Ok, I'm back. Taking a 2nd look at the lousy 21.5KByte picture, I'm not sure if the mid size boxes are TMA's or RRH (remote radio heads). Some of the smaller boxes might be RoF (RF over Fiber), combiners, splitters, or AISGv2 alarm concentrators. Difficult to tell from such a small photo.
I couldn't find any decent photos of the inside of a TMA. If I dig some up, I'll post them.
The internal configuration of the panel antennas varies with frequency. There are all kinds of elements (dipole, folded dipole, LPDA, Franklin/Amos, patch, PIFA, etc). Some examples: If you scroll to the bottom of the above page, there's a Samsung RRU (remote radio unit) photo.
You might find it useful to see what's actually on a "typical" cell tower. This is a spreadsheet of a proposed local Verizon tower than will probably not be built because of local opposition: Note that there are 4 frequency bands and a total of 39 antenna and TMA sections.
Note that the recent AWS-3 spectrum auction will add addition cellular spectrum, which may require additional antennas, TMA's, permissions, hearings, etc.
I hope this helps.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
My favorite is this AT&T camouflaged tower: For those that don't believe that lumber grows on trees.
Various camouflaged cell sites, mostly in southern Calif:
Anything worth doing, is also worth overdoing:
Burning cell towers:
Curing cell tower "negative energy" by "gifting" with Orgonite:
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
TMA (tower mounted amplifiers), RRH (remote radio heads), and small cells have replaced what few cryogenic amps and filters were installed. STI still makes them: The problem is that they are made to be rack mounted and installed in a shelter, and cannot easily be mounted on the tower. The trend is to move as much of the RF to the top of the tower to reduce coax losses, and PIM (passive intermodulation). The only place where cryogenic amps and filters seem to fit are in areas where the shelter is on a rooftop, zoning requirements prevent the installation of additional cell sites, and the benevolent cellular provider wants to increase range or squeeze some additional performance out of the existing system. Even so, using a cooled receiver front end is problematic in an RF polluted site as the increased sensitivity tends to create even more intermod in the receiver.
That's 50 or 60 Hz transformer hum coming from the shelter building (not the tower).
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
Likely a block device fitted with precision machined individual cavities yielding a phased array configuration (ala modern flat panel radar) where then a "diversity receiver" paradigm gets implemented.
No, I haven't had one of those fall off a tower so that I can inspect the guts.
I think you mean beamforming, not steering. Beamforming and SIMO, MISO, and MIMO, have been part of the LTE spec from the beginning. I suspect the beamforming is done like a phased array antenna, or like some 802.11n wireless routers, with phase shifters between antenna elements. There should be a patent buried somewhere that describes how it operated. Google finds plenty of likely hits:
More than you probably wanted to know: Note that it also covers the TDD (time division duplex) version of LTE, which operates on one channel per handset, instead of the more common FDD (frequency division duplex) which requires separate TX and RX frequencies.
The motorized antennas that you are probably referring to are used to adjust the vertical downtilt of a sector antenna. This can be very critical and is difficult to adjust during antenna installation. It can also be done with electrical downtilt, which is more suitable for omnidirectional antennas.
This covers some of the technology, with the added bonus of more photos of the inside of cellular panel antennas:
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
Respectfully - If you had trouble with TDMA, then you are completely screwe d when it comes to figuring out VoLTE. :)
You're in good company though, because I doubt there are even a handful of folks worldwide who understand this technology COMPLETELY from end-to-end. It's a classic case of technology outstripping the ability of folks to ade quately maintain it. (At least, my opinion based on what I've witnessed th us far..)
I got a peek under the hood in a recent in-house presenation from one of th e large carriers. Voice over LTE ("VoLTE") is complicated beyond belief.
