NextWave plunges into WiMAX chip sector with focus on metrozones and Wi-Fi
As Qualcomm faces its demons, its former wireless division president Allen Salmasi is trying to recreate something of the CDMA giant' s business model in WiMAX, through the company he heads, NextWave Wireless. NextWave has acquired a considerable amount of spectrum in the AWS and 2.5GHz bands across the US and plans to launch WiMAX-based services, but like Qualcomm it plans to use these frequencies to develop, test and showcase its own technologies and then transform itself into a silicon-driven company. It has been unclear at what stage NextWave would look to sell what it calls its WiMAXplus designs to third parties, but now it has detailed its roadmap for entering the merchant silicon space, rather earlier than expected.
The company faces a massive uphill battle - its 16e roll-out schedule currently lags that of several larger players, and unlike its role model, it is not clear that it holds significant IPR rights. To create differentiated chip designs for its own systems was in line with many vendors' thinking on how to retain competitive edge in a standards-based market - Motorola, for instance, has returned to silicon design for 16e devices, despite the spin-off of Freescale. But to go to the open market in an already crowded market will require significant design differentiation and the rapid creation of an ecosystem geared up for volume - and the jury is very much out on whether NextWave can achieve that.
On the design side, it has some interesting angles that fit with key operator trends, notably tight integration of Wi-Fi and WiMAX to reduce the early stage capex associated with 802.16e networks, and a focus on the recently neglected license-exempt spectrum options for WiMAX. Both these approaches will enable NextWave to target the metrozone sector, in which WiMAX/Wi-Fi combinations are starting to play a role, as witnessed by the entry of Clearwire - in operator terms, NextWave's most direct rival - into the municipal market. But of course, Clearwire has turned its back on designing its own equipment with the transfer of NextNet to Motorola.
The contrast shows the different strategies of the two would-be disruptive providers. Clearwire acquired its own equipment arm to give it a rapid entry to market with readymade systems, and a measure of control over roadmap and pricing, while it remained in the start-up phase, with limited cashflow and forced to rely on pre-standard systems. But its real aim is to be an operator, and it was quick to turn to a large vendor once it started to enter the phase of implementing standardized 16e kit.
NextWave, however, is using the spectrum ownership and operator model as a springboard into a technology and IPR business, as its new roadmap indicates. Through its NextWave Broadband subsidiary, the company will launch 802.16e chips and make its first direct move into integrated circuit development. It aims eventually to create a system level infrastructure business and a dedicated components business offering baseband and RF chips, as well as licensing IPR and running its own networks (at least for a while - increasingly it seems that, in the medium term, NextWave would look to spin off or sell the spectrum assets and carrier units).
NextWave aims to develop a chip-to-application system that echoes the top-to-bottom strategy of Fujitsu. This will combine WiMAX and Wi-Fi and include systems for metrozones as well as licensed carrier use. It has made a string of acquisitions to fill in gaps in the product chain - including IPWireless with its TD-CDMA broadband wireless technologies, PacketVideo for multimedia software and metro base station maker Go Networks. The Broadband unit, which contains these purchases and the Advanced Technology Group for R&D, now employs 350 people in a total payroll of 1,000.
The company is now close to sampling its first generation NW1000 chips for 802.16e, though it says it has been investigating WiMAX designs for four years. NextWave, which in its previous incarnation acquired huge amounts of PCS spectrum and subsequently went into Chapter 11 bankruptcy protection, made its first stab at chip design in 1996 when it developed CDMA 1xRTT products, spinning that business off into Dot Wireless, which Texas Instruments acquired in 2000.
A second generation 802.16e chip family, NW2000 is expected in early 2008 and will support wave two of the WiMAX Forum profiles. It will feature an integrated WiMAX/Wi-Fi system-on-chip (SoC), designed in 0.65-micron CMOS. The Wi-Fi side will run in 2.4GHz and the WiMAX interface in TDD and FDD spectrum from 1.7-3.8GHz. Salmasi expects WiMAX and OFDM-based Wi-Fi to converge over the coming few years and says baseband chips combining both standards will be the basis of pre-4G, as will new base station formats such as femtocells.
The NW1000 features PCI and SPI host interfaces; highly integrated direct conversion RFIC architecture; scalable channel bandwidths from 1.75MHz to 20MHz; and MIMO support. Its successor will include the NW2100 family of mobile subscriber baseband SOCs, the NW2700 and NW2800 family of WiMAX access station baseband SOCs, and the NW2200 family of integrated multiband RFICs. These will be targeted at multimedia and QoS-sensitive applications and the ASIC architecture promises advanced power save features to extend battery life. The family will support a range of host OS environments and system interfaces, including SDIO, SPI, USB2, and PCIe; the Flexi-MAC architecture for hosted or fully integrated MAC. enhanced MIMO support and optional beamforming; and the 802.16h contention-based protocol to enable WiMAX operation in license-exempt bands. Throughput should be around 36Mbps. The radio will accommodate multiple service types including unicast, multicast, and NextWave's broadcast solution while a convergence sublayer will support next generation IPv6 networks. Reference designs will include handsets, smartphones, PDAs, PC modem cards, fixed CPE modems, USB dongles and Personal Media Players.
Also included in the NW2000 platform are NextWave's NW2700 and NW2800 baseband SoCs, which are designed for light and medium duty infrastructure applications such as in-home access points and femtocells as well as outdoor access points and micro-BTS devices.
If WiMAX does become an important technology in the metrozone and casual access sectors, NextWave is positioning itself cleverly for that space. With most of the large players heavily focused on full mobility and handsets, it could carve out a valuable niche and satisfy the demands of operators that feel unlicensed spectrum has dropped off the mobility-dominated 16e agenda. It is leveraging some key current trends that could help in its task - moving towards picocells and femtocells, which will help bring the metrozone network indoors and so drive new revenue streams; adapting WiMAX so that it could run in the lightly licensed US 3.65GHz band (see separate item), while claiming to remain fully interoperable. The shortcomings of the current Wi-Fi mesh model for metro area build-outs certainly give WiMAX a real opportunity to fill the gaps - allowing operators to add higher value services with superior quality of service and bandwidth. This is something that Clearwire and major metrozone players like Cisco are recognizing. But to serve such potential partners, NextWave will need to get its own supply chain and manufacturing allies in place quickly to give itself the economics to compete in a standards-based and therefore price sensitive space. Of this activity it has so far presented little evidence.
wimaxtrends.com |
