Why Your WiFi Sucks

[tvyankee]

I received this the other day and thought some of you might like to read it. But if this has already been posted i am sorry.

Hope you enjoy.

 

[thiggins]

Jeez, I thought *I* am long-winded!

What were your take-aways from the piece?

 

[tvyankee]

Hey,

I don't know what to really think other then what the article says. The manufactures are happy enough with what they have.

I guess putting out new stuff is more important instead of fixing issues with what they have. As long as people are buying the're flying.

What do you think?

Thanks

 

[L&LD]

I didn't read it like that at all.

With a very quick skim of the 20 page article, with next to useless illustrations and graphs, size and resolution wise, I just see the challenges WiFi manufacturers have.

Things take time to change and that is why I prefer having the chance to try beta products for myself. If they work even slightly better than what I had, great. Otherwise, they are returned and I wait for the next promise to test again.

 

[tvyankee]

Hello,

Is that why really pretty much every piece of firmware that runs these routers are never 100 percent and why they always have some kind of security concern with every upgrade.

Thanks to people like Merlin we regular people have a fighting chance of having a piece of gear that almost works flawlessly or as its been advertised to work. I think they fix what they have now and stop misleading people about the next best thing. Asus as an example has put out new firmware many times for there own product that doesn't really well why should they keep putting out new stuff knowing full well the old stuff does not work correctly. If i had this kinda of approach at my job i would picking up cans on the highway.

I dont think this paper was about graphs and pictures and i think somewhere in there the guy says something about the bad pictures but more about the reasons for why wireless is bad.

Also he says in the last paragraph that the vendors have a good enough mind set that probably wont change in at least my life time. and when i say my life time i mean mine. I believe its true.

Thanks

 

[L&LD]

I still believe you're missing the point.

I do not for one minute believe manufacturers know how to do it right but are not.

If this were the case, we'd have at our disposal easy fixes with a simple google search.


I too appreciate RMerlin's contributions. But that has little to do with transforming theory to a working product. Which we already have.

The complete wifi ecosystem as is may not be perfect in each and every situation, but considering where we were just a few years ago, I am more than content with what we have.

All the issues the paper discussed seem to be addressed at some level in AC Wave 2 and beyond. Patience will prove fruitful.

And when you're so pessimistic that you believe change won't come in your lifetime, I guess you just don't realize how technology actually moves for most people. It doesn't advance in a smooth progression day after day or even year after year, rather; it seems to make radical jumps when all the necessary tech comes together to make an (old) idea finally possible.

 

[htismaqe]

It's not just technology, it's also economics.

I know people hate to think about it but corporations make routers to make money.

It isn't any different with home electronics or appliances. Have you bought a washer and dryer lately, for example? For that matter, take a look at cell phone and satellite/cable TV.

We, particularly in the US, live in a consumer economy. Companies make money by acquiring new users. Supporting existing users COSTS money, sometimes a lot.

Customer service is a lost art in the United States. It isn't a problem uniquely associated with wireless router manufacturers.

 

[azazel1024]

It's not just technology, it's also economics.

I know people hate to think about it but corporations make routers to make money.

It isn't any different with home electronics or appliances. Have you bought a washer and dryer lately, for example? For that matter, take a look at cell phone and satellite/cable TV.

We, particularly in the US, live in a consumer economy. Companies make money by acquiring new users. Supporting existing users COSTS money, sometimes a lot.

Customer service is a lost art in the United States. It isn't a problem uniquely associated with wireless router manufacturers.

And to add in to that, QA costs LOTS of money. Granted, sometimes bone head issues sneak right past the manufacturer and make it in to firmware for issues that darned well should have been caught in any kind of halfway decent QA.

Other times the issues are things that you only find when you put 100 different hardware combinations together with half a million users and THEN you find the problem. You cannot replicate that kind of stuff in QA. At least not without huge expenses.

Routers are commodity pieces of equipment. You aren't paying support contracts on them and you aren't buying $100 hardware with $900 of marketing, QA, support and profit margin built in. You are buying a $20-200 piece of equipment with probably a few cents to a few bucks per router QA and support built in (including warranty support) and a small profit margin.

You aren't going to find a dozen engineers dedicated to trouble shooting bugs, fixing them, QA and then release. Even on the higher end products you probably have a team of half a dozen engineers working on the ENTIRE product line-up...probably including developing new products. So they'll shoe horn in a few hours a month on one of the engineer's busy schedules to take a look at the reported problems for the month and see if they can fix a couple of them and maybe they'll push out a new firmware every month or three.

