When the industry is developing a new technology, those who support the early design processes often know best where things are heading. Keith Dyer spoke to those whose job it is to provide test forms and cases for the early stage developers of LTE, to find out what stage market development is at, and where the steepest hills to climb might lie.
The 3GPP is aiming to have the standards for UTRA-UTRAN Long Term Evolution (LTE) finalised by September 2007. Standards for 3GPP System Architecture Evolution (SAE), which addresses the core network supporting the fast radio access network, “should follow”, the 3GPP stated last October.
” height=”<% height %>” align=”right” alt=”LTE: a testing environment – Fast forward” class=”articleimage” />
And the pressure is on, with seven operators last year forming the next generation Mobile Networks initiative, aimed at making sure that issues of IPR and interoperability didn’t hamper the progress of the standard, and with rival wireless broadband technologies also based on OFDM and using MIMO technology progressing quickly.
Renaud Duverne, European Market Development Manager, Signal Sources Division, Agilent, says: “LTE is certainly going to be one of the buzzwords at 3GSM, because the standard is coming and companies are working on that. I definitely reckon it will be the term that will be adopted as standard, amongst all the acronyms we’ve seen so far for this technology.
“It’s more stable now, although the standard is not finalised until September 2007. So at 3GSM we will show our first LTE solution, following the release of new design libraries for the LTE standard.”
Phil Windred, General Manager, at Aeroflex’s business in Cambridge says:
“We see LTE moving very fast, even faster than previous new technologies. There’s pressure from Japan with DoCoMO, and from WiMax putting the mometum behind LTE. WiMax is a threat to cellular operators and LTE is the answer to that threat.”
Jonathan Borrill, director of market strategy at Anritsu, agrees that there is now considerable market momentum behind the technology.
“The industry has seen a lot of surprisingly quick developments over the past three to six months, since the first draft of the standard, and 3GPP made its selection, in mid 2006. We’ve seen network suppliers moving very quickly to do things on LTE, and it seems quite aggressive to get LTE functionality ready for market.”
Yet some think that the September 2007 date seems a little ambitious. One of those is Othmar Kyas, Director Strategic Marketing, Tektronix Network Diagnostics.
“Most delegates on the committee think that is a very ambitious time line,” he said. “So, basically, it is roughly two years behind where WiMax is today.”
Yet despite these slight differences on the timing, there is clear agreement that LTE development is happening fast – and it is happening now.
Agilent’s Duverne says that the first products from the test houses will come on the design level, followed by Beta releases as early as February this year on signal generation, creating a pre-configured compliant signal and a pre-defined set up on the receiver side.
“We’re already engaged with a few customers and they’re already preparing for transition from one to another — they are looking at the design mode right now.”
Yet the industry is faced with developing physical layer test products for networks and devices for technology that is very different from the current WCDMA and GSM environments. LTE takes mobile operators from a WCDMA environment into an air interface based on OFDMA and MIMO antenna technology.
Tektronix’s Kyas says: “LTE is based on OFDMA in downlink and SC FDMA (Single Carrier FDMA) in the uplink. This is significantly different to CDMA and quite a challenge for operators because managing, planning and optimising these technologies is very different.
“There’s also the addition of the MIMO antenna concept.”
The new air interface also involves a difference both in the range and dynamics of frequencies used.
“UMTS is up to 3GHz and OFDMA up to 8GHz,” Kyas says, so test instruments require higher frequency support.
“Secondly, the dynamics in the air interface are much higher in networks with MIMO – beam forming active antennas with the intelligence to dynamically react to application service profiles make it difficult to measure, assess trends and isolate faults or problems.
“In the past, the RF characteristic was relatively stable for walk and drive tests. But OFDMA operates through a large number of sub carriers to transport the bits, so the transmission of traffic is split onto 512 or 1,024 subcarriers across the spectrum. The advantage is that where the best characteristics are you can select a bundle of subcarriers for transmission.
“So it’s fascinatingly dynamic, but for test equipment vendors the complexity is high, with the need to adapt to RF standards and find new algorithms and processes for identifying faults.
“We need to start with new protocols and find new ways to characterise and manage the network. It’s time to define which KPIs are really relevant and displaying what goes on in the network.”
Anritsu’s Borrill says that although OFDM is a known technology through its use in WiMAx and WLANs, the task for the industry is to learn how it will work and be implemented in cellular networks.
“MIMO is very new for everybody, it’s just come in for WiMax and WLAN and we have to figure how MIMO really works in a cellular network. It’s the same for OFDMA. Today, there’s not an agreed basis for test specs.
“OFDMA as a basic technology is understood but in a mobile network it’s a very complicated algorithm to the handset and base station. MIMO is literally just coming out of the research environment. It’s now about understanding how it will work in the real world, in the street and with real interfaces. So new algorithms will be the big challenge.
“One further challenge is on the handsets, getting into handsets without increasing the price, power consumption and size. That’s one of the critical factors in the choice of MIMO/OFDMA.
“It’s the reason 3GPP specified SCFDMA on the uplink, as you can’t have handsets with such wide RF bandwidths. It would be much too expensive for handsets.”
Of course, LTE is only part of what is happening within R8. Apart from the air interface, the 3GPP is defining the core network architecture, the System Architecture Evolution (SAE), that dispenses with the circuit switch-based RNC, SGSN, and GGSN network elements.
“With the RNC, SGSN, GGSN all that thrown away in SAE, there will be an evolved NodeB , or a superNodeB, to take over everything that the RNC does, making it a much more complex and intelligent piece of equipment,” Borrill says.
“With the main gateways completely removed there will be a new entity called an access gateway, with all the new interfaces based on IPv6.
“So there are questions on fundamental technologies. Customers wanting to build solutions need to know what are the test solutions, what are the important factors in OFDMA signal beahviour, interfaces and bad signals, etc?
“At the system level on IP, with a NGN being all IP, with people starting to put VoIP over the network — how do you plan the network to put all the voice traffic over IP?”
Yet despite the inevitable hurdles to overcome, there is a positive sense too that the industry is now ready to meet challenges, as Aeroflex’s Windred says, that have caused it to stumble before.
”The whole industry learnt a lot in the move into WCDMA and I’ve seen everyone raise their game. The test industry has to be a part of that, taking products to another level, providing observability and analysis of data to let people fix problems. “Because we know there will be problems — the thing is to fix them as quckly as possibile.”