Radwin will present its non-line-of-sight (NLOS) small cell backhaul solution at Mobile World Congress at the end of February. The Israel-based vendor has specialised in providing NLOS solutions for mobile operators in rural areas for eight years, but has now come up with a version that will enable small cell networks, such as LTE, to work efficiently in busy urban areas.
Adi Nativ, VP marketing & business development at Radwin told Wireless why the solution is so ground breaking. ‘The evolution of the cellular networks from 2G to 3G used to rely on the same infrastructure, so when they moved to 2G to 3G to 3.5G to HSPA+ they just changed the hardware in the towers for the base stations.
‘But when moving to LTE they found out that that was not going to be enough, simply because of physics. You cannot be too far away from the tower to get 100Mb to your smartphones and the same number people cannot share the same base stations.’
One solution to this problem is to deploy small cells, or picocells, in much greater numbers at street level where the smart devices can pick up the signal, rather than on the large 2G and 3G towers, which are sited on top of buildings to ensure line-of-sight transmission. But the question then, is how to backhaul the traffic?
Backhaul for small cell networks
‘When you are at street level you need to connect the small cells at street level and that means non-line of sight, because you can’t build big towers on top of buildings which require line of sight between them. At street level you have to go round corners and through buildings. But the other problem you then face at street level is that everything interferes with the signal; vehicles passing; new billboards; changes in the topography of the building; or just leaves coming out on trees. So it is a huge challenge,’ says Nativ.
The roll out of LTE networks requires backhaul solutions at street level. Radwin’s history leaves it well placed to meet the problem. It developed its NLOS technology to help operators in rural areas where the low density of population meant they were looking for solutions to help cut the cost of infrastructure. So, instead of building expensive tall towers, they build smaller ones, but then needed a way to cross lakes, go round mountains and perhaps cover small parking areas.
Radwin developed a product that was small and had a range of up to 120km. It was also designed to be simple to align, deploy and operate. ‘Why?’ said Nativ, ‘Because we have largely been working in emerging economies where low skilled people do the installation and where the product is often carried to remote sites by donkeys or by hand. It had to be deployed very quickly and easily therefore. We are talking implementation in minutes as the requirement.’
Another key requirement was low power consumption. Radwin’s product operates on less than 15W, as many are run on solar power in those locations. Nativ points out that in emerging economies the power supply is subject to black outs or fluctuations (although this also happens in developed countries where infrastructure in rural areas is less developed to save costs, so it is more susceptible to interference and power fluctuations).
‘We shipped 60,000 of these last year to 140 countries, but it’s a niche application in cellular where you can usually get line-of-sight transmission, so operators only go for this when they need to,’ says Nativ.
But now the industry has come to a turning point, according to Nativ. ‘If you want to deploy small cells you can no longer rely on line-of-sight microwave. The mobile carriers came to us, as they knew our equipment from rural deployments, and asked us; can we use your product for NLOS urban small cell backhaul? What we’ve tried so far isn’t really working.’
Reliable and robust solutions
Nativ explains that by ‘really working’ operators mean: it is reliable all the time, even when the topography changes or a bus goes by; it is deployable on street poles and other furniture; it is available today; and it can be integrated into the operator’s main network.
Radwin has been working on the challenge of supplying a small cell NLOS backhaul in urban areas for the last eight or nine months, according to Nativ.
‘We have been working with the top carriers, the leaders of the industry, in the US and Europe. We’ve taken our product out to their LTE small cell trials to see if it worked. This is a very carrier driven industry and we have got to have this. They are pushing everybody forward at the moment; this is not vendor driven, which is great.
‘The performance has been really good, not just NLOS, but maintaining a robust reliable link too,’ reports Nativ.
‘The work is at an advanced stage,’ says Nativ. ‘The product is ready as an off the shelf purchase. We’ve changed the antenna a bit as the distances are shorter and it is smaller than the rural version. Now we are working on how to integrate it into the small cell concept, methodology and network. We are doing a lot of work with picocell vendors and carriers.’
Non-line of sight technology
So how does NLOS technology work? The backhaul NLOS beam works by bouncing it off buildings in a zigzag down a street, for example, until it ricochets round the corner to connect with the next picocell.
It also requires a good modem and a lot of signal processing. The clever bit in Radwin’s solution lies in the signal processing with the development of smart algorithms, which correct errors and monitor environmental changes in the street. This ‘proprietary air interface’ is Radwin’s unique intellectual property.
