In the last few years there has been a lot of activity in wireless networks. Several years after the launch of the 802.11ac (WiFi 5 ,2014), that came after 802.11n (Wi-Fi 4, 2008), the new 802.11ax (WiFi 6) is taking its place as the technology leader in the wireless market. While in the previous standards bandwidth was the main issue, in the new Wi-Fi 6, along with significant increase in the bandwidth, there are also improvements in the latency, reliability and availability, targeting new market segments as we will see later in this article.
First, let’s understand how the bandwidths of wireless networks have increased from the beginning of technology and the 802.11b/g standards to the present 802.11ac and 802.11ax standards. Something happened here, and no one changed the laws of physics.
How much bits per second we can send over a communications channel?
A well-known physical fact is that the number of bits that can be passed over a certain frequency range, also called spectral efficiency (Spectral Efficiency – Bits / Sec / Hz) has a physical limitation. This limitation was formulated about a century ago by the Shannon-Hartley Theorem, which states that the amount of Bits / Sec that can be transmitted over a frequency range is:
When C is the bandwidth in Bits/Sec, B is the modulated frequency range which is the frequency band on which we are allowed to transmit and receive, and S/N is the signal-to-noise ratio, i.e. how noisy the environment is.
Since the laws of physics have already been set before us, what we can do to increase the rate of transfer is:
- Use wider frequency ranges (as seen in the fifth generation) or use multiple frequency ranges simultaneously.
- Modulate with smarter mechanisms and thus transfer more bits per frequency. When the modulation and coding are smarter, the received signal can be better decoded and thereby improve the signal-to-noise ratio. As the devices became more powerful over the years, so too could stronger mechanisms be used.
- Use multiple antennas for transmission and reception (MIMO technology) and then get bandwidth multiplications.
- Transmit in the exact direction of the device and do not dissipate the transmission energy in all directions (Beamforming technology) and thus improve the signal-to-noise ratio.
Let’s see how we use these possibilities to increase the bandwidth over wireless networks.
Frequency ranges and channel width
The width of the channel refers to the range of frequencies at which it is operated. The width of the channel is configured according to the following table and can be set manually or automatically in the wireless network controller. Note that bandwidth should be enabled both by the Access Point and the end node so it is important to verify that the stations will work in the maximum range (usually synchronizes automatically).
Because the larger the frequency range, a higher bandwidth (Bits/Sec) can be transmitted to it, so the bandwidths at 802.11ac and 802.11ax are higher.
What are the frequencies at which it can transmit?
802.11 standards define the use of several frequencies: 900MHx, 2.4GHz, 3.6GHz, 4.9GHz, 5GHz, 5.9GHz and 60GHz. Each frequency has a frequency range in which the transmission is defined. The frequencies accepted and supported by most types of equipment are 2.4GHz where a frequency range of 2.400-2.483GHz is defined (in North America up to 2.473GHz), and the 5GHz frequency where several frequency ranges are defined from 5.170GHz to 5.240GHz (U-NII-1), 5.260GHz-5.320GHz (U-NII-2) and more.
Channels are defined in these frequency ranges, with each channel being 20MHz, 40MHz (starting from 802.11n), 80MHz or 160MHz (starting from 802.11ac). When operating a WiFi network it is important that there is no overlap of the same frequency between two adjacent areas so as not to create interference. In a network where a controller is defined, the controller is responsible, among other things, for managing the frequencies.
In the table above we see the frequency ranges according to the different standards. Some frequencies are allowed in some countries. The 2.4GHz is the most popular while the 5GHz is also allowed in many countries.
Stronger modulation and modulation efficiency
Let’s talk about some basic terms in multiplexing and modulation: OFDM, OFDMA and QAM.
Orthogonal Frequency Division Multiplexing (OFDM) is multiplexing method in which the defined channel is taken (for example, a 20MHz or 40MHz channel), in which several carrier frequencies are transmitted. Each carrier frequency will be further modulated by the transmitted information. For example, in 802.11ac each 20MHz channel has 64 sub-carriers, a 40MHz channel has 128 carrier and so on. The more subject waves there are, the more information we can convey.
In OFDM, as we see on the left, every timeslot is modulated by a single user. In the timeslot is for example 10µSec, every 10µSec will be modulated by a single user. We see this on the left when user-1 modulates first, then user-2m user-3 and so on.
