Hello,

Anyone on this forum that have experience using wAP LTE (RBwAPR-2nD&R11e-LTE) in Sweden? If so how was your experience using it and what kind of speed are you getting?
Thinking about getting one to add redundancy to my network in case of ISP failure but I would like to know how it works in Sweden ans specifically in terms of speed!

I have deployed them only in rural areas here in Finland. The problem is that we have three operators and on band 20 (800MHz) the channel width is only 5MHz (or was it 10MHz?) Anyways the net result is around 20Mbps throughput. I have also used directional 14dBi antennas (yes, I have been asked to help when everything else had been tried).

After I wrote this I realize this is not at all what you asked

I have deployed them only in rural areas here in Finland. The problem is that we have three operators and on band 20 (800MHz) the channel width is only 5MHz (or was it 10MHz?) Anyways the net result is around 20Mbps throughput. I have also used directional 14dBi antennas (yes, I have been asked to help when everything else had been tried).

After I wrote this I realize this is not at all what you asked

Many thanks. Always good to get real world experience
20 Mbit on LTE is a bit low. I would like to be able to push around 40 Mbit but that may be more then the device can handle which is a same if that is the case.

.... Anyways the net result is around 20Mbps throughput. I have also used directional 14dBi antennas ....

20 Mbit on LTE is a bit low. I would like to be able to push around 40 Mbit ....

When talking about LTE speeds one has to keep a few things in mind:

• as @Petri already observed, carriers are of different band width. Which, as everybody around here already knows, directly translates to peak throughput. Carriers with 10MHz band width and in basic RF configuration (2x2 MIMO, 256QAM modulation in DL, ...) have limitation of max 100Mbps per (FDD!) sector. Which gets lower if SINR at receiver is not good enough (higher than 28-30dB) or if receiving terminal doesn't support highest modulations (e.g. supports only 64QAM). Use of pair of directional antennae (to get 2 Rx streams needed to support 2x2 MIMO) certainly helps improving SINR (depending on nature of interference, increase in SINR can be as high as difference in antenna gain (original vs. new antenna), but as low as 0dB if all of interference is coming from same direction as useful signal).
  This set of rules is just the same for every frequency band (anything between 700MHz and 3.5GHz used for current LTE networks) and every band width (net throughput is directly proportional to carrier band width - 20MHz carrier in the same RF configuration peaks at 200Mbps).
  In TDD networks (not really common in Europe, but most of high-frequency, including 3.5GHz, bands will operate in TDD mode) general rules apply as well, but there the sector bandwidth is split between DL and UL statically (can't be changed on the fly because all base stations in certain area have to be time synchronized and use same set of TDD parameters), so throughput yield seem to be quite lower comparing to FDD (even taking into account that with FDD we're actually using total of 40MHz when talking about 20MHz carriers while in TDD we're really using only 20MHz), if I was talking about 100Mbps previosuly, in TDD this would be around 60 Mbps in DL and 20-25 Mbps in UL, depending on DL/UL time share ratio).
• When thinking about minimum signal level, which would allow to get maximum speed: noise floor in urban areas is generally around -100dBm to -105dBm (mostly caused by other LTE base stations operating on same frequency band), so RSRP (not RSSI) should be at least -75dBm. In rural areas, without much of RF noise, noise floor can get down to -120 dBm (which is absolute low due to thermal noise) and good DL throughput could be achieved at RSRP levels higher than -90dBm. Those signal levels are not easy to get if base station tower is not "right next door", specially so if base station tower is "just around the corner" or behind a thick forest. Almost always helps to build a mast (few metres high) to place terminal ... and keep antenna feeder cables as short as possible, it's easy to loose a dB or three there.
• sector throughput is divided among several users (just the same as WiFi), but as user you can't know how many other users are using RF resources at the same time. But if MNO has decent number of subscribers, chances are that in busy hour there are a few (to a few ten) users and resources are divided among them. Hence it's quite normal that user can get almost full speed in off hours (usually that's in the late night), but only a fraction of that during peak hours (might be in early evening).
• usually MNOs enable kind of fairness when scheduling RF resources to users. Hence if in certain sector there's single user with really low RSRP, the whole sector throughput will go down as that user will use un-proportional amount of RF resources to get some basic service. If you use directional antennae, you make much better use of resources available for you, so you might see even bigger net gain than expected.
  In addition to that, when more users actually use data, the interference increases (mostly inter-cell) and RSRP needed for "ideal" service increases accordingly. Or, in reality, achievable throughput decreases even more than what one would expect because of sharing same base station resources by many users.
• don't forget about UL: when path loss between transmitter and receiver is big enough, terminals (phones, modems) hit Tx power ceiling. When they do, UL will start to suck big time and can suck so big, that it'll harm DL (in TCP, at least ACKs have to be transmitted in timely manner to keep DL full speed). This is where MNOs strategies come into play: some MNOs keep base stations Tx power relatively low, which means less bars on phones, but service is available right to the point when all bars are gone. Some other MNOs increase Tx power on their base stations, increasing number of bars (and slightly increasing DL speeds for users in mediocre RF conditions), but users with only one bar (or even two bars) loose service as their terminal is not capable of Tx power high enough to reach base station's receiver.
  This problem is slightly less if carrier band width is smaller as terminal's Tx power gets spread over less band width (same Tx power in 20MHz carrier results in 3dB lower RSRP than in 10MHz carrier). Which means that "use carrier with wider band" rule is counter-productive when SINR is low (e.g. lower than 10dB).

@Kindis Yeah, 20Mbps is slow, but my wAP LTE Kit installations have been very remote with only band 20 available (800Mhz). With directional 2x2 antennas and short (20cm) antenna cables I have succeeded where nothing else has worked. In that sense 20Mbps is really good (twice the capacity of thin Ethernet of the 90's!)

In urban areas I have deployed LTE SXTs (the old ones with only bands 3 and 7) and with them 60Mbps is so common that I suspect the operators cap the throughput. I haven't really dug into it.

Thanks everyone for your input! I will give wAP LTE a try and see what it provides!