Posted: Tue Feb 09, 2016 19:47 Post subject: Does a wireless repeater affect wired AP bandwidth?
Hi all. I've set up a router as a client bridge to act as a wireless repeater, using the main wired router as the AP. I understand doing this forces the repeater to double the bandwidth used. Also the wireless repeater has a different SSID.
Does this client bridge affect the wired AP bandwidth usage? If I were to run a speed test next to the wired AP, would its speed be influenced/slowed due to having the client bridge connected, even if nothing is connected to said repeater? I mean to say, is the half bandwidth outcome only when connected to the repeater SSID, or does it occur with both repeater and main router when the client bridge has been set up? If the former, what can I do to only have this half bandwidth occur with the repeater?
I have about 27mbps (about the max for my ISP) wireless right next to the main AP, and about 13mbps next to the repeater. However, I notice the speed fluctuates when connecting to the main AP when I've recently been using the repeater (again, when I run the tests on the main AP I make sure nothing is connected to the repeater). Is this change just in my head?
This is the same as connecting a laptop to your network and not using it. It should have no impact. People are often overly concerned about the 'half bandwidth' non-issue. _________________ SIG:
I'm trying to teach you to fish, not give you a fish. If you just want a fish, wait for a fisherman who hands them out. I'm more of a fishing instructor.
LOM: "If you show that you have not bothered to read the forum announcements or to follow the advices in them then the level of help available for you will drop substantially, also known as Murrkf's law.."
This is the same as connecting a laptop to your network and not using it. It should have no impact. People are often overly concerned about the 'half bandwidth' non-issue.
Each router you add will add at least 1 hop to your route. That will add latency naturally.
The "Half bandwidth" issue stems from WDS. Since WDS specification/feature is DoA, it is of no concern to most people.
No it doesn't. It stems from splitting the signal. And people seem to think splitting it when they add a router as different from splitting it when they add an active computer. _________________ SIG:
I'm trying to teach you to fish, not give you a fish. If you just want a fish, wait for a fisherman who hands them out. I'm more of a fishing instructor.
LOM: "If you show that you have not bothered to read the forum announcements or to follow the advices in them then the level of help available for you will drop substantially, also known as Murrkf's law.."
No it doesn't. It stems from splitting the signal. And people seem to think splitting it when they add a router as different from splitting it when they add an active computer.
It also stems from WDS everywhere I read about WDS it mentions halving the bandwidth.
Adding more routers just means adding extra hops to the route. That is common sense. Everybody makes a huge big deal about 1 hop on a network when in reality it makes little to zilch difference. I am sitting behind a two routers before I even hit the ISP endpoint.
It also stems from WDS everywhere I read about WDS it mentions halving the bandwidth.
That is because wds splits the signal. So does an active repeater bridge. So does putting another active computer on the network. It certainly does not "stem" from WDS. It is a function of physics.
BTW, WDS is certainly not "DOA". It can work well for some setups and is better in many cases than the repeater bridge hack. _________________ SIG:
I'm trying to teach you to fish, not give you a fish. If you just want a fish, wait for a fisherman who hands them out. I'm more of a fishing instructor.
LOM: "If you show that you have not bothered to read the forum announcements or to follow the advices in them then the level of help available for you will drop substantially, also known as Murrkf's law.."
It also stems from WDS everywhere I read about WDS it mentions halving the bandwidth.
That is because wds splits the signal. So does an active repeater bridge. So does putting another active computer on the network. It certainly does not "stem" from WDS. It is a function of physics.
BTW, WDS is certainly not "DOA". It can work well for some setups and is better in many cases than the repeater bridge hack.
If you get it working, great. The specification is flawed (incomplete / ambigious) if manufacturers cannot interoperate with each other on WDS.
The more WDS clients you have in the mesh (at least that part of the mesh) the less bandwidth there is, to me that really is bad if you have a very busy WDS mesh. Correct me if I am wrong but WDS does not scale up well. Is this halving just with it's peers int he mesh or completely across the entire mesh network?
If you get it working, great. The specification is flawed (incomplete / ambigious) if manufacturers cannot interoperate with each other on WDS.
The more WDS clients you have in the mesh (at least that part of the mesh) the less bandwidth there is, to me that really is bad if you have a very busy WDS mesh. Correct me if I am wrong but WDS does not scale up well. Is this halving just with it's peers int he mesh or completely across the entire mesh network?
WDS works fine if the chipsets are the same, in my experience.
Only ACTIVE clients split bandwidth. It is not the network structure that does this, it is the fact that a device that is actively communicating with more than one other device has to divide its time between those devices. The same as putting two active computers on one router. Exactly the same. A WDS router that is not being used has about zero effect on bandwidth. _________________ SIG:
I'm trying to teach you to fish, not give you a fish. If you just want a fish, wait for a fisherman who hands them out. I'm more of a fishing instructor.
LOM: "If you show that you have not bothered to read the forum announcements or to follow the advices in them then the level of help available for you will drop substantially, also known as Murrkf's law.."
If you get it working, great. The specification is flawed (incomplete / ambigious) if manufacturers cannot interoperate with each other on WDS.
The more WDS clients you have in the mesh (at least that part of the mesh) the less bandwidth there is, to me that really is bad if you have a very busy WDS mesh. Correct me if I am wrong but WDS does not scale up well. Is this halving just with it's peers int he mesh or completely across the entire mesh network?
WDS works fine if the chipsets are the same, in my experience.
Only ACTIVE clients split bandwidth. It is not the network structure that does this, it is the fact that a device that is actively communicating with more than one other device has to divide its time between those devices. The same as putting two active computers on one router. Exactly the same. A WDS router that is not being used has about zero effect on bandwidth.
That right there highlights a problem with WDS, "if the chipsets are the same". WDS is a standard, probably poorly specified if it cannot be implemented correctly across chipsets.
A mesh is a mesh is a mesh, if a mesh must be the SAME hardware, then you're screwed later on as if you want to grow that mesh, you will need to dump all your hardware if it is "incompatible".
I and probably many others, simply avoid WDS becaues of these issues. Anyway, I currently subnet as a client with DD-WRT.
You can pick up amazing business gear on ebay for peanuts and lifetime warranties (HP lifetime warranties on network gear are transferrable - not many people know that, Cisco are not ). Bandwidth is cheap. Why mess with consumer gear.
). Bandwidth is cheap. Why mess with consumer gear.