Příspěvekod David2006 » 11 years ago
Snad někomu něco málo osvětlí následující popis:
Greetings,
I have seen some of the discussion in various places on AF5 range and data-rate performance, antennas/form factor and other miscellaneous topics that might benefit from some further comments.
Firstly let’s look at antennas- the antennas for AF5 have been designed by our team using a tailored IWO (Invasive Weed Optimizer) algorithm. Yes, the way weeds colonize your lawn can produce an effective convergence mechanism for electromagnetic modeling! The antenna structure is really a unified system consisting of multiple reflectors and surface structures to control the RF energy. The antennas make use of tightly controlled edge tapers (non-uniform illumination energy) and decoupling mechanisms that have been designed to allow very independent performance of the transmitters and receivers inside the radio (there are a total of 5 independent radios!). It would be difficult to come up with a design like AF5 that allowed for off-the-shelf general-purpose antennas to be used, so that is why this device is so highly integrated and looks the way it looks. It is also very important to make sure that we do not allow the two antenna patterns to come out of collimation. Having the patterns diverge or converge (cross-eyed) would cause significant degradations in system gain and asymmetric behavior in demonstrated data/error rates. By rigidly containing the overall structure, we have very good control over collimation. I will try to get some patterns published soon, but both the front-to-back performance and side lobe characteristics are incredibly clean when viewed as a system.
The AF5 is capable of both TDD and FDD modes and is also able to utilize GPS synchronization to obtain global timing of TX/RX framing. When sync is deployed correctly, AF5 units will not interfere with each other. In the majority of cases the radios will not interfere with each other even without sync due to the pattern and isolation techniques employed.
FDD is capable of an aggregated 1024Mbps throughput with TDD throughput being approximately half that figure. We are still working on a final version of our recommended deployment guidelines, but yes you will be able to do FDD and only use a single 50MHz channel in certain circumstances. The AF5 has a lot of “knobs” in terms of channel bandwidths, constellation rates (both fixed and auto-negotiated). So to be clear, we will have support for 10MHz, 20MHz, 40MHz and 50MHz wide channels. All of these channel widths will be supporting all constellations so you will have quite a bit of flexibility in situations where spectrum is at a premium.
I have seen some folks speculating about data rates and range. Some have even observed that you will not get 256QAM to go 100km! I am not sure how to write specifications more clearly, but any radio system ever conceived can only support signaling constellations that are directly proportional to the available signal/impairment ratio. Even with AF5, the farther you go, the less SNR is available. See the attached picture for a plot of what you can expect to see in the real world. I expect to see significant improvements as we optimize our PAR techniques. Distance scale is in miles and this is for FDD/50MHz.
There will most likely be situations at greater range or when interference is present that may require users to rely on TDD or offset/split frequency TDD, but that will be dependent on a whole bunch of variables. I will put together a complete set of these charts along with the deployment guidelines. Things scale quite proportionately as you reduce the channel widths all the way down to 10MHz channels and TDD (greatly increased link budgets at the expense of max throughput).
The AF5 comes in two different SKU’s; one with extended high frequency coverage (up to 6.2GHz) for regions that allow it and one that covers the U-NII worldwide allocation plus ISM. The AF5U features band-pass filters to improve co-location with incumbent PMP systems in U-NII mid and world band spectrum. The AF5 covers the complete U-NII world band and US ISM spectrum.
More information to follow… I hope this at least addresses a few questions regarding some of the basic design decisions.
Gary Schulz
VP of Engineering
Ubiquiti Networks, Chicago Design Center
-
Přílohy
-
- AF5 Prelim Distance.png (32.74 KiB) Zobrazeno 2621 x
0 x
nastesti cena brani tomu aby si to pokoupl kdejakej smudla na projeni dvou baraku .... skoda ze to nema odelitelny anteny aby se tam dali dat vice smerovy pak by to melo i slusny uplatneni ...last mile nekam do zapadle visky 5 km ideal ....
jinak na stole rozchodim ns5, sxt , sextant nanobridge a ja nevim co jeste na 100mb... ale bavme se o realnem prostredi venku na vetsine uzemi cr ne o cernocernem lese kde jedine ruseni je to kdyz na tebe bafne v noci medved ..... af5 budou uzivatele regulovat asi stejne jako af24 to jest na max ..... 
protoze s tou velikosti anten to ani jinak na delsi vzdalenosti nejde ... pokud by byly zamenitelne nebo lepe jen jednotka aktomu si pripoj co potrebujes pricemz jedna je tx a druha rx , muzes s dostatecnym ufo talerim udelat i tech 100km a budes v limitu !!! a protahnes vic jak s presmazenym vykonem na malejch antenach a proc? rika ti neco pracovni bod tranzistoru?
