Should MikroTik make more powerful antennas and wireless protocols in 2019?

RouterOS General
1

Do you agree with what is in this topic?

  • Yes
  • No
0 voters

Hi all,

I’m writing this topic to let users write about what they expect for the future and to let staff read different opinions.

First, something that everyone loves: RouterOS.
I think everyone agree that RouterOS is the best OS for antennas, you can take a RB and if you want it to make coffee, you just need to configure it, but I’m not writing this topic for ROS.

Always more users are abandoning MikroTik for outdoor wireless, especially WISPs, because we can’t go on with actual MikroTik antennas with:

  1. Crowded frequencies and noisy towers
  2. Old antennas that can’t handle more than 20-30 CPE stations (I can’t use 12 Net Metal on a tower)
  3. 5GHz PTP that can’t handle more than 100Mbps when you’re near and lucky.
  4. Old protocols that have been “improved” so much time that now they should even resolve transmission problems (Nv2 made is time)

I don’t want to talk about competitors, but I have to:

with Ubiquiti Air Fiber I can make a PTP 16km link with 200Mbps EFFECTIVE throughput, what are they, geniuses? With a Net Metal and the best parabolic antenna I can find, I can barely achieve 80Mbps with 16km link.

Other competitors have antennas for PTMP that can handle 100 clients, with a NetMetal I can barely handle 30 clients.

Nv2, MikroTik’s Achilles heel:

Users always complaint about Nv2 functionality, and they’re totally right.
Why in the hell Nv2 still provides SAME TDMA time slots for clients transmitting when other clients are requesting it? Shouldn’t it reduce (even drastically) time slots for old transmission and give them to new transmission?

GPS synchronization, why MikroTik isn’t interested in this?

Signal filtering:

I’m again talking about competitors, but why there isn’t something like Ubiquiti airPrism technology for MikroTik?

If you sell CCRs that are 3000$ worth, why not making 400-500$ antennas, multi-core and with powerful stuff?

Users want to stay with MK, but they can’t if you don’t make new antennas, but real powerful ones, we need rockets, not fireworks!

I hope that someone from staff will reply, I’m sorry if I write other competitors name, but I hope this would make MK think about the direction they’re taking.

Users please reply with your opinions, don’t let this topic die.

2

I think it all depends on company’s vision, what it wants to achieve / represent, and also business cases for each of the solutions, how big a market is there for it and how much will it cost in R&D.
Technically it is just a matter of priority, right knowledge (which you can hire if necessary) & effort.

3

I don’t think that selling only CCRs will let MikrotiTik make so much profit.

4

I believe it’s all about money.
Ubnt is in USA, it has more resources, more money and more know how.
Mikrotik is a small company in a small country.
The costs of developing such expensive systems might be overwhelming.
Anyway I dont use ubnt anymore because of the failure rate which is at least 10 times bigger.

5

Without ROS7, NO!!!

Sent from my C6833 using Tapatalk

6

100% and we need current wireless drivers, not this basic home improvement software,
With 4 years delay to competitors.

  • all competitors have Spectral Scan
  • all competitors have better performances with TDMA
  • No Roaming features
  • No band steering
  • no advanced Channel (eg 10 MHz)
  • no wave 2

There is no light in the tunnel, after releasing ROS 7.00 it would take a year to bring it to production system, it was the same with Ros 6

If you need 5 GHz stuff don’t buy MikrotiK

7

Guys, please…
If 6.44 would be renamed to version 7.0, it won’t solve all the problems.
First we need new hardware, a Net Metal costs 150$ or something near that? That’s the top of MK antennas, if you look to other vendors you can buy the equivalent of a Disc Lite with only 150$.
There isn’t so much cost for hardware development, RB chipsets aren’t so much tailored, the architecture isn’t unknown.
I think the best would be a totally new Net Metal under hardware specifications and performance.
I would really like to buy 500€ worth Net Metal, when it will come out, they’ll be out of stock everywhere because I’m going to buy all of them.
Also I would like to see more parabolic and sector antennas in MK store.

After this, new hardware features, like GPS, noise filtering and so on.

After all of this MK should really focus on ROS vi and new software improvements, and of course, a totally new TDMA more aggressive in time division, taking slots from ones transmitting too much to have equal time sharing between clients.

8

They don’t have a hardware problem, they have a software issue.

The same device with ipq4018 so performe well from another vendors, have more functions and deliver more speed to more devices, this is all about software, a piece of software that has an 6 years old kernel.

There is no issue with functionality they are added more and more, but there are issues with the basic.

If you use an old kernel and you want new hardware support. You have do to that by your own, and this is the problem, Mikrotik Wireless Drivers support newer chipsets only with basic features, and that is why they

  • don’t perform well
  • we Are misssing Spectral Scan, Wave 2 and more
  • they have compatibility issues

On 4011 all Intel ac clients only connecting with 54MBit, my friend have new Asus Router with the same Chipset, and what does my Intel ac Laptop?
Connecting with 1300 MBit, welcome to Mikrotik legacy Wireless world, they had done an Business decision, and there is no Future for
Outdoor Wireless, but they don’t Tell US.

