Hacker News Reader

Comment by singold | Hacker News Reader

"warns that sleeping next to your smartphone—the one that emits 3G and 4G signals all night—affects your brain patterns, restructuring your brain cells and likely preventing you from allowing your brain to clean out waste material properly."

Does anyone have any kind of evidence/study of this kind of effects?

View on HN

If you read the linked article, they find "...sensitivity to low-level radiation to a subtle degree...". Except they strapped cell phones to users' heads.

Given that cell-phone radiation falls off at a factor of 1/r^2, I would imagine the effects disappear very quickly as the cell phone gets further from your head. Also, the 2nd study used 10 sleep-deprived men. That's a pretty small sample size.

Thats what studies do, they use extremes to find clearer evidence.

10 is pretty small sample size by small sample size standards.

Yes it is, but that has nothing to do with what I said or what I replied to.

Emits signals all night is quite an exaggeration. Unless the phone is actively transmitting or receiving something, it only checks in with the network every once in a while.

I'm not sure about exact times in 3g/4g, but 2g (GSM at least) phones caused audible interference with poorly shielded audio equipment when they were transmitting. I recall an otherwise idle phone only checking in once an hour or so, though the interval was/is probably defined by the network, so that maybe different at different locations. Still, for power saving reasons, I expect an idle phone to just sit quiet most of the time.

I'm not certainly most phones are "otherwise idle" given push notifications, apps phoning home, etc. But, most of those devices are probably on wifi when their owners are at their homes asleep in their beds.

Phones turn off wifi whenever they can to save battery life. For example, Apple: "Notifications use Wi-Fi only when a cellular connection is unavailable." [1]

[1] https://support.apple.com/HT201925

I can send a text message to my phone and have it be received within a second or two, so my phone must be connecting with the cellular network at least once a second.

It wakes up around once a second to listen on the paging channel, but it doesn't transmit at that time.

It's funny how developers have completely lost the idea of real push messages, and ended up thinking that the client always have to initiate the connection.

> an otherwise idle phone only checking in once an hour or so

The phone network has to know which cell tower you're nearby in order to send a call to you. If you're driving, you could get far away from your original cell tower in one hour.

If your phone hasn't checked in for an hour and you've traveled a great distance in that hour, how does the phone network know where to send the incoming call for you?

I can think of only two solutions: Either your phone checks in much more frequently than once an hour, or the network tries pinging you on wider and wider concentric circles of cell towers in order to locate you. Which is it? (Or something else?)

The phone itself wakes up and listens for the base; if it finds it (and the ID matches) then it goes back to sleep. No need to transmit. If the ID doesn't match, it pings the new base to register and the network will inform the old base to drop that client. If it doesn't find the base at all it initiates a longer scan to find active bases on other frequencies. Once it finds the best one it has to re-negotiate and auth before the base will start accepting traffic on behalf of that client. Handoff bypasses the delays by having one base communicate to another to bypass the negotiations and the backhaul network automatically re-routes packets when this happens. The clients also get info about neighboring cells and towers which makes handoff easier and faster. As the baseband wakes up in its lowest power mode every minute or two it senses its connected base signal getting weaker while simultaneously one of its neighboring signals getting stronger and during one of these wake sessions it will elect to power up, transmit, and hand-off. Then it can go back into low-power monitoring mode until the next boundary. (This is an extremely abbreviated description of a complex process.)

The minimum ping time (let's say it is one hour) is just so the base can prune its client list.

When an SMS or call comes in the network doesn't broadcast that to all cells in the country. It knows where your device was last registered and sends it there even. If the base doesn't get an acknowledgement after a timeout period the SMS is dropped / the caller gets voicemail. If your device isn't near that location it should have handed-off on its own which means it is probably turned off or in an area with no signal.

As far as I know none of the current generation of cell systems bother to broadcast SMS, phone, or data packets to a wider area near your last known location.

Groups of cells are tagged with what is known as a "location area." The phone monitors a list of neighbor cells during its periodic wake ups (around 1-1.5 seconds) and reselects the strongest one. If the phone needs to reselect to a cell that is in a different location area, it sends a message to the network known as a "location area update." The network updates a database known as the HLR (home location register) that tracks the location area where the mobile last checked in. There is a policy called a "cell reselection hysteresis" that keeps it from ping-ponging between location areas if you are on the fringe of two.

The mobile also needs to send what are known as "periodic location area updates," to keep its entry up-to-date. Most networks I've seen have a 56 minute interval for that.

When the network needs to page a mobile, it sends the message to all cells in a location area.

Thankfully my cat sleeps on top of my head and shields from my mobile charging on the bedside table.

That's what your cat is programming you to think while it's on your head all night. You buy the expensive cat food now too, don't you...

He won't eat the cheap one. His food costs more than the tuna cans I eat.

This sounds like pseudo-science so I'd be extremely wary. Small study making bold claims based on an extremely subtle effect almost always turns out to be statistical noise or a flaw in the study.

Reply on news.ycombinator.com