‘The big difference between our technology and others is that we assume the environment is changing very fast,’ explains Nativ. ‘During night and day a building’s temperature changes, so if I am bouncing a beam off that building I need to know how it will behave when the building’s temperature changes and make adjustments if necessary. Our technology is built to follow those changes every millisecond and then change the transmission and reception parameters or communication algorithm to cope with that. This is what is really unique about our solution.’
Nativ points out that where an operator can use fibre for backhaul he would recommend that first. ‘Fibre is available, very reliable and has huge capacity, so if you can get it use it, But the problem is that while you can usually get fibre to a building, you don’t often get it to the street pole. You might only need a very short range connectivity, 20m-50m, but you can’t always get right of way to the street pole – who owns the fibre? Not necessarily the carrier – probably BT or Virgin in the UK.
‘So, the questions for Radwin is, how can we be as reliable and much cheaper for them to run backhaul over wireless? Bear in mind that in many parts of the world wireless is the main backhaul application, in others it is fibre,’ he adds.
Network planning and integration
Radwin’s technology is designed to work with any picocell or other product produced by RAN equipment vendors, but Nativ stresses it is the operational side that needs careful planning, integration and deployment. The site of the picocell must work in conjunction with the backhaul requirements too. ‘We must all work together to find the optimum solution,’ says Nativ.
What is being discovered is that the architecture of small cell networks is very different to roof-mounted transmission networks. The picocells have to be sited not just to provide the best signal, but also to enable the most efficient NLOS backhaul. This requires considerable testing at street level before an operator commits to a site acquisition or rental contract.
The backhaul unit is sited close to the picocell and connected via an Ethernet cable. But the power supply to the picocell also has to be considered. It is easy enough to get power to street level, but it then has to be connected to a picocell, which might be mounted on a lamppost or the side of a building.
‘Assessing where does the power come from is where the operation starts to change,’ says Nativ. ‘The acquisition of a site is usually a contract between the carriers and the municipality; not only to get permission to use the site, but to get the power in – that’s their issue. What we need to do is help simplify that interaction between the carrier and the municipality.
‘What we provide is a unit with low power consumption (15W), so the carrier doesn’t need to change the power to suit us; you get 15W of power to every street,’ says Nativ. ‘The unit also needs to be mounted very quickly and be small, so they can install it quickly and then it can be configured remotely. We’ve been doing this for years, as many locations in rural areas are remote, so we work closely with the carriers to enable that.’
Nativ believes that using Wi-Fi for offload or backhaul is an interim stage until the operators build out their own picocell networks (although many Wi-Fi vendors argue that even with new LTE spectrum and networks, the operators will still struggle to cope with the rising demand for data).
‘Carriers are using Wi-Fi, but they say Wi-Fi is good for Wi-Fi. It is no good for delivering 3G or LTE grade services, so they are looking for equipment to deliver much more than Wi-Fi is capable of. Wi-Fi does not have the robustness to deliver a proprietary solution – it does not have the right protocol,’ says Nativ.
‘Our technology is designed to be used in multi-path environments where there are a lot of reflections at street level. Yes, Wi-Fi is too, but when a bus crosses and interrupts the Wi-Fi signal it fluctuates. When we deliver non-line-of-sight carrier class backhaul it must remain constant with minimal change. This is what we have been doing for eight years, so that is why we meet the demands of LTE small cell architecture better than most other solutions out there,’ argues Nativ.
Network management and performance monitoring
‘It is very important that the operator is able to manage the network remotely and manage the performance of the system without having to get maintenance people down there. We run standard interfaces with network management and comply with standards for network performance monitoring. This is another difference between us and other solutions that have mostly come from the enterprise space and try to go into the carrier space, but they were not designed to do this kind of thing,’ says Nativ.
He cites the example of a network outage. ‘As the power comes back, with our system everything is still logged on the equipment and you can transfer that back to the control room. That is a requirement. The operator must understand what happened. We are part of any end-to-end integrated systems and we can do it very simply.’
Radwin is now working with the leading carriers to assess how the network will look and how it will be assimilated into the architecture and what the operational guidelines are on how to deploy it.
‘I don’t see it being deployed commercially until next year, but more trials are underway this year,’ says Nativ. ‘Our message is that there is no longer an obstacle to deploying small cell networks, as we have NLOS technology, so why not try it.
‘Solving the backhaul issue is the main challenge for the deployment of small cells. If it doesn’t happen, then I don’t know how fast LTE will happen or how fast it will be profitable, so the question is: can the operators serve as many customers as they want to with all those applications that they want to use?’