 

[htismaqe]

We also live in a connected world.

Just look at the latest video game release and how many patches it gets over the course of the first few months.

Companies wouldn't be nearly as eager to use their customers as QA testers if they didn't have the at-hand capability to roll out patches and hot fixes as early and often as they can afford...

 

[RMerlin]

Thanks to people like Merlin we regular people have a fighting chance of having a piece of gear that almost works flawlessly or as its been advertised to work.

My work does next to nothing to help with the issues raised in that article however. Those have to do with the RF engineering side of the router, while I only work with the firmware environment itself that surrounds the radio code. At best I can downgrade a broken driver to a last known good version, but that's pretty much it.

 

[sm00thpapa]

Guess I'll read this in full since my wifi sucks on the 2.4 band. LOL!

 

[tvyankee]

I haven't missed anything but if you want to believe that you are in entitled to your opinion.
I hope in the future that these companies put together better equipment. Easy to overlook shortcomings but its harder to fix them.

I know im a pessimist .

 

[sm00thpapa]

Just sad that routers are getting better and costing more but the WiFi is still the same range as 10 years ago.

Sent from my LG-D800 using Tapatalk

 

[L&LD]

Just sad that routers are getting better and costing more but the WiFi is still the same range as 10 years ago.

Sent from my LG-D800 using Tapatalk

Did the physics change since then?

 

[stevech]

Just sad that routers are getting better and costing more but the WiFi is still the same range as 10 years ago.

Sent from my LG-D800 using Tapatalk

Laws of physics prevail.
RF attenuation vs. distance, add obstructions.
Only so many bits per Hz per second. It does increase to a point, but then you just run out of signal to noise ratio, and hit the wall on the transmitter's amplifier linearity vs. its cost. Higher speeds of OFDM in MIMO leads to need for greater linearity and that's expensive.

 

[htismaqe]

the WiFi is still the same range as 10 years ago.

I don't think this is even remotely true.

 

[azazel1024]

I don't think this is even remotely true.

I'd agree. If you look at it, both based on reported numbers and my, and I would say many people's, experience ranges have increased. At least modest amounts. 802.11n DOES have at least a modest increase in range over 11b/g. It isn't going to be 10x the distance and indoors, it might be hard to notice because of what obstructions can do.

Beam forming can SOMETIMES increase range, though in general beamforming doesn't do much to increase signal strength at longer distances.

Also, it could be completely my perception and not reality, but I do think that access points have generally gotten somewhat better at SNIR discrimination compared to older access points, which does mean more range.

Nothing has dramatically changed the picture of wifi, in terms of range, over the years, but there have been at least some minor incremental improvements over time.

Speed has deffinitely jumped by leaps and bounds. I remember back in college testing 11b wifi and being astounded that I could get 4Mbps over the air. That's 500KB/SEC! I mean, I could nearly saturate my cable connection over WIRELESS.

You want to talk about paltry improvements. 2001, home cable internet connection 5/768, wireless connection ~4Mbps effective. 2014 home FIOS internet connection 75/35, wireless connection ~200Mbps effective (and that is just 11n).

Over that time I have seen wireless increase by 50x, yet my home internet connection has gone up in speed by ~15x (okay, around 50x if you look at upstream, but if I was on cable it probably would be more like 7-8x instead).

When it comes down to it, physics still rule. The abilities of amps is still bound by physics and economics as are signal processors. FCC hasn't suddenly lifted broadcast power restrictions on 2.4/5GHz (okay, they did raise part of 5.2GHz range, but not above existing max on the rest of 5.2-5.9GHz) and clients are generally still power limited in their radios too.

Increased range is going to be at best very minor improvements from better amps and signal processors as well as possibly tweaks in algorithms, broadcast technologies, etc.

 

[thiggins]

Yes, physics have not changed over the years. And there have been improvements in receiver sensitivity.

The largest change has been in modulation, coding and MIMO to make more effective use of channel bandwidth. This results in higher throughput at a given signal level, which often translates to increased effective range.

The attached plot illustrates the point using N150, N300 and AC1750 routers (and clients that can take advantage of the higher link rates).

If you are sitting at 54 dB attenuation location, the AC1750 connection would provide a usable connection while the N150 and N300 would not.

The plots all end around the same point due to physics (signal attenuation).