What router should I buy for home use

[jlsstriker]

I am playing my pc smoothly without lag. When my family streams on youtube or downloading filed and etc my game started to lag and so much delays.

What router should I used to make my pc run smoothly and without delays and lag even there's wifi users downloading,streaming and etc.

My budget is $100.

 

[thiggins]

I am playing my pc smoothly without lag. When my family streams on youtube or downloading filed and etc my game started to lag and so much delays.

What router should I used to make my pc run smoothly and without delays and lag even there's wifi users downloading,streaming and etc.

My budget is $100.

There is no simple answer to your question. Much more information is needed such as what router you have now, number and type of devices, how many users are doing what at the same time, how far away from the router users are located, how many neighboring wireless networks and much more.

If you don't want to do the homework, buy a top-ranked AC1200 class router in your price range and give it a try. No guarantees.

http://www.smallnetbuilder.com/tools/rankers/router/ranking/AC1200/rev8/6

 

[L&LD]

Is your pc connected wired or wirelessly? How fast are your ISP speeds (up/down) and do they actually match (consistently) what you're paying for?

If you're running any N class router (even the venerable RT-N66U) or even the same basic hardware first generation AC class router (such as the RT-AC66U), then moving to an newer AC1200 class router (or higher) as Tim suggests will be an improvement if your ISP speeds are not to blame for your issues.

However, I would suggest the RT-AC56U for about $50 to $100 depending on the sale you can find instead of anything from TP-Link, Linksys or Netgear (let alone the other no name routers available).

With RMerlin firmware support and forks based off of RMerlin's work (john9527 and hggomes) available for the RT-AC56U and other select Asus routers, considering other brands is not in your long term interest for many aspects of router ownership. Including continuing support for security risks, performance and feature improvements and the forum here whose members can answer almost any question regarding the supported Asus routers.

http://www.snbforums.com/threads/asuswrt-merlin-378-55-3_hgg-final-mod.26524/page-2#post-199549

 

[JTravers]

Sounds like you're suffering from bufferbloat. Run a speedtest at dslreports.com to confirm.

Get a router with proper fq_codel based QOS. Something with Streamboost or adaptive QOS from Netgear or Asus. Or get a router which is supported by OpenWRT if you're adventurous and technically inclined.

 

[pete y testing]

There is no simple answer to your question. Much more information is needed such as what router you have now, number and type of devices, how many users are doing what at the same time, how far away from the router users are located, how many neighboring wireless networks and much more.

also to add to tim's list what type of internet do you have and what speed plan are you on as it could just be a case os supply and demand , your internet can only supply so much and if ppl are asking for more that what its capable of someone is going to suffer

if your stuck on some slow as adsl you have issues to start with

so OP explain exactly what type of internet and plan , what type of modem and model , what type of clients and how they connect

the more info you give us the better we can inform you

one thing i can tell you now

My budget is $100.

that isnt going to help and you prob need to expand the budget if and only if new hardware will help , if it is a case of supply and demand and you have bugger all supply then only changing usage patterns of the users in the house is going to help

 

[NightOwl326]

It's almost like Smallnetbuilder doesn't have every tool imaginable to make a router selection. And after hours of testing, evaluations and rankings, the "what should I buy" and the "which router A,B or C router should I buy" questions never end. I wonder if other sites are full of "what car should I buy", "what pet should I get" and "which women should I marry questions"? Sheeeeessssshhhhh

 

[woe_and_fury]

It's almost like Smallnetbuilder doesn't have every tool imaginable to make a router selection. And after hours of testing, evaluations and rankings, the "what should I buy" and the "which router A,B or C router should I buy" questions never end. I wonder if other sites are full of "what car should I buy", "what pet should I get" and "which women should I marry questions"? Sheeeeessssshhhhh

Im from head-fi and its the same with questions on headphones. I have x amount what should I buy?

