Wifi (or Wireless Fidelity) is a communication standard for Wireless Local Area Network (WLAN). This standard is termed as IEEE 802.11. Wifi works in Physical Data Link Layer of OSI Model. Wifi carries data over of Radio Wave Carrier.
Types of Radio Waves
Waves as we know, transmits energy/information from one point to another in the form of oscillation. There are three types of waves. First is Mechanical – this requires a medium for transfer. Example, waves over water, sound waves etc. Second is Electromagnetic Waves – it is created with fusion of magnetic and electric fields. These waves do not require medium to travel. Ex. Light Waves, Radio Waves, etc. Third is Matter Waves – matter can exist both as waves and particle.
We are concerned here about only Radio Waves. Below figure explains various Electromagnetic Waves :-
Light waves are just in the middle between Gamma Rays and Radio Waves. Radio Waves (includes Microwaves) have longer wavelength and thus can travel lesser distance because of lower energy in comparison to say Gamma rays or Light rays.
Position of Wifi Signals in Radio Waves
Next we will see where does Wifi stand in Radio Waves as below :-
As we can see, Wireless LAN occupies just a small fraction of the Microwaves (Radio Waves) section. More so it is overlapping with C-Band Satellite and some Space borne and Martime Satellites. Specifically currently Wifi Signals works over 2.4 GHz range and 5 GHz range. In near future it will also work over 6 GHz range (Wifi 6E). Radio Waves operates in 30 Hz – 300 GHz range.
Wifi Signal Line of Sight
Wifi signal does not always get a clear Line of sight from Transmitter (Point A) to Receiver (Point B). Below diagram explains this :-
Wifi waves are absorbed into following objects. Hence, placement of routers/acess point really matters.
Dry Objects < Brick < Concrete < Water < Metal. Example, Refrigerator has metal body and water content. Any wifi router kept over/near that will not be able to work properly.
Reflection/Diffraction : Wave encounters hindrance and gets reflected. Example, Computers may receive signal not directly from source but as reflection. Similarly, it may get diffracted if the receiver has sharp objects near them.
There are two modes of wireless signal : Adhoc WLAN and Centralized WLAN. Adhoc WLAN are Mobile Hotspots – peer to peer. As more users join speed decreased and security also.
Wifi was introduced in 1997. A bunch of Wifi Protocols have since then have been established – IEEE 802.11b, a, g, n, ac, and now ax. To avoid complex names, IEEE has revised alternate names starting from Wifi 1 (802.11b) to Wifi 6 (802.11ax). IEEE is currently working on Wifi 6E (E for Extended) – on 6 GHz band.
All wifi protocols are half duplex – can either transmit or receive at a time. This is unlike Ethernet which can do both at same time. Hence, Ethernet LAN CAT cables are considered GOLD standard. But surely, Wifi is catching up now as people are moving towards mobile technology.
Below, table shows comparison of all Wifi Protocols in details :-
Frequency Band – 2.4 GHz
The 2.4 GHz Band Channel are the most used channel in Wifi. Practically, all devices works over 2.4 GHz for wireless transmission. Be it Radio Control, Phones, Bluetooth, Microwave Oven, AV devices etc.
Below figure shows entire spectrum of 2.4 GHz :-
There are 14 Channels for signal transmission. In some countries, channel 14 is not used. Each Channel is about 20-22 MHz wide. If router is connected to channel 5, other router in the vicinity is in channel, it may cause interference among them. It happens mostly in crowded vicinity with many routers. The best combination is Channel 1, 6, 11 as all are separated by some gaps. Each Channel is separated by Guard Interval of 5 MHz.
Frequency Band – 5 GHz
It was introduced in 2003 by 802.11a standard (Wifi 2), and since then many channels have been developed. Below figure clears the entire 5 GHz spectrum :-
The entire spectrum of 5 GHz is divided into 4 subsections. Section 1 & 4 (in Green) are mostly allowed. Section 2 & 3 (Light Blue) are allowed with restrictions in some countries and require firmware router configuration. As we can see, 25 channels of each 20 MHz width channels are available for use.
Channel Bonding allows 2 and more than 2 channels to bond to make bigger channel. So, Channel 36 and 40 are bonded to form Channel 38 with 40 MHz frequency. Similarly, channel 38, 46 are bonded to form channel 42 on 80 MHz etc.