Orthogonal Frequency Division Multiple Access (OFDMA) is a smarter multiplexing method, where the modulated waves can be divided between number of users. If in OFDM a certain computer (Client) transmits in front of the AP, then as long as a certain the user is transmitting the channel is busy for all other users.
In OFDMA, as we see on the right, the same timeslot can be used for multiple users when each gets separate Carriers. For this reason, when there are many users, we get a higher bandwidth than OFDM, when the latency also decreases, since we don’t have to wait to the next timeslot.
When multiplexing is about dividing the channel between different users providing them different timeslots and frequencies, modulation is about how to carry bit/sec over the frequencies.
Quadrature Amplitude Modulation (QAM), used in Wi-Fi and cellular networks, is a modulation method in which we change the intensity and occurrence of the carrier wave, and in this way many bits in each transmission. As we see in the following image, according to the QAM level so is the number of bits that can be transferred at any given moment. The principle is that instead of transmitting one bit per transmission, we will transmit 4, 16, 64 or more. On the left we see QAM in order of 4, QAM-4, where in each transmission (in each signal) we pass 2bits (00,01,10 or 11), on the right we see QAM in order of 16, QAM-16, where there are 16 different options for each signal, from 0000 to 1111.
802.11n uses up to QAM-64 modulation, 802.11ac uses up to QAM-256, and 802.11ax uses up to QAM-1024. As we increase the QAM level we will of course get a higher transmit/receive rate, but as we saw in the Shannon-Hartley formula as the amount of noise increases the transmission rate in bits/sec will decrease. The higher the order of the QAM, the greater the difference between the signals (angle and intensity), and the higher the possibility that the signal will be decoded incorrectly.
As we see in the next figure, as we rise in the modulation level the efficiency will not increase in direct proportion but less than that. What enables the raising of the QAM level are of course smarter algorithms and more powerful processors, and as we progress, we will get more efficient and powerful modes. The parameter that measures this is called Modulation Efficiency or Spectral Efficiency, which is measured in Bits/Sec/Hz.
In the graph above we see that as we rise in the modulation level, we get higher efficiency. At the high modulation levels and in good signal/noise ratio we get up to 10-12Bits/Sec/Hz and even more.
Multiple Input Multiple Output (MIMO)
Multiple Input Multiple Output (MIMO) is a wireless technology that increases the data capacity of a RF radio by using multiple transmitting and receiving antennas. MIMO is relatively old technology, which was already implemented in 2007 in the 3GPP R.7 standard (HSPA+) in cellular networks and shortly afterwards in 802.11n.
In 802.11n came out Single User MIMO which allows transmission up to 4 antennas for a single user and allows a significant increase in bandwidth in front of the same user.
802.11ac introduced Multi-User MIMO technology that enables transmitting and receiving of multiple antennas to multiple end users.
Beamforming is a technique by which an array of antennas can be steered to transmit radio signals in a specific direction. Rather than simply broadcasting energy in all directions, the antenna arrays that use beamforming, determine the direction of the end-user, and sends and receive a stronger beam of signals in that specific direction.
In the figure we see an example to beamforming. The way it works if by changing the amplitude and phase of the antenna arrays.
What’s new in Wi-Fi 6?
To see what is new in Wi-Fi 6, let’s have a look in the following table.
First, we see higher order modulation, more MIMO streams and Multi-User MIMO that brings to higher link efficiency and higher bandwidth, but this is not necessarily the most important thing.
Another thing we see from the table is OFDMA and lower OFDMA spacing. Both brings both higher bandwidth but also lower latency, that is more important to real-time applications like IP telephony, conferencing and so on. If you look for example on IP Telephony over Wi-Fi you will see that there was dramatic increase in Voice over Wireless LAN (VoWLAN).
There are also mechanisms like BSS coloring for eliminating inter-AP interference, Quality of service mechanisms, better support for small packets and more.
How to design a wireless network
Now that we have talked about technology, let’s get into planning. We will look at the design from two perspectives – the client and the planner.
For the customer, the issue is simple. All a that a smart customer needs to do is set requirements and let the engineer of the vendor who installs the network do the planning according to the requirements. This design way comes from the fact that each manufacturer has its own APs, its antennas and so on. The requirements we will define will be:
- Functional requirements – the number of SSIDs, security mechanisms, working with or without a controller, how to connect to the corporate network and a few dozen more functional requirements.
