Unless you’ve been hiding in some dark cubicle night and day for the past year (which, admittedly, is easier for this audience than for most), you’ve heard about WiMAX, a.k.a. IEEE 802.16, and the promise that it brings for both the fixed and mobile sides of the broadband wireless market. Learning more about it is a bit like swimming backward through acronym soup with a side of evolving standards, but such is life with a ground-breaking global standard.
WiMAX, which stands for worldwide interoperability for microwave access, is an evolutionary step in wireless technology that may finally cause the industry’s long-anticipated global revolution. It takes on and solves many of the challenges that the wireless industry faces today. But rather than acting the role of standards bully, coming in and displacing all of those before it, WiMAX is more of a facilitator, building on the value of existing standards to achieve better quality of service, increased flexibility, and reasonable economies of scale.
There are two different versions of the IEEE 802.16 standard, called WiMAX 802.16-2004 and WiMAX 802.16e. Unlike the evolutionary path of the Wi-Fi (IEEE 802.11) standard, with its a, b, and g versions, the two flavors of WiMAX are designed to address demands for different markets, and both will continue to exist on parallel paths. Version 802.16-2004 is a bit more established, having been standardized for more than a year at this point. It supports fixed and nomadic access in line-of-sight (LOS) and non-line-of-sight (NLOS) environments. Version 802.16e, optimized for the needs of the mobile market, was ratified in December 2005. The completion of the 802.16e standard is a significant milestone for the design community because it eliminates ambiguity in what can be architected, designed, or built. Now designers know the exact focused set to build their product around for interoperability and potential global adoption.
As its definition implies, WiMAX is targeting a global audience. Creating a standard that plays well with others, particularly on an international scale, is no easy task. The WiMAX Forum was created by industry leaders to support the IEEE standard. The goal of the forum is to achieve global acceptance for WiMAX, and that demands tight coordination and cooperation with international standards organizations like ETSI (European Telecommunications Standards Institute) and their HiperMAN (high-performance radio metropolitan area networks) standard, as well as with Korea’s WiBro (wireless broadband) technology. The WiMAX Forum has also created a certification process to ensure that new products are fully interoperable and support metropolitan broadband fixed, portable, and mobile applications. This process is currently active for 802.16-2004 products, and they estimate that certification of new 802.16e products will start in mid 2006, with the first certified products available in early 2007.
WiMAX Takes OFDM to the MAN
The WiMAX Forum has chosen to adopt orthogonal frequency division multiplexing (OFDM) and orthogonal frequency division multiple access (OFDMA) techniques for the 802.16 standard. Intel Corporation has been playing a key role in the WiMAX Forum since its inception. Chris Knudsen, CTO for the Service Provider Business Group at Intel, says, “Traditional telecom systems have either been GSM or CDMA and those are timed sliced, which gives you time diversity, but OFDMA also has frequency diversity. We’re moving into an era where you can schedule and optimize wireless transmissions to a channel based on both time and frequency, and we should see further evolution around that technique. OFDMA adds a dimension of processing, or the ability to be more granular for the architects. You now have another dimension of your palette with which to design your architecture. I think you’ll see the efficiency of these systems both in power and the ability to deliver capacity to users increase over the next 5-10 years, and OFDM will be the architecture of choice.”
Knudsen says he is excited about the new palette to paint from with more flexibility and complexity, providing more opportunity to be creative with design. “I think back to designing architectures, and one thing that OFDM does is that each individual tone is 10-11 kHz wide, so the requirement for channel equalization is orders of magnitudes less than for CDMA technologies. So you can process complex signals and get better spectral efficiency with less power. You also have the granularity to divide users between all those different tones.” He points out that the design challenge is far from over. “The challenge becomes how you digitize all of this and take this information and divide it up and process it among multiple users. It’s a very challenging, complex architectural problem, but something that gets engineers excited.”
You may be wondering about overlap with established standards like Wi-Fi. Wi-Fi is the indoor, unlicensed, wireless access for personal broadband devices. If you consider the fact that there are more than a billion cell phone users globally, how do you take broadband and cost-effectively move the services to a mobile platform? Enter WiMAX. With WiMAX, it’s the first time that a wireless metropolitan area network (MAN) has been taken to a standard like IEEE, and the first time that OFDM and OFDMA have been applied to a MAN. Although OFDM is incorporated in 802.11, this is the first time it’s been standardized on a MAN, and industry and technology has accepted it as the access mechanism for the physical layer and the evolution of MAN technology.