Then we ask, what is their current setup, their music source, what content, genre, open or close preferences.

But if you help out the newbies like I once was, sometimes you get people who help the community and stick around!

 

[sfx2000]

Sounds like you're suffering from bufferbloat. Run a speedtest at dslreports.com to confirm.

Get a router with proper fq_codel based QOS. Something with Streamboost or adaptive QOS from Netgear or Asus. Or get a router which is supported by OpenWRT if you're adventurous and technically inclined.

bufferbloat concern is vastly over-rated, IMHO as a former IETF/3GPP2/IEEE standards engineer...

I have issues with how DSLReports is 'testing' this, as many times, it's not on the LAN/WAN interface, but it's mostly upstream at the carrier level...

 

[sfx2000]

My budget is $100.

WRT54GL and OpenWRT...

http://www.linksys.com/us/p/P-WRT54GL/

 

[JTravers]

bufferbloat concern is vastly over-rated, IMHO as a former IETF/3GPP2/IEEE standards engineer...

I have issues with how DSLReports is 'testing' this, as many times, it's not on the LAN/WAN interface, but it's mostly upstream at the carrier level...

I think the point with these solutions is to move the bottleneck from the carrier to the customer, and then to solve it with things like traffic shaping and fair queueing.

I had the exact same issue with maxed out uploads affecting my downloads or vice versa. Had an F score on the bufferbloat test. Loaded OpenWRT onto my router and turned on fq_codel and sqm-scripts and the upload/download problems went away, web browsing remained speedy even when maxing out an upload (ping times barely increased), and bufferbloat score went up to an A.

Maybe the issue is overblown. I trust your opinion, as I'm not an expert. However, the tools to "fix" it seem to also fix problems like those the OP is describing. So I think my recommendations still make sense.

 

[sfx2000]

I think the point with these solutions is to move the bottleneck from the carrier to the customer, and then to solve it with things like traffic shaping and fair queueing.

I had the exact same issue with maxed out uploads affecting my downloads or vice versa. Had an F score on the bufferbloat test. Loaded OpenWRT onto my router and turned on fq_codel and sqm-scripts and the upload/download problems went away, web browsing remained speedy even when maxing out an upload (ping times barely increased), and bufferbloat score went up to an A.

Maybe the issue is overblown. I trust your opinion, as I'm not an expert. However, the tools to "fix" it seem to also fix problems like those the OP is describing. So I think my recommendations still make sense.

It's a hot button topic - no doubt - but at the same time, most home Router/AP's are actually not that bad with regards to buffer bloat due to limited memory for buffers on the device itself (many AC1900 class devices are still limited to 256MB RAM). Don't believe me - do some internal testing across a WAN port inside your LAN... there are tools that can measure this, and most routers one can buy on the BestBuy/Fry's/Amazon/NewEGG shelves - they're fine with regards to this...

So there's little that can be done from an end-user/customer perspective - most of it is upstream, either at the carrier, or the upper tier providers from them - e.g. backbone of the internet at large...

At the last mile/local loops - DSL and Fibre are a bit better as a transport than Cable, due to architectural differences, but at the same time, with CDN's, best cost routing (cost is a driver here) and the inherent bias of bandwidth towards the downstream side (from operator to customer), there will always be some level of buffering...

 

[Nullity]

It's a hot button topic - no doubt - but at the same time, most home Router/AP's are actually not that bad with regards to buffer bloat due to limited memory for buffers on the device itself (many AC1900 class devices are still limited to 256MB RAM). Don't believe me - do some internal testing across a WAN port inside your LAN... there are tools that can measure this, and most routers one can buy on the BestBuy/Fry's/Amazon/NewEGG shelves - they're fine with regards to this...

So there's little that can be done from an end-user/customer perspective - most of it is upstream, either at the carrier, or the upper tier providers from them - e.g. backbone of the internet at large...

