More Examples of Interference From Unlicensed Devices
February 21, 2007
The IEEE Consumer Electronics Society celebrated its 75th anniversary at the International Conference on Consumer Electronics in January, along with the Consumer Electronics Show in Las Vegas.
| Fig. 1|
| Fig. 2|
| Fig. 3|
| Fig. 4|
| Fig. 5|
| Fig. 6|
I presented two technical papers: "New Challenges to Designers of DTV Receivers Concerning Interference" deals with the recent FCC actions to allow unlicensed transmitters to operate in terrestrial broadcast spectrum. Readers will recall my dire warnings about a resulting digital citizens band arising Feb. 18, 2009 when the sharing begins.
Broadcasters and DTV designers need to be aware that while the strongest DTV signal--desired (D) or undesired (U)--will be about -5 dBm at the receiver input from a transmitter five miles distant, much stronger signals may be received from unlicensed transmitters operating in first adjacent channels. In other words, the protection ratios the FCC developed from for DTV-DTV interference will not protect viewers from this new kind of interference.
Those protection factors took into account the location of DTV transmitters, but the location of digital citizen band transmitters will be unknown. DTV receivers will be subject to additional forms of legal interference. If it is legal, it cannot be "harmful," or can it?
The accompanying figures here deal with one, two or three undesired 6 MHz signals on TV channels, all at the same power. There are no desired signals shown because I want to illustrate the intermodulation--or IM--products these undesired signals produce in an overloaded receiver. Each signal is close to 0 dBm, so they produce lots of IM and spread it over a number of channels.
Fig. 1 shows one signal filling channel (n). This could be a DTV signal or an unlicensed signal. Note that the IM spills over into both (n-1) and (n+1) where the 'noise level' is -43 and -45 dBm.
Fig. 2 shows the same signal on both channels (n-1) and (n+1). Note that there is IM on channels (n-4) and (n+4) at -40 dBm, and more IM on channels (n-3), (n-2), (n), (n+2) and (n+3) each at about -34 dBm. This is spectrum spreading of the IM.
This has been shown in previous columns and in a paper I co-authored with Gary Sgrignoli in "IEEE Transactions on Consumer Electronics" May, 2005. This spectrum spreading is being studied more closely here.
Fig. 2 demonstrates that a pair of strong signals on two channels can cause interference to reception on any of seven channels nearby. Either or both of these can be a DTV signal, or from an unlicensed transmitter.
Broadcasters may be unconcerned about their digital coverage now, but that may change in 2009 when those will be the only signals in operation. All these channels can be jammed by two unlicensed transmitters.
As this column has noted, the IEEE Working Group 802.22, which is working on a protocol for unlicensed transmitters, does not think these devices should operate on a first adjacent channels. However, the FCC doesn't have to agree.
Opening first adjacent channels would be impractical not merely from the DTV interference standpoint, but because of radiated IM. Think of it this way, a DTV transmitter may radiate up to 1 megawatt (30 dB above 1 kW) of power in its allocated UHF channel. It is also radiating about 36 dB less power in each first adjacent channel, so the power in each adjacent channel is 6 dB below 1 kilowatt.
How can a base station with an ERP or 2.4 watts serving as an entry point to the Internet be heard over the noise in the channel? How can anyone with an unlicensed transmitter be heard in this community?
Fig. 3 shows a pair of signals with two channels between them. The familiar IM on both first adjacent channels is seen, but now there are clusters of other IM products below the lower channel and above the higher one. These are IM products generated between the two signals. If either signal is removed, both of the extra cluster vanish, too.
Suppose there is a DTV signal on one of these channels, and an unlicensed transmitter signal on the other. Spectrum spreading will produce a plot as in Fig. 3, and any of six TV channels may be jammed by this combination of signals. Two unlicensed transmissions could produce the same results. Note also that in Fig. 4, there are vacant channels between the occupied channels. Note also the increased spectrum spreading in Fig. 4. Ten channels have IM.
It is convenient to think of pairs of signals causing interference, but there are many cases where three or more U signals can intermodulate as shown in Fig. 5.
The IM is produced in the receivers. If the RF amplifier gain were reduced by 20 dB, or the RF signal attenuated by 20 dB, these IM products would vanish as shown in Fig. 6. All three signals could be received if the overloading shown in Fig. 5 ceases.
I have suggested that the low VHF band be re-allocated to unlicensed transmitters. By isolating these devices in a band of their own, the interference problems would be solved. As matters stand, those 40 or so broadcasters seeking a DTV allotment in the low VHF band are in for a shock when they find out how little coverage they actually have.
Their only solution is to request a substantial increase in power. If they get this power increase, whatever economic advantage they sought in seeking a low VHF channel will be lost. Has anyone told them? Will they get their power increase? Why not put it to the FCC?
Then there is the matter of broadband over power lines. I sure wouldn't bet the BPL threat to the low VHF band will simply go away. If BPL doesn't interfere with DTV reception because the signal is carried conductively over power lines, why should it be incompatible with unlicensed transmitters radiating elsewhere in the low VHF band? Perhaps the BPL folks should populate the lower portion of the band first, and unlicensed transmitters from the top down.
Recently, I've been studying ways to design receivers to operate with up to -5 dBm of U signal power present. I think I've come up with a way to do it and that will be the topic of one of my ICCE papers. My concept kicks in at D = -48 dBm. What about the viewers in the DTV fringe area?
Viewers receiving DTV at -70 dBm are now protected against adjacent channel interference by the FCC DTV planning factors. If a station is received at -70 dBm on channel (n) inside the station's coverage area, there won't be any other DTV signal on channels n-1 or n+1 above -42 dBm (D/U = -28 dB).
But what about one or more unlicensed transmitters on the same block, which may put -14 dBm into nearby receivers. What about more than one unlicensed transmitter in the neighborhood? I see no way to avoid jamming in weak signal areas where the RF amplifier must run at maximum gain to capture the desired signal more than 15 dB above the noise generated in the receiver's mixer.
My other topic was "Non-Invasive Testing Methods to Determine the RF Performance of Consumer DTV Receiving Appliances."
My plan involves testing with two undesired DTV signals. My undesired DTV signals include sideband splatter akin to that of real DTV transmitters. My desired DTV signal is a real-life signal received with a log periodic antenna over a 13-mile, direct-ray path and amplified up to -5 dBm without measurable sideband splatter.
These tests cover the entire range of desired signals from -84 to -5 dBm measured in 5 or 10 dB increments. Once it is determined how two undesired DTV signals affect a receiver, further testing of one specific undesired signal can be readily carried out.
Charlie Rhodes is a consultant in the field of television broadcast technologies and planning. He can be reached via e-mail at firstname.lastname@example.org.