Wednesday, August 14, 2019

Starting to Look at 5G Cellular

In the academic sense, I'm not physically looking at a 5G link.  Smallville here will probably not get 5G for years and years.  I don't even get a particularly good 4G/LTE signal here at home.  Still, I hear the buzz and it makes me want to understand what the story is.

There's lots of buzz about 5G for the new technologies it's going to bring, but those largely come down to higher bandwidths and faster downloads for the users.  Numbers like 10 Gigabits/second get thrown around - orders of magnitude faster than 4G.  To the network providers, the attraction is that the number of subscriber devices they can support (sell) jumps up dramatically.  Just as 4G is faster than 3G was and every generation before it.  Despite the fact that 5G services won't roll out anywhere until - maybe - the end of calendar 2019, the buzz seems to center on the idea that whoever controls the networks will know everything about every business and every person in the country.  Why wait?  Avoid the rush and just get one of those "OK Google" or Alexa monitors to bug your own house now.  This security seems more like an encryption issue rather than whose base station it goes to.  No, I don't trust Huawei and the Chinese, but I don't trust the US either. 

That part about security comes from what's actually a completely unrelated technology to 5G, the drive to use internet connectivity to improve services of all kinds, call the Internet of Things, usually referred to as the IOT, or (as I like to call it) the Internet of Things That Just Don't Quite Work Right (IOTTJDQWR - which doesn't quite roll off the tongue as easily).  The IOT and it's cousin, the Industrial IOT, will do simple things like put water pumps, flood gates, and all sorts of civic and utility infrastructure on the net.  The vast majority of these things are not high bandwidth users and don't need the faster data rates that 5G promises, although the lower latency/faster network might be useful.  They could be done today, and many are done over industrial radio links. 

The promise of higher data rates brings the inevitable trade that higher rates usually require higher bandwidths and higher bandwidths usually require higher operating frequencies.  This also scares a lot of people.  Back to that in a minute, but first, let me show you a list of frequencies that 5G can use - not all of these will be in any particular device.  The list is from a document from Keysight, the former Hewlett Packard test equipment group, and still arguably the industry leader in this sort of test equipment. 

It's a bit of an eye chart if you don't enlarge it, but two things to notice are that the new frequencies are all printed in a greenish/teal sort of color; and that frequencies currently used for other purposes that have been reassigned for 5G (or dual purpose, it's not completely clear) are in purple.  Those in black, in the second vertical block, remain for LTE of various flavors.  The only new frequencies capable of high bandwidth, that are higher than the existing LTE services (virtually all below 3 Gigahertz - 3 GHz) are in the top right block and are in a few bands from 24 to 40 GHz.  Why not 60 GHz, where WiGig is currently all alone?  This chart shows why:

This shows the attenuation (loss of signal) due to dry air (in red), water vapor (blue) or the sum of the two losses (in black). You will see a huge peak of attenuation centered right on 60 GHz, going from around 0.3 dB/kilometer up to (guessing) 15 dB.  That means a transmitter on that peak has to put out almost 16 dB more power, or 40 times more power, than the lower frequency transmitter.  That huge a power difference means much more expensive hardware.  It turns out that dip in attenuation you see between the huge peak at 60 and the much smaller peak at about 22 GHz covers the 24 to 40 GHz range for the 5G signals. 

One thing that people don't seem to understand is that it's hard to get radio frequency energy into places you want it.  Even in a controlled, lab environment, on well designed hardware, we regularly fight to get thousandths of a watt that we expect to get but that nature just doesn't want to give up.  Furthermore, the higher the frequency, the more struggles that come. 

In the case of the 5G network, those struggles will be because the signals don't penetrate walls and other things as well as the lower 4G frequencies do.  It's going to call for more towers more closely spaced and fancier antennas.  The buzzword is MIMO - Multiple Input Multiple Output - it's a way of combining antennas to get better performance out of the network.  If you have a WiFi router with two or more antennas on it, you have that now. 

Is 5G going to be dangerous for people carrying handsets or walking around the city?  The power density requirement in FCC-OET Bulletin 56 is the same as for frequencies we've talked about here before, 1 milliwatt per square centimeter.  Industry is experienced with designing for these levels and I don't see any reason why it should be any more dangerous than existing networks.  If anything, because 24 to 40 GHz doesn't penetrate as well as lower frequencies, it might be safer.

I've written many times on RF safety and can do so again.


