# how much power is radiated by cell towers?

I want to know how much power is radiated by cell towers of GSM (1.8 Ghz), 3G (2.1 Ghz), 4G (2.6 Ghz) ?

I want links to references if exist.

• I'm not an expert on the topic so I'm not going to give an actual response, but I believe this highly depends on the configuration - they might have a single, very powerfull tower in areas with very low population density, and very weak (and many) towers where there is a lot of population. This has to do with the fact that every tower can only serve a maximum amount of connections. I have seen amplifiers for LTE with rated powers of 200W, If my memory serves me right – Joren Vaes Apr 16 '17 at 11:52
• Where I live with a lot of trees my understanding and from the timing of some dead spots, in the fall after the leaves fall they lower the power and in the spring when the leaves come back they raise it. (word of mouth from someone who works in the business) – old_timer Apr 16 '17 at 12:54
• It depends how you define it. Input Power? Effective radiated power ERP at 10m? Try again. Some mobile providers use higher power Tx and have 50% fewer towers or twice the spacing in urban areas. – Tony Stewart Sunnyskyguy EE75 Apr 16 '17 at 13:25
• @TonyStewart.EEsince'75 this is one of the rare cases where OP's lack of clarity is actually a feature of the question. I was about to vote to close it (due to misconceptions and ambiguity), but I took the chance and tried to explain what OP should have considered, in hopes of doing something that might be helpful in the general future / future readers. – Marcus Müller Apr 16 '17 at 13:37
• Thanks @TonyStewart.EEsince'75 sorry for my ambiguous question, I was about to try to edit it but I find that Marcus Müller answered it in great details. Thanks for you all. – Abdelaziz Mokhnache Apr 16 '17 at 14:43

First of all, you have a misconception about GSM, 3G, 4G:

The frequency bands you list are some of the frequency allocations for these networks. These are different between different operators and in different countries.

Then: Cellular networks are not broadcast transmitters. They don't work with constant output powers.

The power they transmit depends on what they need to achieve. As noted in the comments above, a cell tower that covers a huge rural area will blast out more power per user on average than a small-cell tower in a city centre.

Since power consumption is one of the biggest costs in operating a mobile network, carriers are extremely interested in keeping transmit power as low as possible.

Also, lower maximum transmit power allows for smaller coverage area – this sounds like an anti-feature, but it means that the next base station using the exact same frequencies can be closer, which becomes necessary as operators strive to serve very many users in densely populated areas, and thus need to divide these users among as many base stations as possible, to even be theoretically able to serve the cumulative data rate of these.

Then, as mentioned, the transmissions will be exactly as strong as necessary to offer optimal (under some economic definition of "optimal") service to the subscribers. Which means: when there are only a few devices basically idling in the cell, the power output will be orders of magnitude less than when the network is crowded and under heavy load.

Loads have a very high dynamic. You can watch an LTE load monitor from a city centre live here.

This goes as far as shutting down base stations or reducing the number of subbands served at nighttime – something we were able to see very nicely happen every night from the uni lab where I spent a lot of my days (and, obviously, far too many nights).

So, there can't be "this is how much power all towers emits" number, since it depends on usage.

Now, as also mentioned, there's completely different cell types. With 3G and 4G, we saw the proliferation of micro-, nano- and femtocells. Those are just radioheads that can be placed nearly anywhere and serve a very restricted space – for example, a single room.

These obviously would use much less power than a single antenna mounted on a mast somewhere high.

Antenna systems can be very complex, too – a modern base station will make sure to use a combination of antennas to form something like a beam that hits your phone as precisely as possible – motivation for that, again, is less necessary transmit power (lower cost) due to not illuminating anyone who's not interested in the signal you are receiving, and of course, possibility for denser networks.

Then, there's aspects like interoperability. A carrier might offer both 2G and 4G, often closely co-located in spectrum, on the same mast. Now, turning up the 2G downlink's power too much might lead to saturation in 4G receiver (phone) amplifiers – and to drastic reductions in possible 4G rate for a slightly improve in 2G quality.

This problem might get even more important as operators move to deprecate and shut down 2G, and might very soon be broadly adapting schemes where 2G service is "interweaved" into 4G operation in the same band (2G is very slow, and takes only very limited "useful" bandwidth, but still occupies very precious frequency bands, so it's only natural to use the very flexible 4G in a way that says "ok, dear handsets, this is our usage scheme, where we leave holes in time/frequency so that 2G can work 'in between'. Please ignore the content of these holes."). Then, the whole power/quality trade-off might become even more complicated.

In essence, it's also important that when you're carrying around a phone, the most radio energy involved in the operation of the phone network that hits you is not the downlink power (i.e. base station -> phone), but the uplink power, simply because power goes down with the square of distance, and you're darn close to your phone compared to the base station antenna. An important corollary of that is, by the way, and I've seen dozens of people not understanding that, is that the more base stations there are, the less power you get hit by. Very simple: Your phone will need more power to reach a base station far away, and the power that the base station needs to reach your phone will always be adjusted so that your phone will have good reception (if possible!), but not more.

• Important takeaway: building more towers will actually reduce field strengths, for both up- and downlink. – Simon Richter Apr 16 '17 at 18:02
• SO MUCH TEXT SO LITTLE useful info. – Tommixoft Jun 17 '19 at 16:52
• @Tommixoft better than one comment with zero insight. Sorry, things aren't any easier than I described. – Marcus Müller Jun 18 '19 at 6:45
• Do you have a source or reference or a back-of-the-enveloppe calculation for the power consumption being one of the biggest costs in operating a mobile network? Not saying it isn’t, just curious, that would not have been my first idea, spontaneously. – jcaron Dec 27 '19 at 13:37
• @jcaron pheww um, wait. Here: 5G cost evolution – Marcus Müller Dec 28 '19 at 9:51

It depends how you define it. Input Power? Effective radiated power ERP at 10m? 1km?

Some mobile providers use higher power Tx and have 50% fewer towers or twice the spacing in urban areas. define question in terms of dBm or dBuV or dBW ERP @distance or input W vs technology etc, regional specs. Cell towers only transmit around 10 watts usually.

Sometimes up to 50 or so in urban areas. Your phone can transmit up to 2 watts. antenna gain depends on losses and diversity gain (Sphere/beam coverage)

Try again.

This small 4G antenna lists:

20 Watt for GSM 900 and LTE 800
10 Watt for GSM 1800 & UMTS
6 Watt for WLAN & LTE


And this 4/3/2G antenna has a max power use of 50 watt.

Considering exponential power loss with distance and that a microwave dinner requires about 10 minutes right next to a 630 watt antenna in a box designed to focus all that power into it, i would be more worried about my own body heat than sleeping several meters away from a busy cell tower with a roof and/or wall in between.

Even finding ERP a 10 m is difficult. Peak power allowed in the USA can be found near 500 wars per channel. Likely typical power near 100 watts per channel. Intensity measures such as watts per meter are not so easy to obtain because the average number of channels are not included.

• Welcome to EE.SE, Don. You might want to proof-read your post before hitting the "Post Your Answer" button. You have spelling errors and are missing a verb. – Transistor Dec 27 '19 at 4:49