Sudhir Tangri, Marketing Director of “innovation, technology and outsourcing” company Aricent, explains to David Murphy why femtocells will be a vital part of future high-speed mobile networks
DM: So give us the low-down on Aricent if you would please Sudhir.
ST: Sure. Aricent was formed in 1992 as an offshoot of HNS and initially, we were called Hughes Software Systems. In 2004, we became part of Flextronics Software Systems, and then in 2006, a private equity firm invested in us, and we became Aricent. But we have always been extremely focused on the telecom space for the past 16 years, and we have more than 8,500 consultants and engineers working in this space on a variety of technologies.
Our headquarters are in Palo Alto, California, and we have 19 offices and delivery centres based in India, China, Germany, the UK, S. Africa, Mexico and Ukraine, which means we can be close to our customers.
DM: And how long have you been working on femtocells?
ST: We started quite early, in 2004-05.
DM: And what have you been doing?
ST: There are three areas that we have tried to address. The first are the semiconductor vendors. These are the people who come up with chipsets so that the equipment vendors can build the hardware. This is the first phase of the market.
Then the equipment vendors started building their products around these chipsets. This started back in 2005, and we were involved in one of first commercially available femtocell products in February 07. In fact, we can say that we have more than two thirds market share in the equipment maker space, and we have worked with a multitude of equipment vendors and architectures.
Then in the last three quarters, we have been working with operators, helping them to set up test labs and do friendly user trials, working with the engineers in the trenches rolling out the trial networks. Right now, there are seven or eight different operators doing these trials, and the two biggest challenges they face are interference and radio resource management.
DM: So seven or eight operators conducting trials. What about the rest?
ST: A lot of the others are slightly lagging behind, still setting up test labs and such like.
DM: But are they all agreed that this is something they need to do?
ST: Generally, yes. The vast majority of operators believe there is a need for it and a business case for doing it. But different operators are at different stages of adoption. Some are doing user trials, with others, we are trying to help them set up test labs, and there are others who are still trying to define the business case. But in mature markets, the operators who have rolled out 3G networks know that they need some kind of femtocell solution, because indoor coverage is a problem, and around two thirds of mobile calls are made by people when they are indoors.
DM: So what have you learned so far?
ST: We’ve learned that the three sub-segments, the chipmakers, the equipment makers and the operators, each have their own challenges.
Three years ago, the chip vendors’ challenge was to push the product to the market. Now it is not to push it out, but to bring the costs down. Currently, the BOM (bill of materials) cost is around $80-100 dollars, so the box price will be $160-200. These price points are very high from a subscriber’s perspective. So the question is, can they reduce the BOM price to $35-50, so the box price would be $80-100?
We have been talking to a wide spectrum of vendors, and there are two different approaches to the problem. At one end, you have the guys who are trying to shave certain features off a large platform to make it good for a femtocell, and at the other, you have the guys who are trying to do an application processor upgrade themselves to support the power and requirements for a femtocell chipset.
DM: But doesn’t the price of chips always come down eventually?
ST: From a price point and volume elasticity perspective they are stuck in a chasm. In the semiconductor industry, volumes go to hundreds of thousands of units and the price points drop, but in the femtocell business, the industry expects the cost point to be $100 dollars a box for the volumes to kick in and the big semiconductor players are not seeing those volumes so they have not started investing in it to bring the costs down, so it’s a Catch 22 situation.
So the vendors are waiting for this to happen, but they also have another challenge, which is that the operators are asking them to ensure that these products work. But when they do final testing, it is a closed lab environment, which is very different to putting in a live network. Things may work fine in a controlled environment, but in a live network there is interference and handover issues and all sorts of other things to contend with. So you get teething problems, and this is what the equipment vendors are trying to fix.
DM: So how do you think things will pan out?
ST: There is a business case, and the industry is excited, but the question is, how quickly can the industry work to resolve these challenges? For the moment, we’re talking about technical challenges, but if they can sort these out in the next three or four quarters, then the next set of challenges are more around marketing and distribution. How do you market and distribute these devices? It has to be a situation where the customer buys it in the shop, brings it home, switches it on, and it works, because if it needs an engineer call-out, it will not be viable.
DM: And how do operators get round the fact that for most people, this indoor coverage problem is one that does not exist for a lot of people?
ST: I think this is less of a challenge, because looking two years ahead, LTE (Long Term Evolution) networks will have a very high data rate, 10Mbps, and if you are used to it outdoors, you will expect the same user experience in your home or the office. But LTE works on 2GHz or higher frequencies, where indoor coverage will be a bigger issue.
At the same time, demand for these faster networks is being driven by innovative applications like Facebook and YouTube on your mobile. So if you have this LTE network, and you can see YouTube videos on your handset, you will expect the same user experience and performance inside the home that you get when you’re outdoors, and for that, you need a femtocell. So the opportunity is to make the femtocell part of the package
DM: So are you confident, then, that femtocells have a future?
ST: LTE has to happen, because new applications and new devices are creating the demand for that kind of bandwidth. There’s a huge difference in subscribers’ hunger for bandwidth pre- and post-iPhone. AT&T’s network in the US has been choked since the iPhone launched. As soon as you touch that device you are generating data over the air, and in the next couple of years, there will be a plethora of devices like the iPhone, coming from the operators or from Yahoo and Google, so this happening. Then the other element is the social networking applications, which will become more and more popular and will drive the bandwidth requirement. So I believe it will happen, and we will keep chipping away to make it happen.
Recent Comments