There's also something called "carrier aggregation" in LTE where you can ta ke 2+ channels (even in different bands) and merge them as if they were one wide pipe. In this market, the plan is to aggregate 700 LTE, 850 cellular LTE, and AWS band LTE (1.7GHz UL, 2.1 GH DL), to create a "single" 25 MHz channel. I'm told where it's already rolled out that downlink speeds are s ustainably greater than 75mbps (with theoretical peaks much higher than tha t). I still have lingering doubts about the ability of any handset to "pe rform as intended" when it is simultaneously handling frequencies that are nearly 1 GHz apart from each other. It must work though, or the carriers w ouldn't bother to invest the capital to roll it out.
The way it seems to work for me is that as soon as I understand a fair percentage of the current technology, a new technology arrives where I have to fight the learning curve from scratch. It's been like that (for me) since IMTS phones. Worse, my "fair percentage" seems to be decreasing as 3GPP cranks out bigger and more complexicated standards.
I suspect that they all work for the chip vendors. When those that know move on to something else, it's a simple matter to bring in the next generation of cognoscenti, who will create a new and improved protocol Tower of Babel. The real innovation is SDR (software defined radio) which (within limits) makes it possible to design incredibly complexicated systems in software, and change them as required. Just don't let the consumers know, or they might wonder why they're stuck with a vendor locked handset.
Maintenance? Surely you jest. Most of what I've seen is wholesale rip out the old, and shove in the new. Offhand, I would guess about an 8 year cycle time, possibly synchronized to FCC spectrum auctions.
Yes, but it allows for billing voice at data rates, which allegedly cuts costs. Basically, VoLTE converts voice into just another data service, along with all the other digital cellular services (SMS, internet, tracking, NSA snooping, etc. Having totally separate circuitry and frequencies for voice was wasteful. Most of the complexity is in firmware and software, which do a good job of hiding the complexity from the user.
In cable and DSL modems, it's called "channel bonding". Same idea but different names, as has been traditional in the telecom business since Ma Bell was asked to get out of the computah business and complied by renaming everything.
Carrier aggregation offers some realy bad ideas, such as:
That's currently being done in a few DSL modems, all DOCSIS 3.0 cable modems, and some dual-band wi-fi routers. Just about anything that runs OFDM modulation already does channel bonding of a sort. Even
802.11g OFDM also does this (to reduce the effects of frequency selective fading).
The idea is simple. Chop up the incoming data in a number of relatively narrow band channels, and recombine them at the receiving end. If one channel is lost or trashed, data continues to flow, but at a slower rate. It really doesn't matter what carrier frequency these small channels are using, as long as the frequency can support the modulation rate, and that the different delays don't make a mess of the phasing. For example, the current DOCSIS 3.0 modems can conglomerate up to 8 channels, spread anywhere between about 110 MHz and 1GHz. If it can be done with cable RF, it can also be done OTA (over the air) RF.
Sure. On a lab bench using test equipment, in a simulation, in a perfect RF environment, or with one one user. In the field, there are multiple sources of interference, reflections, noise, etc which tend to reduce preformance. I can believe that 75 MHz is possible mostly because I've seen 20 Mbit/sec downloads on LTE using 2x2 MIMO. Higher orders of MIMO will probably go correspondingly faster. T-Mobile is starting to deploy 4x2 MIMO and Sprint 8x8 which should produce impressive advertised rates: Note that going from 2x2 to 4x2 doubles the number of antenna elements at the cell site, which means either bigger or more antennas. Although it's in the 3GPP Release 8 specification, 8x8 is currently science fiction. Visulize 8 antennas on a handset or 4 times the number of antennas on a tower. This sorta illustrates the problem: Also add a NFC (near field communicaitons) loop, and a cordless charging loop. The coffee cup warmer loop will be in the next update.
Don't worry about the 1GHz separation. It's the percentage bandwidth that's important. As I previous mumbled, as long as the frequency can deliver a small part of the data stream for later reassembly, it should work. That's why T-Mobiles mixing of 800 MHz and 5.7 GHz scheme is technically functional.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.