sufficient spectrum in the bands of their choice, the next critical decision to maintain the user experience is the backhaul design and capacity. In highly competitive markets such as India, the backhaul cost, defined by the mix of underlying technologies, also becomes a key consideration.

The drivers for backhaul remain clear increasing traffic density on 4G, bundling of fibre-to-the-x (FTTx) with mobile services, likely deployment of small cells, cloud radio access network (RAN) and 5G.

Traffic density (Mbps per square km), in contrast to traffic levels (GB demand), is the parameter of interest. Traffic density reflects the adoption of data services by customers in a specific cluster, their data usage as well as the level of simultaneously active users. Our modelling for some Tier I cities suggests increasing traffic density, as data usage becomes more widespread across the city, and absolute consumption of data continues to increase.

Such a concentration of demand results in high traffic in hotspots and dense urban locations within cities, and loads wireless sites 20 per cent of the sites are estimated to carry more than 50 per cent of the traffic. The next 30 per cent of the sites will be carrying about 30 per cent of the traffic. Operators are identifying these sites and migrating them on a high capacity backhaul to manage the surge in traffic.

Traffic density will continue to increase. The spectrum band under consideration for 5G in India is the 3.3 GHz to 3.6 GHz band. A pilot deployment on these frequencies by an original equipment manufacturer suggests more than 20 nodes per square km, representing about three to four nodes per macro cell. Such a concentration of 5G (and small cell) nodes, especially with a significant quantum of access spectrum and high spectral efficiency will put pressure on the backhaul.

Finally, with increasing competition in mobility, operators in India are also considering the deployment of various versions of the FTTx (x being home, node, tower). Such FTTx deployments will also need fibre backhaul.

Operators and network planners understand that a strong backhaul, combined with a ubiquitous small cell infrastructure, is going to provide them with a competitive advantage and improve quality of service (QoS).

Currently, 80 per cent of the operator backhaul is on microwave and the remaining 20 per cent is on fibre. Operators will have to find the right technology mix for their backhaul that supports the capacity needs and is also cost efficient. To defend and build their market position, operators, independent fibre companies and tower companies are coming together to develop a high capacity backhaul network.

Our engagement experience suggests four different access media for designing a capable and cost-efficient backhaul, including microwave, millimetre wave, dark fibre and lit fibre.

We believe the initial transition, which is already taking place in metro markets, is the upgrade of microwave backhaul and migration of select microwave rings to fibre links.

Right-of-way (RoW) charges for underground fibre can become a limiting factor in select markets, as well as select circles such as Maharashtra. However, in markets with the expected deployment of FTTx, data densification on 2300 MHz, 2500 MHz, 5G deployments on 3.4 GHz+, there is no real alternative to underground fibre. The planning needs to identify the locations that will need fibre from all the above areas and to migrate these points of presence (PoPs) to fibre, with the rest of the network backhaul using a mix of other technology options. The Smart Cities Mission is helping a few operators and towercos in reducing these costs, but this option is not available to multiple operators and is applicable to only a restricted number of cities.

The next best solution to underground fibre is aerial fibre, although the regulatory framework and the supporting infrastructure for laying a structured aerial cable on poles are limited to a few cities. Aerial fibre is being increasingly used in business-to-customer (B2C) use cases by local cable operators and for specific small cell backhaul in Tier II markets, although we do not expect this to be a sustainable long-term solution for high capacity and reliable backhaul.

Regulators are also providing access to mmWave E band and V band spectrum for backhaul. The available capacity of mm Wave spectrum can be 1 Gbps per link, subject to propagation issues. E-band spectrum is expected to be a clear substitute for fibre in tough patches and areas with high RoW charges. A mmWave radio hop is currently priced at about $5,000 per hop, although this is expected to rapidly decline with scale. A combination of mmWave links with microwave links and fibre can help optimise the overall cost of backhaul, and can be configured at a market level.

Such a mix becomes especially important when planning for small cell deployments. The cost of backhaul as a percentage of the total equipment cost is double for a small cell network (point-to-multipoint and point-to-point) when compared to that of a macro-cell network (20 per cent).

The final decision in planning and optimising the backhaul cost is the mix of dark and lit fibre. For sites that are carrying the majority of the data traffic, dark fibre provides a clear business case the availability of dark fibre is likely to improve due to the high earnings before interest, taxes, depreciation and amortisation margins (60 per cent+) for the dark fibre business for the suppliers, as well as the expected consolidation and organisation of fragmented metro fibre networks in India. Lit fibre, including E1s and STM-1s, will be used only for cell sites with predictable and limited backhaul capacity.

The cost of backhaul (or leased bandwidth) within the overall cost per GB is likely to continue to grow as operators deploy additional access spectrum, refarm 3G to long term evolution, and introduce small cells, C-RAN, 5G and FTTx into their network design. The expected availability of mmWave spectrum and dark fibre will help address this capacity demand, with a sweetspot of functionality versus cost achieved through a mix of these new technology mediums and microwave.



Copyright © 2010, All Rights Reserved