Weve all heard the ebb and flow of WiMax enthusiasm ranging from "it'll take over the world" hype to "it's already a bust" anti-hype, and everywhere in between. What's for sure is that a lot of companies, ranging from start-ups to (most famously) Intel, have bet a lot on the future of WiMax.
Most discussions of WiMax these days tend to center on mobility (whether the broader WiMax market will develop around fixed systems, or wait for the advent of mobile) or the role of "pre" WiMax gear in developing the market.
We met with some folks at Supercomm last month, however, who brought up an entirely different WiMax angle - WiMax in the LAN instead of the WAN. The folks at Cygnus Communications were at Supercomm making an interesting case for WiMax access points as an adjunct to (or replacement for) Wi-Fi APs.
Problems with Wi-Fi
By now, we're all pretty familiar with the benefits and shortcomings of Wi-Fi. The benefit side is pretty darn compelling: cheap (and getting cheaper), ubiquitous (and getting more so), easy (and slowly getting easier), and fast enough for most home networking needs. Sure, it's not as easy as it should be (how often do you get calls from non-techie friends and relatives for help with their Wi-Fi networks?), and network speed is like money in that you can never have too much. But for most folks and for most purposes (i.e. Internet connection sharing and data networking), Wi-Fi has got what they need.
Wi-Fi security is an issue, but one that is rapidly being improved by the adoption of WPA and now WPA2 encryption solutions. Wi-Fi range is another issue, also being solved by new technologies.
Wi-Fi QoS - the big bugbear
One area that we haven't seen enough progress on to date is QoS. We have talked before about the need for QoS, and for service-level agreements in the residential broadband network, and for mechanisms for enforcing and ensuring service quality within the home.
It's a problem if you If you're just trying to let a broadband customer surf the Web, this doesn't really matter; if you're trying to sell VoIP services or video (or gaming, or music services, or whatever) well it matters a whole heck of a lot.
For home wireless networks there have been two main "futures" which promised to help out here: the Wi-Fi Alliance's wireless multimedia (WMM) certification program, and the parent standard which enables this (802.11e). WMM gear has begun to hit the streets ahead of the final 802.11e standards approval (802.11e is still in draft mode).
WiMax is predictable, Wi-Fi is not
WMM and 802.11e do provide a much needed mechanism for improved prioritisation for traffic (which is why some Wi-Fi phone vendors have been offering pre-standards variants for some time). But the underlying 802.11 MAC relies upon Carrier Sense Multiple Access With Collision Avoidance (CSMA/CA), which sets a random "clock" for retransmissions of packets when a collision between stations occurs. "QoS" mechanisms like WMM can help prioritise these retransmissions, but they dont help avoid the initial collisions in the first place.
That's where we come back to the WiMax LAN. Unlike Wi-Fi , WiMax provides for a predictable, scheduled approach to station transmissions - avoiding the contention-based system used by Wi-Fi. So a WiMax AP could provide "real" QoS in the home, for things like video and VoIP.
But will it happen?
We're not placing bets on anything overcoming the Wi-Fi juggernaut in the near term. But the folks we spoke with at Cygnus were confident that WiMax LAN chipsets in volume could reach prices near those of 802.11. It certainly is worth watching as the WiMax market matures. Carriers still need to find a solution to in-home distribution the problem becomes more acute with every new high speed service offered. WiMax LANs may be one of the solutions.