WiMAX: Where and Where Not?Tuesday, April 03, 2007
Don’t know what FDD and TDD mean? Don’t worry, read on and you will understand!
TDD stands for Time Division Duplexing; FDD stands for Frequency Division Duplexing.
At our recent Wireless Data University (renamed Andrew Seybold University, School of Wireless Mobility), I talked about WiMAX, where it fits and what spectrum it will be deployed on. I stated that it would not be deployed on the new AWS spectrum that was just auctioned, nor on the 700-MHz spectrum that will be auctioned later this year or early next year, because WiMAX is a TDD technology and the AWS and 700-MHz bands are reserved for FDD technologies.
I was questioned by an attendee about this statement, but fortunately for me, Dr. Mo Shakouri from Alverion, a WiMAX equipment vendor with a lot of product already in the marketplace, was a speaker and after his discussion of WiMAX and where it fits, I asked him the question about TDD and FDD and WiMAX. His answer was that today, WiMAX is a TDD-only technology and that it would be at least two years before there was an FDD flavor.
So what does all of this mean? What is the difference and why does it matter? Let’s start at the beginning with two-way radio systems that were developed in two ways. The first was when the base station and mobile radio transmitted on the same channel. This is known as simplex operation (or, in first responder speak, these channels are known as tactical channels). This type of radio system worked, but range was limited, so a system was developed where the base station transmits on one channel and the mobile unit on a second channel. The mobile unit’s transmission is then received by the base station and rebroadcast so all of the other units can hear both sides of the conversation. A later refinement of this technique was developed and it is known as trunking.
There are many advantages to having the base station transmit on one channel and the mobile units on another, and when the first cellular licenses were granted, this is the way the 850-MHz band was allocated. Both A and B block operators transmit in the upper portion of the band and the mobile units transmit in the lower portion. One reason for this is to minimize interference between operators. Since all of the high-power base (cell site) transmitters transmit on the same portion of the band and the receivers at the cell site receive on a different portion, the operators are less likely to cause each other interference.
This becomes even more important when you look at the PCS (1900-MHz) spectrum because now there are more than two license holders in most cities and four or five different operators can be sharing the same band. PCS channels were also allocated for one portion of the band for base station transmitters and one portion for mobile transmitters. The AWS and 700-MHz bands have also been allocated in this way.
WiMAX has been designed for both transmitters to operate in the same portion of the band on a simultaneous basis. This works fine, I do not have a problem with the technology, which has been in use for a long time. Frequency Division Duplexing (FDD) and Time Division Duplexing (TDD) both work well EXCEPT when you try to mix them. This is the issue for WiMAX with the AWS and 700-MHz spectrum.
Let’s look at an example. In a given city, there could be six AWS license holders, all with either 5- or 10-MHz blocks of spectrum. They will build their systems so all of the base station transmitters use the upper portion of the AWS spectrum and the mobile units transmit in the lower portion. This makes good engineering sense and it is relatively easy to control interference from one operator to another.
However, if one of the six (for example) license holders decided to deploy WiMAX in the spectrum, it would use either half of its allocation for WiMAX, or perhaps both halves, but it would be transmitting from both the base station and the mobile units in each half. This means WiMAX could cause interference to the license holders of the technologies deployed as FDD systems, and the FDD systems could easily interfere with the WiMAX system even though all of the network equipment for both types of systems would be fully compliant with FCC rules and regulations.
Is this clear yet? It can be confusing to those not familiar with wireless or radio frequency systems, and many of you probably don’t need to know all of this, but I thought it worth discussing because I have heard several AWS license holders talking about WiMAX as being under consideration for their system rollout. The allocations in the 2.3-, 2.5- and 3.5-GHz bands are all TDD in nature, that is, the 10, 20 or more MHz of spectrum per license is not mated with another 10, 20 or more MHz of spectrum for the mobile side of the system.
Just to confuse you more (I hope not), a license in the 2.5-GHz band for 20 MHz of spectrum is just that, 20 MHz. But when an AWS, PCS, or cellular operator has a license for a 10-MHz block of spectrum, it really means it has 10 MHz for the base stations (cell sites) and 10 MHz for mobile operation for a total of 20 MHz. Got it?
When you hear about the 30 or 60 MHz of spectrum that will be auctioned in the 700-MHz band now occupied by TV channels 60-69, in reality, there is 60 MHz of spectrum available. However, since it is allocated for FDD operation, each winner will be licensed, for example, for a 10-MHz chunk at the high end of the band and 10-MHz chunk at the low end for a total of 20 MHz.
One reason that the use of FDD systems becomes more important over time is the fact that as more sites are built to handle new demand, and better coverage, many cities and counties are requiring cell site sharing by multiple wireless operators. Today, this works out well. Let’s say that Sprint, AT&T and Verizon Wireless all share the same cell site and all three are using the PCS or 1900-MHz band for that cell site.
All of the transmitters installed at the site are transmitting in the upper portion of the PCS band and all of the receivers are listening for mobile traffic in the lower portion of the band. No matter how good the transmitters are, there is the potential that they will cause interference outside of their licensed spectrum. If AT&T had a transmitter and Sprint had a receiver in the same portion of spectrum, it is possible that the AT&T transmitter would interfere with the Sprint receiver or that Sprint would have to spend a lot of extra time, effort and money installing filters to keep out the stray signals. If you add to this the fact that AT&T and Sprint use different technologies, one based on spread spectrum and one based on time division technology, there is an even greater chance of interference.
By keeping all of the receivers in one portion of the band and all of the transmitters in another, this potential for interference is minimized. This is a major reason that Nextel and the first responder community are going through what is being called a rebanding process, moving radio systems in the 800-MHz band so Nextel cell sites no longer interfere with first responder radios.
Does this mean WiMAX is bad and CDMA, GSM and UMTS are good? No, it simply means that WiMAX, as well as technologies from IP Wireless and others, use the same portion of the spectrum for both base and mobile transmission and that CDMA, GSM and UMTS use one portion of the spectrum for base transmissions and another for mobile transmissions. Both of these techniques work well, unless you try to mix them in the same portion of spectrum, in which case you could end up with interference issues that would have to be resolved and this would add cost to the systems. This extra expense can be avoided if the type of technology is matched to the spectrum allocated for it.
I have said it before, and I will repeat myself and others. The use of radio frequency technologies is part science and part black magic, which is one reason most people who are not directly involved in the technology aspects of the wireless industry don’t have an appreciation for the many challenges that have to be overcome for our systems to work as well as they do. There are laws of physics and we have bent some of them, but no one is about to break any of them!
Andrew M. Seybold