The light blue section are accessible to routers with DFS. DFS stands for Dynamic Frequency Selection. This allows unlicensed devices to access already allocated frequency bands without interference to the existing radar system. This means if it detects an overlapping radar, then the router vacates that channel and selects an alternate channel.
Most routers don’t have DFS but maximum client PC or mobile already has DFS. As we can see only two channels of 80 MHz are accessible to routers with DFS. Also, since other routers will be using sub-channels also, it may become crowded. With DFS, 80 MHz channel can access total 6 channels – 42, 58, 106, 122, 138 and 155. Also, 160 MHz channels also becomes available – 50, 114.
Hence, at present any router with maximum DFS channels and MIMO technology is the best.
Multiple In Multiple Out (MIMO)
MIMO was introduced in 2009 with 802.11n and since then has evolved. With MIMO, speed and reliability was increased. MIMO is expressed as A x B. Mostly A = B. So, A means Transmitter and B means Receiver. A x A implies A number of receiver and A number of transmitter. But, MIMO speeds are as good as client devices. So, if Router supports 4×4 and client devices is 2×2 then MIMO will work over 2×2 only.
There are two ways of using MIMO : Spatial Multiplexing and Space Time Coding.
In Spatial Multiplexing, data is split at transmitter into multiple streams and then recombined at receiver end by MIMO technology. Every stream transmits unique data to increase the speed. A 3×3 MIMO will transmit unique data into 3 spatial stream. So, speed increases 3 times. But the bandwidth remains the same. It is like multiple flyovers over the base road. Nowadays only Spatial Multiplexing mostly used at consumer level.
In Space Time Coding, same data is sent through multiple antenna and received via similar antenna on the receiver device. This ensures reliability but not the speed.
We need carrier waves to transfer a data. Modulation is the process of changing some property of the carrier wave – amplitude or frequency of the carrier wave with data signal as input.
Why do we need modulation : Data signals are low frequency waves and are susceptible to interception that high frequency. Modulation is a way to convert them in to high frequency waves. High Frequency have high bandwidth and can carry more data also. Low frequency requires large antenna and can travel high distances.. And, high frequency waves require smaller antenna but can travel small distance only. So we see X-rays are used close to body for scans. Also, FM can carry more data but distance is lower than AM. But, FM has better sound quality.
Below figure shows the modulation of data signal using ASK, PSK, FSK :-
CCK (Complementary Code Keying) – PSK
Below diagram shows the data modulation of CCK :-
As we can see, each 4 bit blocks are taken from the data stream and separated to into two blocks. First two blocks are phase modulated: 00 – 0 angle, 01 – 900, 10 – 1800, 11 – 2700 degrees. Next two blocks are given code words a CCK word: 00 – +i, +1, +i, -1, +i, +1, -i, +1, and like wise for 01, 10, 11. This way data is modulated for 5.5 Mbps over carrier wave. A similar but more complex method employed with 11 Mbps transmission. This is a phase modulation of the carrier wave – a more complex version of QPSK.
PSK is Phase Shift Key Modulation. Other PSKs used in modulation are – BPSK, QPSK, 8PSK, 16PSK, DPSK, DQPSK, etc.
An ASK (Amplitude Shift Key) Modulation takes place by changing amplitude of the wave. Example: 1 may represent HIGH for crest and 0 represent LOW for trough.
QAM (Quadrature Amplitude Modulation)
QAM is a combination of ASK and PSK over a single channel. It maps the bit data into a constellation diagram. This diagram x-axis and y-axis is defined by two vectors I and Q. I represents a cosine function and Q represents a sine function.
Below figure shows the constellation diagram :-
As we can there are 4 bits represented here (orange). Bits are laid just like coordinates. To read a coordinate specific Amplitude and Phase modulation is required. Below diagram explains working as below :-
Here, QAM bit (A3, B4) = A3 * cos(wct) + B4 * sin(wct). Above constellation diagram is a 64 QAM as 64 bits could be represented. Each quadrant is a 16 bit representation. Therefore, 26 bits = 64 points. So, each point is represented by 6 bits of information (in 0’s and 1’s) in 64 QAM. Similarly, 256 QAM, a point is represented by 8 bits as 28 bits = 256 points. Currently, 1024 QAM is operable in 802.11ax.
