As the CTIA Wireless show got underway in Las Vegas lastweek, there was lots of talk of emerging wireless technologies. CTIA is the primary trade organization for the cellular industry, so most of those discussions centeredonthe evolution of cellular data services, in particular the 4^th Generation cellular options. The two main contenders at this stage of the game are the 3G Partnership Project's Long Term Evolution (LTE), and WiMAX. The outcome will have a major impact on the range of wide area services available for mobile unified communications and fixed mobile convergence.

US cellular networks have been a standards mess since the introduction of digital cellular in the early 1990s. At that time, two different digital cellular standards were proposed: TDMA and CDMA. The TDMA networks eventually transitioned to GSM, the standard used for about 75% of the world's cell phones; GSM is used in AT&T’s and T-Mobile’s networks. CDMAwas developed by Qualcomm and used by Verizon, Sprint, as well as regional carriers Alltel and US Cellular.

Once the different groups started down their respective tracks, there was no diverting them. Eventually the industry focus shifted to data services,and each camp developed their own solutions. The GSM data progression went from GPRS, to EDGE, to UMTS, and now to High Speed Downlink Packet Access (HSDPA), or simply HSPA (because they’ll soon be adding “Uplink” as well). On the CDMA side, 1xRTT begat 1xEV-DO(Evolution- Data Optimized or Data Only- no one seems to be sure about that) that eventually evolved into 1xEV-DO Rev A.

The bottom line is that whether you're using the best GSM has to offer (HSDPA) or the best CDMA has to offer (1xEV-DO Rev A) your downstream data rate is about the same, 500 K to 700 Kbps, the low end of the ADSL range. The main difference between the two is the coverage area as the higher speed devices with revert to the earlier, lower-speed technologies if the high-speed service is not available.

True to form, as the industry got set for the migration to the fourth generation (4G), the GSM camp proposed a solution called Long Term Evolution (LTE) while the CDMA zealots pushed one called Ultra Mobile Broadband. At that point, a funny thing happed: Verizon broke ranks.

In a rather surprising move last year, Verizon, our largest CDMA carrier, announced that they would begin testing LTE (i.e. the GSM-proposed solution) as their 4G service, in essence joining forces with AT&T and the GSM camp. In the meantime, Sprint had declared WiMAX their 4G solution, but that decision was based largely on the fact that they owned a big chunk of the 2.5 GHz BRS spectrum that was being proposed for WiMAX. Eventually the standards groups defined Sprint's 4G WiMAX solutionas a "3G" technology, but that's the difference between marketing and engineering.

It does seem rather clear that UMB is becoming the orphan child in all this. One of the big discussion points is whether carriers should pursue LTE or develop networks that incorporate both LTE and WiMAX.

The Same, Only Different

Wireless data protocols address Layers 1 and 2 of the OSI model, and the amazing thing about LTE and WiMAX is how much they have in common, particularly at Layer 1. In a radio protocol, Layer 1 describes how the transmission (i.e. the 1's and 0's) are encoded and sent over the radio channel, and Layer 2 deals with addressing, message formatting, and how the users share the radio channel. Motorola recently announced a base station product that could support either WiMAX or LTE. They claim that roughly 75% of the technology is reusable.

The dual goals in the design of radio systems are bandwidth efficiency and reliability.Bandwidth efficiency defines how many bits per second you can transmit on one cycle of radio bandwidth (i.e. bits/second/Hertz). The fundamental limiting factor in all radio systems is how much radio spectrum (i.e. "bandwidth") you have available.Given that the radio spectrum is a limited resource, the more efficiently you can use it (i.e. the more bits per second you can transmit in it) the better off you are. The trade off is always that if you send at a higher bit rate, the signal will be less robust and suffer a higher incidence of transmission errors. There are other factors that come into play as well like the amount of power you lose at different frequencies, obstacles in the transmission path, the difficulties in communicating with a device that's moving, and so on.

The good news is that engineers have been getting better at the art of optimizing both the efficiency and the reliability. The three elements that we find in common with most digital radio interfaces today are Orthogonal Frequency Division Multiplexing (OFDM), Multiple Input-Multiple Output antennas, and forward error correction.The first two I described in an earlier article on 802.11n.. Forward error correction is the idea of adding additional redundant bits to the transmission, and the receiver then uses that redundant information to detect and then correct a certain percentage of the errors that occurred.

Virtually all of our modern radio links incorporate those three elements. That would include the 802.11n WLAN interface, WiMAX, and LTE. The 802.11 radio link standards provide a good view of how these new technologies impact bandwidth efficiency; the progression is summarized in the table below. The use of OFDM in the 802.11aand g interfaces resulted in roughly a five-fold increase in bandwidth efficiency, and the MIMO technology used in 802.11n nets an additional four-fold increase.

This type of mathematical analysis is somewhat contrived, as all of these devices use adaptive modulation, which means they reduce their efficiency in adverse conditions. While the LTE proponents claim that their technology can provide bandwidth efficiency up to 16.32 bits/sec/Hz versus a maximum of about 4.8 bits/sec/Hz for WiMAX, in reality they'll probably come out pretty close to equal.

Conclusion: So, Which is it Going to Be?

The funny thing about technology businesses is that technology counts for so little in the final result. Politics, human nature, and all of the frailties inherent therein trump the technology issues in virtually every case. The result: LTE wins. Both Verizon and AT&T said as much when they announced the plans for 700 MHz spectrum they acquired in the FCC’s Auction 73 last month.

The reason for that outcome is deceptively simply: the cellular carriers rule the roost, and they are more comfortable with technologies that are developed by vendors with whom they are familiar and whose judgment they trust. The cellular carriers control the market, so they control the decision. Now if the WiMAX camp could come up with something that was phenomenally better, then the cellular carrier would have to look more closely at this, but unfortunately, physics works the same for everyone. With no compelling technical advantage for WiMAX,LTE wins.