Alrighty. WiMAX and Wi-Fi work together. Now, how does WiMAX fit in the market vs. traditional voice carriers? Well, all of the systems deployed today were designed for voice first, and data represents a market growth opportunity. If you use your cell phone to manage data today, you probably feel like your carrier does a decent job, but you probably also cut them a little slack because it’s an emerging market and the technology isn’t optimized for data.
WiMAX is architected to optimize cost on both the client side and the infrastructure side and to align the network architecture around an IP-centric architecture. Comparing those two, voice networks primarily use 3GPP or 3GPP2, and both of those are migrating to interfacing to IP structures, but the infrastructure is very complex. WiMAX is focused on structuring around the Internet architecture and taking advantage of the cost efficiencies with the standard Internet features.
One of the key technology variables with WiMAX is spectrum. The spectrum target is predominantly the new spectrum that is coming out. Current voice systems are impacted in capacity by their current spectrum, and they’re looking for alternatives to expand new data services. WiMAX is a complement here as well, because it will be the most cost-effective method to deploy those services in a new spectrum.
Evolving Market Favors Programmable Logic Flexibility
The promise of growth that WiMAX brings to the wireless segment is by no means lost on the programmable logic community. WiMAX is like any other wireless technology. It’s heavy in physical (PHY) and media access control (MAC) layer processing, which is all real time. Those applications are highly appropriate to programmable logic, and the vendors are very much involved.
Xilinx has been leading the programmable logic charge in WiMAX, both as a member of the WiMAX Forum since 2004, and with involvement in the IEEE 802.16 standard since 2001. “At Xilinx, we have a vertical market focus, and we chose wireless as a key market to support years ago,” said David Gamba, director of Vertical Marketing at Xilinx. Xilinx sees opportunities for programmable logic in both the short term and the long term, and they have created comprehensive WiMAX solutions, including devices and IP. Gamba indicates that there are many ways to leverage the flexibility that programmable logic offers for WiMAX applications. One example is finding a way to deal with different air interfaces available for WiMAX. “The most common are licensed bands, at 3.5 GHz and 2.5 GHz. There’s also 5.8 GHz, which is unlicensed. Using programmable logic, you can sit behind your PHY device, implement your MAC functionality, and adapt to the different types of chips that could be used for a licensed vs. an unlicensed band. For example, if you’re using a 5.8 GHz unlicensed band, you may encounter a lot of interference that may require additional logic for advanced error correction. In a licensed band, you may not need this error correction because the interference is less.”
Arun Iyengar, senior director of Altera’s wireless business unit, sees a strong play for programmable logic, particularly on the fixed side, with WiMAX 802.16-2004. “We’ve been focusing on scalable OFDMA, which allows you to pick a number of sub-channels on the fly. OEMs using Altera FPGAs with our scalable OFDMA solution can support any WiMAX profile for their customer. Perhaps one operator requires 1024 subchannels, and another operator needs 512 subchannels. It’s as simple as changing a setting to support each customer’s WiMAX profile.” Will it finally reach a point where you can standardize on a specific PHY and a specific MAC? Iyengar believes the answer is yes, but that it may be 5 or more years out. This is a brand new technology, and, in terms of operator deployment and decision making, there’s still a lot of fluidity.
Lattice Semiconductor has also recognized the importance of WiMAX for the wireless market. “Lattice has developed an OFDM transceiver reference design targeted for WiMAX 802.16-2004,” says Clement Lee, Director of Corporate Product Planning at Lattice. “In addition, we are updating our IP to support 802.16e.” Lee indicates that the short-term market for the company is in fixed broadband wireless, with a more long-term opportunity in mobile broadband. “We’re seeing a lot of activity in the 802.16e-based designs, and we want to be part of that.”
The future looks bright for demanding technophiles. If WiMAX fulfills its promise, we’ll have this technology standardized in a form that’s readily adoptable anywhere in the world. The question really becomes, what on earth will we demand next? I’m sure we’ll think of something…