I hear complaints from people all the time about their home WiFi situation. In this post, I’ll address some of the common complaints, explain a bit about how wireless works, and provide a couple of examples and solutions to better enable your decision making when it comes to outfitting your home or small business with WiFi.
WARNING: SEMI-DEEP DIVE AHEAD
My wireless doesn’t reach all the corners of my house! I’ve tried $40 routers from Wal-Mart, and $300 routers from amazon/newegg. They’re all virtually the same. Help!
Stop buying wireless routers. You’re much better of going with many smaller range, lower power Wireless Access Points (WAPs) than you are one giant multi-antenna routing monstrosity.
First of all, you’re paying for quite a few features you’re not using. For example, the actual routing portion, DHCP server, NAT, etc. Those are all things that are important at the edge of your network (where your internet provider connects to you), but not so much in the rest of the house. This can also cause problems because unless you put the device in bridge mode, then you end up having to create another network segment (say: 192.168.2.0/24 instead of the 192.168.1.0 of the rest of your network, for example). Often this will break certain applications or hardware features that rely on a single layer 2 broadcast domain, or maybe multicast traffic. Certain speaker systems, smartphone applications for home automation, etc. often won’t work if they’re not on the same layer 2 network.
Secondly, just because your wireless may be more powerful, does not mean it’s better. Something many people don’t realize is that a wireless connection, like any other type of multi-party communication, is a conversation. Cranking up the power on your router is akin to using a bullhorn to talk to someone on the other side of the parking lot. Sure, you may hear me fine if I speak through it, but unless you have a bullhorn, I won’t be able to hear your reply. In the wireless and microwave world, we use a term called EIRP (effective isotropic radiated power). It’s not important you learn what that acronym means in detail for the sake of this post, but in short it’s the combination of the power used and strength (gain) of the antenna. Increased output power only increases the strength of the signal in one direction. Using a larger antenna on the other hand, increases the strength of the signal in both send and receive (tx and rx). The problem we’ll often run into with this is normally one or more of cost, aesthetics, impracticality, or legality in extreme cases (for instance, the FCC or Federal Communications Commission, puts limits in place for the amount of EIRP on a given frequency). Something else to consider is that a larger antenna means you’ll hear more. Hear more of what? Well, everything. Not just your device, but the devices and wireless access points of everyone around you.
So, if increasing power only works in one direction, and increasing the antenna size has diminishing returns, what should you do to increase of the coverage of your home or business?
Use more wireless access points at lower output power
Keeping the output power similar to that of the devices you use has a dramatic effect on the quality of communication.
What about those wireless mesh systems. Should I use those?
There are a couple of different communication systems used in wireless mesh design. The oldest, cheapest, and most common one is based on the principle of “Store and Forward Repeating”. These are single radio, single frequency designs, and frequency is a shared medium. In the simplest terms, anything your WAP hears in a store and forward design is stored in memory, the radio waits for the frequency to clear, and then transmits this data up to whatever other WAP it’s uplinked with. Because of this design, your maximum throughput is effectively halved for each wireless mesh hop. With 2 wireless mesh hops on a store and forward design, you would effectively have a maximum throughput reduction of 1/4th that of what you would have in communicating with a single wired WAP. This problem is compounded when multiple people are connected to the same mesh system and are trying to send and receive data!
Better “mesh” designs have slowly started to work their way into the market
Multi-radio, multi-frequency designs often eliminate the problem in using a single shared medium, as now they are using 2 (or more) frequencies: 1 to talk to the client, and 1 or more to uplink to another WAP. You can still run into the same problem here though, if multiple WAPs are uplinked to the same WAP on the same frequency – you’ve simply shifted the problem to the uplink frequency instead of the access frequency your devices are using to talk to the nearest mesh WAP.
The best design for a distributed wifi system isn’t even technically mesh in the sense of needing to uplink to the rest of the network wirelessly, they’re simply wired. These systems all wire back to one or more switches, and often use the same wireless name (SSID), and often the same password. This helps eliminate throughput restrictions imposed by the width of the frequency (I’ll explain this later), the strength and quality of the uplink frequencies to each mesh WAP, etc.
In short, if you need the maximum amount of coverage and throughput, use multiple WAPs all wired to one or more switches.
I have great signal strength, but my older (802.11b, 802.11g, 802.11n) router or WAP just can’t meet my throughput needs. Should I upgrade?
802.11ac, the newest mass-adopted WiFi standard (yes I know 802.11ad is out, but that’s a very different situation), brings with it many performance enhancements while being backward compatible with existing 802.11a/b/g/n equipment.
Before I go any further, let me point out that the 802.11ac standard itself is mainly based around increasing performance of the 5GHz band. 802.11ac routers and WAPs will often be dual radio designs that support older 2.4GHz gear as well, but the vast majority of performance increases on a “pound for pound” basis come from either the increase in channel bandwidth usable in 802.11ac, or the number of spatial streams. In 802.11b and 802.11g, we were using 20MHz wide channels (aside from certain vendor-specific tweaks). In 802.11n, that possible channel size increased up to 40MHz wide. In 802.11ac, we are now looking at 80MHz and 160MHz (80+80) wide channels.
See the chart below (courtesy of Cisco):
As you can see, even in keeping with 1 spatial stream (SISO) which most cell phones and IoT devices have, we’ve seen a possible 10-fold increase in bandwidth when going from a 20MHz wide channel up to the 160MHz wide channels 802.11ac (wave2 and wave3) support. The possible maximum throughput is then again doubled for each spatial stream. As a note, this is only possible in the 5GHz band, which has far more available frequencies than the 2.4GHz band. The 2.4GHz band, as it stands in most countries, only has a maximum of 40MHz available, or one “fat” 40MHz (20+20) wide channel.
One final note here, is that for each time you double the channel bandwidth, you also decrease receiver sensitivity by half. This is something to keep in mind if you design a coverage model around the more common 20MHz wide channel, only to turn on 80 or 160MHz wide channels to find out that your coverage is now terrible. Also, 5GHz attenuates twice or more as bad as 2.4GHz does – it’s a higher frequency, and the higher the frequency, the harder it is for it to penetrate structures and environmental factors (humidity, rain, snow, etc.).
So I took your advice and bought a couple of WAPs, and now I’m trying to figure out where to put everything. Is there a method I should use to go about this?
Absolutely! In professional terms, we call it a “site survey”.
A site survey entails placing as many WAPs as you think you will need in the places you think they will give the best coverage, and then walking around in that coverage area with some kind of tool to take a look at the signal level you have at various locations. This can be as simple or as technical as you want to be. There’s an Android app called “WiFi Analyzer” which can tell you the signal level for each SSID (wireless network) in the area. This is good for a baseline. Or, you can go further and heat map the area.
A wireless heat map uses a combination of hardware (wifi device) and software (including a floor plan) to map the coverage area at various locations. In the above example, the areas with stronger signal are marked in red, and the areas with weaker signal are marked in green and blue. Depending on the settings in the software, those colors correspond to different signal level thresholds (-50, -60, etc). As a rule of thumb, I want no less than a -74 any where in my home, so I aim for the appropriate number of wireless APs to make this coverage possible on a low or medium power level.
Hopefully this blog post was informative to many of you out there who may be having problems with your WiFi, or simply had questions about certain topics. Wireless is a very deep topic, from antenna designs, log scales, frequency “personalities”, radio designs, indoor vs outdoor, modulation levels, noise filters, PHY technology, SoC capabilities, etc. I tried to keep it informative, without being too technical. In the future if you’d like me to expand on a certain topic, feel free to send me an email and I’ll try to make that happen.