  1. I'm not totally convinced that the constant low level RF exposure from cell phones is safe in the long run. It is amazingly difficult to study long term risks to such exposures in the real world. RF, like ionizing radiation is a proven risk at high levels but we truly don't know the threshold exposure at which the risk moves from theoretical to measurable.

    1. I am. Totally convinced that constant low level RF exposure is safe, that is.

      The reason I'm convinced is there is no known mechanism by which radio emissions can cause cell damage except heating, and if the exposure doesn't cause heating, there's no damage. Just like light, and radio is just low frequency light. Ionizing radiation is orders of magnitude more energetic than light: UV or visible. That's why it's dangerous.

      If light is strong enough to burn it's dangerous too; and if it's not that strong, it's safe to be exposed to. Same as radio. You can burn things with a very strong light like a laser or you can burn with daylight by collecting light over a large area and focusing it - like the sun and magnifying glass. Of course, around here in Florida, standing outside in the sun can cause you to spontaneously combust but that's also just heat.

      Sunburn is caused by higher energy, higher frequency light, but the same rule applies - that damage depends on power the cells are exposed to. In radio, the dangers vary with frequency and that's encoded in the exposure standards. Fun fact: the most dangerous frequencies for people are in the FM broadcast band, because most adults are about a quarter wavelength tall in that band and will absorb the most energy. You don't hear complaints about that.

      As you point out, the kind of study you'd need is just about impossible to do with free-living humans. How do you separate every other thing those people were exposed to over a lifetime to figure out what caused the illness? You'd need a very controlled exposure and a population living in metal boxes their entire lives. Since that's both unethical and ferociously expensive to do with people, it has been done over and over with animals. Nothing has been found. There have been odd studies that show other damages (other than from heat) that aren't duplicated.

      A good rule of thumb in statistics is that if something hasn't shown up in multiple studies, it's either not there or it's a very weak effect. Very weak effects show up better in larger groups. When larger numbers of animals are used, the effect doesn't stand out.

    2. Yes....ionizing radiation is much more energetic than RF...I play with it every day. Yet 125 years ago we didn't even know it existed. In the 50's it was safe and useful.... Ordinary people used it to size shoes and treat acne. We now know how amazingly wrong we were about ionizing radiation. It is entirely possible that we are at the same level of pseudo knowledge regarding LONG TERM CHRONIC RF exposure. We may very well find in another half century that the increase in formerly rare medical conditions is due to
      cellphones and other pervasive RF energies we bathe in 24/7.

    3. Well, the stupidest thing anyone could say is that we know everything there is to know, so I won't go there. I do think the probability of discovering some totally new phenomenon is pretty small. The major advantage we have over 125 years ago is 125 more years of science, and more people investigating the world freely. 125 more years of "shoulders of giants" to stand on.

      Remember the axiom that "the dose makes the poison?" That's what the requirements of power density work out to be. The investigators find a dosage that does no detectable damage and then set a safety margin on it. If you read the specs, they set a lower margin (higher exposure) for people who work in the field, assuming they'll be more careful about exposure than the general public. Some countries have set lower power limits simply on the precautionary principle. There's no data forcing them to go for lower limits; the simply think that going to a larger number of nines in the chances that no one gets hurt is better. There's always a balance between the uncertainty and the cost.

  2. the buzz seems to center on the idea that whoever controls the networks will know everything about every business and every person in the country

    The government has tanks and nukes and supreme courts and printing presses and the populace can't fight them and win REEEEEEEEEEEEEEEEEE

  3. Louisville Ky. has had 5G for more than a year. The antiquated infrastructure won't support it. It runs at something less than 50 to 90 MPS IF it is having a good day. Sometimes it works. Sometimes upload is dial up slow (I have seen test speeds on my system under 6 MPS upload) Simply put. The government can DEMAND all the control it wants to. Is it going to spend the BILLIONS needed to rebuild the out of date infrastructure? I Think not. I live in one of the bedroom community's around Louisville. They are working around the clock to get the new wire up on poles. First Thunder storm or tornado, and the whole region is gone for a year or more. Too many buy outs. Too many tax breaks. Too many users on an already over loaded 20++ year old infrastructure. Control everybody with the internet. Yeh. Right.---Ray

  4. The bodily mechanism which signals for the maintenance of bone mass is piezoelectric voltages generated by bones bending. That's why bone regrowth after injury can be medically stimulated by externally applying small voltages.

