Richard Deasington, Director of Market Development at iDirect, believes that recent developments in satellite technology combined with small cells will make rural 3G mobile service viable even for smaller communities. We asked him why this is the case, why it's not just relevant to developing countries and what the industry is doing to extend 3G service to off-grid places today.
Satellite backhaul seems to be very expensive. Can anything be done to change this?
There are two major technical advances in satellite systems that are about to fundamentally change their economics.
The first is the move to Ka band, at 20-30GHz which delivers several benefits:
- New spectrum adds extra system capacity. Compared to the 2GHz bandwidth in the Ku band at 10-12GHz, we now have access to over 10GHz. Clearly, this adds several times more capacity across each satellite footprint.
- Higher frequency means smaller dish size. Since wavelength is inversely proportional to the frequency, high gain dish antennas are a third of the diameter of those used for Ku band. Smaller dishes are easier to transport and install, and are also less obtrusive.
- Satellites can use more spot beams. Rather than broadcast the same signal across their whole footprint, the satellites can reuse the spectrum many times over because they have been fitted with a number of small spot beam antennas for specific geographic coverage. The same spectrum can be reused in every second spot beam. This greatly increases the overall system capacity and total throughput available.
The second major advance has been the change from using dedicated bandwidth to packet switched architecture.
- The older fixed capacity allocation method (SCPC) left bandwidth unused, wasting system capacity that could have been used elsewhere. Allocating bandwidth on demand means that statistical multiplexing gains can increase total system capacity by anything from 30 to 80%.
- Rather than pay for a fixed bandwidth link, regardless of how much of it is being used, satellite operators can be more creative in their tariff plans, and for example, they can charge on usage rather than on a fixed capacity basis.
- For small cells, where there are potentially thousands of sites to be connected, it doesn't make economic sense to use dedicated bandwidth, so solutions that can centrally manage bandwidth will be used.
How has satellite backhaul been used in mobile networks to date?
Satellite backhaul is being used in many networks worldwide to connect macrocells in remote, rural regions. These include many islands and remote pockets of population who would otherwise not get a service.
Often this has been done with very limited amounts of bandwidth. For example, in Africa, a 2G voice basestation may operate with as little as 100kbit/s up to a few hundred kbit/s. Optimization techniques can squeeze the transmission down to as little as 4 kbit/s per voice call.
For example, one of our partners, Altobridge, has deployed many such remote sites in Africa and Asia, with their combined 2G/3G Altobridge Lite-Site.
The combination of new satellite capacity leading to lower rental costs per MHz of satellite bandwidth together with increases in efficiency from packet switching and dynamic bandwidth allocation means that OPEX costs (annual running costs) have dropped by a factor of 3 or 4 times over the past 4 years, making satellite backhaul more viable for rural communities.
Satellite links are also used in developed countries to reach difficult locations. It's not unusual for a typical rural 3G site to be provisioned with 5Mbit/s downlink and 2 Mbit/s uplink.
How would a mobile operator rent capacity on a satellite system?
Satellite networks typically leasefixed bandwidth to their customers. Ground station equipment at the remote cellsite would connect in the same way as a leased line, using equipment from manufacturers such as iDirect.
At the core switching center the satellite ground infrastructure is set up including a larger antenna, a hub system and management software that is used to configure and manage the satellite network. Traffic from multiple remote sites would be consolidated at the switching center and tied into the mobile operator's core network.
What else have iDirect been doing to enable small cell satellite backhaul?
We've been very busy working with many of the popular small cell vendors to check interoperability with our equipment, which already supports the service. We are supporting many trials of small cells from Southern Africa to Europe, with very good results to date.
We've also joined the Small Cell Forum where I chair the Rural Special Interest Group.
We are seeing a lot interest in the small cell over satellite approach for rural connectivity, not just for 3G but also for LTE.
What commercial arrangements are you seeing now?
One of the significant differences with small cells is the greater fluctuation in traffic demand. Larger macrocells with a few thousand users tend to show a pretty flat traffic flow or profile during the busy times By contrast, small cells which support tens or a hundred users see much wider variation in usage levels. In very remote areas with a handful of users, traffic spikes would be more extreme. This makes solutions that dynamically allocate bandwidth very attractive for small cells.
A commercial solution to this problem is for a mobile operator to rent spectrum from a satellite operator for use across a specific area. This can be constrained to use specific spot-beams from the satellite itself, but can be freely used by any of operator's basestations in that geography. This gives the mobile operator a total backhaul capacity which is available on-demand to all sites, allocated dynamically based on instantaneous usage rather than installed capacity at any given time.
So for example, if a village has an event which draws the crowds who all want access at the same time, then the small cell serving that community can operate at full capacity for that short time without any additional cost to handle other less hectic periods.
Does the business case stack up?
If you take a typical Western European ARPU of $33/month, then a site with 100 subscribers could raise a previously untapped income of $3,300/month. A proportion of that could easily pay for a satellite bandwidth of 5 Mbit/s. Satellite operator Avanti have presented figures indicating a backhaul cost per subscriber as low as $5/month for voice and data in a village served by small cells.
The CAPEX cost of the link is somewhere around $3 to $6,000 for the hardware, which equates to less than $100/month over 5 years.
How could site sharing affect these arrangements?
The Small Cell Forum is already looking at what's involved for rural small cell sharing. From a satellite backhaul perspective, there's no reason why multiple operators couldn't share a satellite ground station serving multiple small cells – one for each network operator.
Some operators believe in a land grab for remote rural communities, because it may not be commercially viable to install multiple small cells from different operators. The total revenues would not justify further investment when one operator has already installed a site and captured all of the local users. While this may be valid in developing countries with low ARPU, we may see more open site sharing by multiple operators in developed countries. Some small cells can already accommodate multiple radios from different operators, making this both practical and cost effective.
iDirect sponsors ThinkSmallCell. Visit their website for more insight into how satellite broadband can be used for cost effective rural small cell deployments.