Reihl sheds light on pair of white space device tests
Aug 19, 2008 11:35 AM
When the opportunity arose to chat with Edgar Reihl, Shure director of advanced development, about the Aug. 9 white space device test at FedExField in Washington, D.C., and the Aug. 12 test at the Majestic Theater in New York City, “HD Technology Update” jumped at the chance.
HD Technology Update: What were your general impressions of the white space prototype device tests at FedExField in Washington, D.C., Aug. 9?
Edgar Reihl: The white space devices performed poorly, which is consistent with what we had observed at earlier field tests conducted by the FCC. On the other hand, the test logistics went smoothly. We had excellent cooperation from the NFL, ESPN, the FCC and the Redskins. I’m very gratified with all of the hard work and assistance from all of those folks.
I have been observing the tests at the FCC Lab since January of this year. We’ve obviously had concerns all along that these devices were not going to do very well in a real-world environment. That’s part of the reason we asked the commission to do at least a couple of tests of this nature.
HD Technology Update: How many devices were tested, and can you tell me about them?
Edgar Reihl: Only two devices were tested on Saturday (Aug. 9) at FedExField and on Broadway in New York City this week. They were the Philips device and the device from I2R (Institute for Infocomm Research), which is a Singapore-based company.
They were selected because they are the only devices that purport to detect wireless microphones. In these tests, the FCC was seeking to determine how well white space devices detect microphones.
These are not smart radios. They don’t have any cognitive ability. They are basically radio scanners that are manually controlled by a PC interface, which you can command to scan a particular TV channel or a group of channels.
The device will perform a scan, and when it is finished it will identify a probability of detection or indicate whether a particular channel was in use. There are some differences between the Philips and the I2R box, but basically, they are both supposed to be able to detect television and wireless microphones. Once they do a scan, you get a readout of whether or not they found something on each TV channel.
HD Technology Update: Can you give me some more details about their performance at FedExField?
Edgar Reihl: The two devices had somewhat different problems. Outside the laboratory, the Philips device consistently exhibits a tendency to identify all of the TV channels as being in use. On a couple of occasions, it identified one or two channels as being open, but most of the time, it identifies every channel as being in use.
The problem with that is these devices are supposed to sense when something is in use or not. If it decides every channel is in use, it would never operate. As a result, there would be a great incentive for people to modify it or somehow defeat the protection, because the device wouldn’t do anything for you; it would never work.
On the other hand, the I2R device was never able to sense any of the wireless microphones. We saw this in some of the other field tests, like one that was conducted at one of the residences selected for DTV tests. It couldn’t detect a wireless microphone at a distance of 10ft. When we got out in the field, it couldn’t detect the presence of wireless mics either.
On top of that, it would routinely return different results — as far as which TV channels it found were occupied or not occupied. Without knowing more about the algorithm it uses, I’m not sure what would cause that particular problem, but apparently something about the way it does its TV detection causes it to yield inconsistent results. The I2R device definitely did not indicate every channel as being in use, but the list of occupied channels was different with each scan.
As you know, I am not from the TV industry, but if you were relying on that device to protect television from interference, it would not be successful for that either. As far as I am concerned, the device definitely did not detect the wireless microphones, even though the microphones used at FedExField were not the normal power level that most people use. They were SK250s, which are 250mW bodypacks operating at the maximum power allowed for Part 74 wireless microphones.
ESPN, which wasn’t covering the game, brought all of this gear down for the tests so they could do on-demand testing for the FCC. These bodypacks were not actually on bodies during these tests. The important thing to know is that when you wear a wireless bodypack, the human body actually attenuates the signal very significantly, typically on the order of 10db or 15dB. Had those been on the body of a person, the actual signal would have been considerably lower than it was for the test.
Even so, the devices were not detecting those microphones consistently or at all in some cases.
HD Technology Update: Can you describe the methodology for the test?
Edgar Reihl: We did a walk around prior to game day to identify a number of different locations in the stadium we thought would be useful places to test from. We started the testing on the 50-yard line on the south side of the field. The second location was the Tailgate Club, one level above the parking lot. The idea was to try to replicate the tailgate situation out in the parking lot.
For the third location, we originally planned to test from the crow’s nest, on the south side of the stadium, but we ended up at a slightly lower level because it would have been too difficult to get the gear up to the top. We wanted a location that was high up in the stadium for two reasons. First, it would be a fairly long distance away from the microphones, and second, one of things that makes the field tests realistic is that you have a fairly high ambient environment from the DTV signals present. The fourth location was in the press box.
