[Originally appeared in December 2007 print edition of Mobile Enterprise Magazine]
The United States Federal Communications Commission will auction spectrum in the 700-MHz band in January 2008.
This is called beachfront spectrum because 700-MHz means fewer towers for the same coverage as existing systems and better penetration into buildings and other structures. Yet, it's close enough to existing spectrum to be included in multiple-band devices.
We won't delve into the politics of the 700-MHz auctions here. However, the outcome is clear-the spectrum will be available about the same time that two of the three next-generation mobile technologies are ready for commercialization.
Thus, we'll likely see this new technology first in the 700- MHz spectrum. These networks will be IP-based end-to-end and will be built out using next-generation technology from the very beginning.
What are the 4th Generation technologies that will be jockeying for position? And will they make an appreciable difference in data speeds?
There are at least three contenders for next-generation, or 4G, wireless: Long- Term Evolution (LTE), Ultra-Mobile Broadband (UMB), and WiMAXMobile, an all-new version of the WiMAX being deployed today.
Amultitude of standards is nothing new. There are three different standards for our third-generation networks: UMTS/HSPA, EV-DO Rev 0 and Rev A, and WiMAX. Not only that, but there are two versions of WiMAX, although it appears that WiMAX-Mobile will be the most popular. This is the version being built out on the Sprint/Clearwire network in the United States. While it's called WiMAX-Mobile, it also will be used for point-to-point and point-to-multipoint services.
The fourth-generation LTE technology comes to us from the GSM/UMTS community by way of the 3rd Generation Partnership Project (3GPP).
It's considered the likely migration path for networks that have deployed UMTS/HSPA. UMB is being developed by 3GPP2, a sister organization to 3GPP that develops standards on the CDMA side of the technology fence. The WiMAX Forum and IEEE are developing WiMAX-M.
According to the 3GPP, LTE could be deployed as early as 2010. UMB, which is further along in the process, could be ready for commercialization in 2009. We don't yet have a date for the availability of WiMAX-M.
The Bandwidth Claims
Both LTE and UMB are designed to use a number of different bandwidths, starting at the current CDMA channel width of 1.25 MHz all the way up to 20 MHz of bandwidth. There are a couple of reasons for this wide range:
- Spectrum allocations around the world are different and there needs to be flexibility in how much bandwidth a new technology requires;
- Higher bandwidth provides for more data capacity and faster rates.
If you're evaluating data speeds, it's essential to compare these formats in the same amount of bandwidth. For example, in comparing existing 3G formats, WiMAX proponents present today's WiMAX as better than either EV-DO Rev A or UMTS/HSPA.
They base these claims on a larger amount of bandwidth. CDMA EV-DO operates in a 1.25-MHz channel (called a carrier) and UMTS/HSPA requires 5 MHz of bandwidth. Usually when comparing the three technologies, the WiMAX side uses 8 MHz or more of bandwidth to achieve its "better than" UMTS and EVDO speed and capacity measurements.
Our own comparison is based on 10 MHz of spectrum (see Figure 1). In 10 MHz of spectrum, you can have two UMTS/HSPA carriers, seven EV-DO carriers or one WiMAX carrier. When you compare the three technologies using these metrics, you see that WiMAX is equal to, but no better than, either of the other two technologies.
Other factors to keep in mind when evaluating data speeds:
- Stated speeds and capacities usually have been determined in a laboratory setting, and represent the maximum speeds obtainable by a given technology. Real-world numbers will always be lower.
- Distance is a factor. The further you are from the access point, the slower your data speed will be.
- Wireless bandwidth is shared.
Consider a hotspot at Starbucks. If you're the only person using Wi-Fi in the cafe, you'll have the full capabilities of the Wi-Fi speed all to yourself. The speed will only be limited by the data rate of the transport relaying your data from the access point up to the Internet.
However, if five or six customers are using Wi-Fi simultaneously in the same Starbucks, you'll share bandwidth, and data speeds will be lower.
And that Starbucks access point isn't discrete. Each cell site is usually divided into sectors, and total bandwidth is available on a sector-by-sector basis. Since cell sites generally cover larger areas, you'll be sharing the total available bandwidth with far more customers than the ones sipping lattes around you in Starbucks.
What does all this mean with regard to next-generation technologies? The bottom line is that 4th-generation technologies will improve data speeds. But our actual user experience won't be anywhere near what the engineers claim are the optimum capabilities of the given technologies.
Still, the Still, the next generation technologies (at least LTE and UMB) will provide much faster speeds than 3G for both sending and receiving data.
As an example, if you refer to figure 1, you'll see that today's technologies are providing data rates in the 1 Mb-2 Mbps maximum range.
Both LTE and UMB should provide data rates in the 30 Mbps range down to a device, and about 15 Mbps back up to the system. This is based on 10 MHz of bandwidth. The peak numbers for LTE and UMB are in the 100- Mbps-plus range-again, based on using 10 MHz of bandwidth (see Figure 2).
Keep in mind that there are also additional data speed increases coming to 3G technologies such as CDMA EV-DO and UMTS/HSPA.
With the rollout of 4G services, we'll have next-generation, 3G and 2.5G data rates available across a single network. 4G will be implemented first where data demand is greatest: near airports, hotels, convention centers and centers of business. As more sites add next-generation capabilities, and Voice over IP (VoIP) is implemented across these networks, there will be a natural transition from 2.5 and 3G to next-generation services, though this transition will take a number of years.
When it comes to wireless bandwidth and higher data speeds, laws of physics come into play and there is a theoretical maximum data rate per Hz of spectrum (Shannon's Law). While we can't break the laws of physics, the wireless engineering community is certainly becoming adept at bending them a little.