    How many "accidental" rectifiers are present in the body, which would generate small voltages in an RF field? What is the signaling modulation which controls bone regrowth? What modulations would be accidental security penetrations which activate bodily mechanisms we don't understand? Humans didn't evolve to resist "random" inputs from a complicated RF environment. Selectivity to resist undesired rectified RF inputs has not been improved by a long duration evolutionary pressure.

    The "thermal bulk heating only" idea is wishful thinking. I suspect the short term risk of RF fields is minimal, but the long-term risk could be alarmingly large.

    1. The bodily mechanism which signals for the maintenance of bone mass is piezoelectric voltages generated by bones bending. That's why bone regrowth after injury can be medically stimulated by externally applying small voltages.

      It's also why baseball pitchers have larger bones in their throwing arms than their other arms. As well as being a perfect example of the idea that "the dose makes the poison" that's embedded in the RF safety standards. If those bones get bigger in a pitcher, or in therapy by applying voltages, what about the smaller voltages or throwing 50 mph instead of 95? If applying one volt causes bone growth, what a microvolt? A nanovolt? Is there a threshold below which bones don't change? My guess would be there is.

      But still, as always, got any data to show? All I'm saying is that before I retired I read studies from the US, the EU, Australia, Japan and other countries. Nobody had any proof of other effects other than heating, in lab animals (in vivo) or in cell cultures (in vitro) experiments. Odd things show up from time to time but the experiments don't replicate. Is it contamination or other issues with the experiments or is it a weak effect that shows only rarely? Bigger experiments should allow weaker effects to show up.

      I worked with a Ph.D. who had managed the radio side of some of those experiments, too. People who work in radio aren't typically suicidal and look into what seems to be dangerous.

      If food science was one tenth as good as radio science, we wouldn't be as messed up as we are.

    2. If a study put the 100 MHz signal from a rock music station into an ethernet card, it would never accidentally input correct ethernet frames. The study would conclude the ethernet card was not sensitive to voltage, and only subject to bulk heating effects. Those RF safety studies have the same problem of using the wrong protocols, and concluding the body isn't listening. Nobody knows if a slight variant of an existing bodily electrical protocol is a poison. What's the RF version of Roundup for humans, a fake regulatory input that kills? I mean, other than television news.

    3. Do you honestly think that the only thing that has been studied is bulk heating effects? That they haven't been searching for other effects?

      I have some stuff here from some various agencies that I'll extract and post later.

    4. Yes. I may be factually wrong and I would be happy to learn otherwise, but, yes, that is what I believe today. I believe the government positions on the risk from 60 Hz, cell phone, WiFi, overhead airplanes, coal power, nuclear power, automobile exhaust, fluoride in the water, rare violent SSRI side effects, non-thermal lighting spectrum, sugars, starchs, fats, meats, cholesterol, statins, diabetes, debt, democracy; anything that would disrupt the status quo, is as scientifically bankrupt as analysis of HotColdWetDry.

      What about flashing lights triggering epilepsy? I know a person who gets migranes from the visual flickering of fluorescent lights, which tells me at least one person has a 120 Hz frequency sensitivity big enough to notice. Maybe the 120 Hz frequency sensitivity is only sub-clinical for most of us because it's hidden in the noise from the bath of anxiety-producing factors. Why is the human birth rate in the first world low despite that we finally can feed and support lots of children? It doesn't seem to work that way for any other mammal.

    5. Although you say that, " I may be factually wrong and I would be happy to learn otherwise," I get the feeling that there's no evidence that could be presented that would mean anything to you.

    6. When you're troubleshooting an intermittent electronics problem on the bench, you don't decide you're on the track of the problem until you can turn the flaw on and off at will. I push on the circuit board here, it stops working; I release it, it starts working again. Ten times in a row, still does it after lunch break, and so on.

      I believe the studies you're quoting don't demonstrate an RF emission that causes for example bone growth, then reduce emission power until the bone growth stops, then declare a safety factor of 100, then say that's the maximal safe level? Do the complete set of studies as a group prove maximum external inputs which causes no detectable effect for every known electric communication mechanism in the body? Instead, I believe the studies are a circular argument. First the study writer assumes bulk heating is the only mechanism, then he demonstrates what level is irrelevantly small bulk heating.

      I think the analogy to Roundup is a good one. What RF input is a poison to a signaling mechanism, malware, whose tiny dose acts all out of proportion to its size because the body's signaling mechanism provides the gain? What RF input causes a big effect like a tiny amount of jellyfish sting does?