At the last mile/local loops - DSL and Fibre are a bit better as a transport than Cable, due to architectural differences, but at the same time, with CDN's, best cost routing (cost is a driver here) and the inherent bias of bandwidth towards the downstream side (from operator to customer), there will always be some level of buffering...

Without CoDel (& traffic-shaping) my ping during a connection saturating upload is ~600ms. With CoDel, latency never exceeds 55ms with an average of ~35ms during said upload test.

Granted, it could be coincidence, but I have done enough testing (routers: Speedstream 4100, Sagemcom 1704n, multiple PCs with multiple NICs running both Linux & FreeBSD/pfSense) and the only thing that fixed my ping during upload saturation was CoDel.

Download saturation causing latency spikes is very likely an upstream problem though.




My perspective: Which node is the throughput pinch point during connection saturation? With upload, it is very likely the customer's modem or router (depending on whether traffic-shaping is used). With download... I'd say it was a device within the ISP's network, assuming that your ISP limits the throughput they transmit to you to whatever service level (10Mbit, 25Mbit, 100Mbit, etc) you are paying for.

 

[jlsstriker]

I really have a bad download and upload speed about 2.mbps downstream and 0.81 mbps. Number of devices connected are only smartphones and tablets, and then a maximum of 7 devices connected.
3 users streaming at the same time.
There are about 2-3 neighboring wireless networks and users are about 3m-5m close to the router.
My pc is wired connected

 

[sfx2000]

Without CoDel (& traffic-shaping) my ping during a connection saturating upload is ~600ms. With CoDel, latency never exceeds 55ms with an average of ~35ms during said upload test.

With and without QoS/Media Prioritization, whatever - my latency is around 12mSec - and I'm no one special with regards to my home broadband (they're actually a competitor, but they offer better BW than...)

CoDel is a good thing - turn key and generally works without user interactive tweaking

But like I said earlier - a lot of bufferbloat is not on the last mile to the home/business - much of it is upstream - and there, it depends on traffic. I have concerns about how DSLReports.com is reporting this, as many factors are out of their control - so the feedback their reports here are informative (nice to know) rather that normative...

 

[RMerlin]

Just a random thought: testing bufferbloat using a ping might not give accurate results IMHO, as ICMP packets might have a lower priority than TCP packets.

Sent from my Nexus 9 using Tapatalk

 

[Nullity]

Just a random thought: testing bufferbloat using a ping might not give accurate results IMHO, as ICMP packets might have a lower priority than TCP packets.

Sent from my Nexus 9 using Tapatalk

That just means that your real world results would be equal or better than the ping latency. Underestimating is probably a good thing.

 

[RMerlin]

That just means that your real world results would be equal or better than the ping latency. Underestimating is probably a good thing.

True. But it also means some people might be wasting their time trying to improve a situation that's already working fine (or, that a fix they applied does work, despite their ping latency remaining poor).

Just something to keep in mind while trying to work out what can have the appearance of being problematic. Or, it might lead someone into devising a better test method / test tool.

 

[Incarnate]

It's almost like Smallnetbuilder doesn't have every tool imaginable to make a router selection. And after hours of testing, evaluations and rankings, the "what should I buy" and the "which router A,B or C router should I buy" questions never end. I wonder if other sites are full of "what car should I buy", "what pet should I get" and "which women should I marry questions"? Sheeeeessssshhhhh

I could be wrong, but it looks like the title of this particular forum is "Wireless Buying Advice". There seems to be many other sections of this forum that may cover topics other than making a router selection. Maybe you should hang out in one of those.

 

[sfx2000]

I wonder if other sites are full of "what car should I buy", "what pet should I get" and "which women should I marry questions"? Sheeeeessssshhhhh

Routers are like cats, and tastes vary - some like their pussy furry, and others like them less so...

(sorry I couldn't resist...)

 

[Nullity]

I swear that picture of the hairless cat is immoral or illegal.

Ugggh... :/