- Performance requirements – bandwidth, delay, jitter (changes in delay) and the allowed packet loss level. It is important to define in situations of rest and movement, when in movement it is important to define the speed required by the organization, for example a person walking inside a building, a vehicle traveling at low speed, etc. It is possible (and desirable) to also set radio data as reception intensities at any point in the organization where network access will be required.
- Information security requirements – When we transmit to the air, the transmitted information can be listened to, so it is important to check the level of security required from two aspects – authentication of users entering the network (Encryption) and encryption of the information passing through the network. There are new security standards and technologies, for example WPA3. I will write about information security in wireless networks in a separate article.
- Marketing/Management Requirements – Here are some non-technical requirements. For example, a chain of stores will ask that everyone who connects to the network receives pop-up advertisements on a computer, a hotel will request a billing option for connecting to the network, an organization with employees who visit the site can request location and location services for employees, and so on.
What is important to note is that the bandwidth we see in the standards is in optimal condition, that is when there are maximum antennas (also for the end units), the environment is not noisy and more.
It is also important to note that as we rise in frequency, the distance get smaller in inverse squares ratio, i.e. when we rise from 2.4GHz to 5GHz, the distance is theoretically four time smaller. In practice, the distance will be higher that due to mechanisms such as MIMO and Beamforming.
Another consideration is how much bandwidth we really need. It is quite easy to get carried away for maximum design, which will give us extremely high bandwidths. The question is whether we need it, and how much it will cost us. If for example Wi-Fi is intended for a network of industrial controllers, reliability is important, but the bandwidth is usually extremely low, whereas if the network is intended for viewing stationary cameras then you need high bandwidth but only for places where there are cameras.
Another issue is whether we should use the latest technology or not. You might be surprised of this question but on the time this article is written (January 2021) the latest Wi-Fi technology is Wi-Fi 6 (802.11ax) that provides about 10Gbps per AP, and the one before it is WiFi 5 (802.11ac) that provides about 7Gbps per AP, with access points that cost 30-40% less then the new ones. Wi-Fi 6 APs provides. If the requirement is bandwidth, 802.11ac will do the job. If the requirement is low latency (for example for VoWLAN) and you are having strong budget constrains – take 802.11ax and work with 2.4GHz to increase the distances and decrease the number of APs.
Cellular or wireless?
As cellular networks become faster and cheaper (especially with the advent of fifth-generation technologies), it may well be, especially for out-of-building network deployment, that we should purchase a cellular service, perhaps even set up a private network. There are technological and economic/business considerations for and against each one of the solutions.
When there is a possibility and there is no impediment to working in a cellular network (for example from information security constraints), it is worth examining the possibility. Although we do not have a commitment to performance in a cellular network, we usually get good performance. It also depends on the number of end units – when it comes to several hundred units, at a price of few USDs per Data SIM, it could very well be that the cellular option would be worthwhile. Another advantage that the cellular network is the responsibility of the supplier and the whole issue of maintenance is in the provider responsibility.
In places where communication is critical, you should also consider cellular backup to the wireless network, in such a way that a fall or poor reception in the wireless network is a response by the cellular network. There are many routers today that give this a solution.
Like the technical aspects, the “how much does it cost” question has an equally important meaning. In terms of standard, almost any equipment you purchase will support all devices. The high costs will be in the controller (controller), in software licenses (security, location, etc.).
In terms of the number of APs, if you do not need high bandwidths, you should consider using 2.4GHz to reduce the number of APs. If controller costs are expensive, there are many manufacturers that sell controller services in the cloud at a monthly cost, and if there are no security restrictions you should consider this.
In terms of backups, it is important to truly assess how critical the network is. You can cram the deployment of APs ($ 300 to $ 800 or more per unit, depending on the model and whether internal or external), use a primary controller and a secondary controller (thousands to tens of thousands of dollars), use cellular backup, and double the lines that connect to the corporate network. Each of these options cost money, sometimes a lot of money, and the question of whether you really need it and what it means to fall into the net.
In this article we talked about the various technologies in wireless networks, emphasizing the new standards, especially Wi-Fi 6. As we move forward with the time, more technologies are developed, and there are many interesting one we are going to see in coming year. Keep follow our knowledge base for things to come.