Below figure shows the QAM modulation and demodulation procedure :-
Below table shows the available QAM modulation points. Constellation point 25, 27 are not used due to noise and signal ratio not achieving a significant upgrade.
|Constellation Points||Modulated Bits (per point)||QAM|
|1||2o -> 0/1||BPSK|
|4||22 -> 2 bits||QPSK|
|16||24-> 4 bits||16 QAM|
|64||26-> 6 bits||64 QAM|
|256||28-> 8 bits||256 QAM|
|1024||210 -> 10 bits||102 QAM|
Also, the QAM reduces the range of signal. So, 1024 QAM signal is more susceptible to noise interference than 256 QAM. Hence, the realistic speed may vary from the marketing hype. But there are other hypes too which we will discuss later below.
MCS (Modulation and Code Rate)
MCS is defined as useful bits which can be carried by 1 symbol. This is the useful or the parity bits.
A code rate is = Useful Bits / Total Bits (Useful + Redundant bits)
Redundant bits are for the error correction. So, 64 QAM 3/4 means 3 bits are useful and 1 bit is a redundant bit. Or in every 4 bit stream only 3 bit is usfel and 1 bit is the redundant bit meant for error correction.
QAM Data Rates
Below table shows 802.11a, b, g, n ,ac, ax for easy reference :-
20 MHz channel upto 64 QAM 3/4 is available on 802.11 a/g. On 802.11n, 64 QAM 5/6 MCS is available. Further with channel bonding 40 MHz channels are available on the 802.11n. We can also see, with each MIMO increase from 1×1 to 4×4, the speed is increased 4 times. Please note 256 QAM and 1024 QAM is a non-standard and may not be supported. So, do not fall over the marketing hype if any by manufacturers.
802.11 ac and ax for 80 MHz :-
With Channel Bonding, 80 MHz Channel width is achieved in 5 GHz bad. As can be seen 802.11ac supports max. speed of 1733 Mbps at 4×4 MIMO 256 QAM 5/6, while 802.11ax supports 4803 Mbps at 8×8 1024 5/6 QAM..
802.11ax 20 and 40 MHz :-
About only 70% of the above speeds are realistic speed. As the rest 30% of the speed goes to the Wifi overhead.
Let us take some examples to understand the marketing hype on routers. Router AC5300 – “Router can run upto 5300 Mbps speed”. First of all 5300 is the combined of all the speed it can support on both 2.4 and 5 GHz. But the router is triband router. So, it can support 2.4, 5 and one more 5 GHz bands (802.11ac). On 2.4 GHz it supports 1000 Mbps (802.11n 40 Mhz, 1024 QAM 4×4 = 1000 Mbps (non standard). Now on 5 GHz, 1024 QAM 5/6 4×4 = 2166 MHz. Other 5 GHz channel also supports 2166 Mbps. So, on total = 1000 + 2166 + 2166 Mbps = 5300 Mbps.
But if we take realistic standard speed it will be way different. So on 2.4 GHz example – 1000 Mbps on 2×2 client will be realistically, 64 QAM 5/6 = 300 Mbps and 150 Mbps on 1×1 client. Similarily, on 5 GHZ – 2166 Mbps will be 256 QAM 5/6 = 866 Mbps and more realistically at above 32 feet distance = 650 Mbps and 4333 Mbps on 1×1 client. So most important devices are your client devices to decide the speed selection.
The list is endless but below figure will show the breakup of major router speeds :-
Frequency division Multiplexing
A multiplexing technique involves multiple subcarriers to use a single channel. So multiple carrier waves can be transmitted in parallel. These carriers are separated by guard band and are not overlapping. Example, in cable television broadcasting.
OFDM (Orthogonal Frequency Division Multiplexing)
In OFDM, the subcarriers are in overlapping state without the guard band and modulated in a way that when one is passed others at 0. Below figure shows OFDM model :-
As can be seen, at any point in time, only one carrier has high or low amplitude (power), the other two carrier is then at 0. This makes better utilization of the same bandwidth and gives better throughput. Since 802.11a, g, n, ac use the OFDM technique.
OFDMA (Orthogonal Frequency Division Multiple Access)
OFDMA is more advanced version of OFDM. With this, every user will be allocated a part of the subcarrier such that delay time response is reduced. Below figure shows major difference between OFDM and OFDMA:-
Currently, OFDMA works only with 802.11ax. And both router and end device should have 80.211ax capability.
FHSS : Frequency Hopping Spread Spectrum
FHSS is an legacy multiplexing technique. It allowed multiple carrier to exist within the same time frame by hopping the frequencies. Introduced in 1940, it helps in secure transmission of radio waves mostly used in military then. But later was reused in Wifi transmission for reducing overlaps.