Shure they can build something like AF, but this would take years, you would no see anything in 2019 and. 2020, and we don’t need more new ARM Based ac Hardware until the Software issues are fixed.

This issue is known for 12 month now..,
http://forum.mikrotik.com/t/arm-devices-and-nv2-protocol/120717/1

9

You should not see “hardware” only as chipsets.
What about antennas? I use for example UBNT Titanium sectors with Net Metal, they work really well.
Why MK can’t do different sector antennas?
They made LDF that can be used with every parabolic antenna, but why not making also a parabolic antenna for them?
I also would like to see multi core CPEs, a Net Metal doing nothing has 20% of core usage.

10

LTE antennas with more gain (antennas only) for LTE PCI-e cards.

Regards.

11

SXTsq AC is cheap as a CPE and Multicore , but the ARM staff don’t work in Nv2 environment so useless for most.

RF Element sectors are much better then UBNt ones, most UBNt guys use RF Elements horns.

Really ?? Bring a Sat Dish to the market ??
You can buy one on every corner in the world!
The intention here was shipping cost, why ship big antennas to the world, and everybody can buy one in various dimensions in every Electronic market…
This is Intension for LDF…

12

I tried those RF horns and with a signal of -60 I have really bad performance, with UBNT Titanium sectors clients go from 5Mbps down to 30 with same signal.
I don’t know why, but if I use UBNT antennas with UBNT rockets, Nano etc, they work better, they’re specifically made for them, like mANT for MK, but I can’t use 120 degrees sectors.

I was talking about Net Metal’s core, I don’t care if an SXT is multicore, I can’t use it as AP for clients like Net Metal.

I also would like new software improvements and feature, I’m not dreaming of v7, but as I wrote, a more powerful TDMA and GPS sync (that also requires new hardware).

13

We have complete another experience with mt/ubnt and Mimosa Electronics , RF Elements Sector Performance really well, we use 30°/40° ones and uses smaller channels with really good speeds

CCQ‘s over 98% on all clients…
300MBIt real TCP over 40Mhz with Mimosa

M11g with AC Modul ist multicore to

We really need is software renew, getting wave 2 feature for hotels etc, roaming…

And for WISP working TDMA with good speeds.
Spectral Scan to find clean channel
Rf filtering and gps would be great but they will need 3 years to get them running…

14

The basic problem is that we use the old RouterOS V6 from 2013. Based on the old kernel 3.3.5 :frowning:
With that is not possible to do miracles in 2019

15

Any other comments please?

16

People are on vacation and don’t read forums

17

As other posters have said, new antennas and protocols don’t mean anything when we’re still forced to use a six year old kernel with a hacked-together wifi driver that barely supports any modern features. I want an up to date kernel and non-proprietary wifi drivers far more than I want new antennas and (yet more) proprietary protocols.

18

You know that we will not see new kernel on current RBs? Update actual RBs or make new ones with new kernel, better hardware and so on, what’s the best choice for MK that is selling their products?
Come on, it’s not so different than smartphones.

19

It is different no one buys Andoid 2.3 phones today…
but Mikrotik is there at this moment
All routerboards that on the market would profit from newer Kernel, it is easy, put openwrt to you RB and try it with 4.14 kernel (4.9 depends on board)… Mikrotik is in the dead and in a very smal street, i don’t No if the could turn around the car… or stop it in the end of the road… they don’t Tell. They don’t care…

20

I don’t see why not, there isn’t that much hardware differences between all the models and they are all supported in upstream kernel (except TILE which is a dead end). In fact I just re-purposed an RB951G-2HnD for running DNSCrypt, it shows it’s possible to run a modern kernel with old board no problem:

root@OpenWrt:~# cat /proc/cpuinfo
system type             : Atheros AR9344 rev 2
machine                 : MikroTik RouterBOARD 951G-2HnD
processor               : 0
cpu model               : MIPS 74Kc V4.12
BogoMIPS                : 299.82
wait instruction        : yes
microsecond timers      : yes
tlb_entries             : 32
extra interrupt vector  : yes
hardware watchpoint     : yes, count: 4, address/irw mask: [0x0ffc, 0x0ffc, 0x0ffb, 0x0ffb]
isa                     : mips1 mips2 mips32r1 mips32r2
ASEs implemented        : mips16 dsp dsp2
shadow register sets    : 1
kscratch registers      : 0
package                 : 0
core                    : 0
VCED exceptions         : not available
VCEI exceptions         : not available

root@OpenWrt:~# uname -a
Linux OpenWrt 4.9.120 #0 Wed Aug 15 20:42:39 2018 mips GNU/Linux