We got started with the testing at about 10:30 a.m. The equipment itself all basically fit on a single Rubbermaid non-metallic cart that the FCC staff brought with them. They had a high-performance spectrum analyzer with a preamp and a monopole antenna with a ground plate that they would set up at each location.
The first thing they did at each location was a background scan, which was recorded digitally on the analyzer. Next, they would turn on one of the prototype devices — either the I2R or the Philips. During the testing, they require any other equipment to be switched off. So the analyzer was turned off and the other prototype device was turned off. Only the white space prototype being tested was powered up.
The Philips device is in a small, black plastic box. It basically contains the innards of a TV tuner and is controlled by a desktop PC, which was on the cart. There’s a user interface written by Philips that displays at the end of each scan what channels it found. The I2R box is a larger black metal box. It was a little more finicky in that it has to warm up for a certain period before use. This one was controlled by a laptop. The output came out as a text file with probabilities of detection at each TV channel.
As soon as they would do a scan, the FCC engineer would read off a list of channels it had found occupied. Neither of these devices was configured to tell you what it found. They would just say whether or not a channel was occupied.
After they finished the testing of each device with the ambient environment, we would radio over to the ESPN folks to switch on their microphones. They had a number of microphones set up in various TV channels that the FCC staff had identified as being free and clear of any ambient signals.
They would bring those up, and the FCC would repeat their scans on their analyzer. Then they would test each of the white space devices, having them scan to see what they detected. Obviously, because the Philips device normally finds all of the channels occupied, when you switch the microphones on, the results were not noticeably different, but for completeness, the commission did test every device in every location.
HD Technology Update: What was the proximity of the prototypes to the wireless mics ESPN turned on?
Edgar Reihl: When we were on the sidelines, it was the width of the football field. That was the closest situation. The farthest was on the upper deck below the crow’s nest, and that would have been at least twice as far away. But all were clear, unobstructed paths, even to the press box. These were all unobstructed straight-line shots.
HD Technology Update: Can you briefly describe the tests conducted this week on Broadway of these prototypes?
Edgar Reihl: The same devices were tested Aug. 12 on Broadway. In New York, it’s quite a bit different situation, because for the most part, we were indoors, with the exception of one test in which the antenna was sitting on the sidewalk outside the theater.
These tests all occurred at the Majestic Theater, which is a Broadway theater owned by the Shubert Organization. It’s in the Times Square district. It’s a fairly high RF field environment, but I’m not sure it’s as strong as what we had out at FedExField because we were down at street level. I don’t think the RF levels there from the DTV signals were as high. The rest of the tests were run inside the theater.
HD Technology Update: What were the results in New York?
Edgar Reihl: Very similar. In almost all cases, the Philips device showed all channels occupied. There was one location where it showed two channels unoccupied — Channel 32 and Channel 37, I believe. I should say that most of the time, the Philips device shows Channel 37 as occupied when of course it is never occupied, because it is reserved for radio astronomy. It’s also used by very low-power medical telemetry devices inside medical institutions, but we weren’t in any of those areas.
In one of the tests with the device closest to the microphones (from the 10th row of the theater), it showed Channel 32 and 37 open with the microphones switched off. When they switched them on, it showed both of those channels in use, but we didn’t have microphones operating on either one of those channels.
The I2R device never indicated any occupied channels on any of the channels where microphones were operated. It did show some of the TV channels in use. But it would show one list of channels during a scan with the microphones switched off and a slightly different set with them switched on. This kind of inconsistency is something we saw in the other tests that we observed as well.
HD Technology Update: Although you are coming at this issue from the wireless mic perspective, you share concerns with broadcasters about the introduction of these devices into the TV band and the potential for interference. Do you have any thoughts about the DTV side of this issue?
Edgar Reihl: We share many of the same concerns. Obviously, ours is more on the content side because the wireless mics are critical to producing the content. But broadcasters obviously have immediate concerns about interference with over the air.
There’s another area of concern as well. That’s ingress into cable. We have informally done some testing with ingress into cable television reception and found there’s potentially a very serious problem with that, especially when these devices are used in the consumer’s home in close proximity to a television connected to a cable system. The main thing is if you are using a system that relies on sensing to detect a signal. You can’t sense the presence of the cable channel, because it’s not on the same frequency. I have observed one test in the living room of a member of the FCC Office of Engineering and Technology. They switched it on, and it completely froze the TV picture with a few milliwatts of power.