October 13, 2014 at 18:17New Wi-Fi for the Internet of Things (Part 1)
- From the sandbox
- Tutorial
At the end of the last century, computers drove people out of many areas of activity related to the performance of routine operations. United by the Internet, computers have broken the boundaries of information dissemination.
In the early 2000s, social networks and mobile gadgets united people from all over the world: replacing personal communication, they presented us with a network that would be more correct to call the “Internet of People”.
Both phenomena, computers and the Internet, unpredictably changed even people's habits, prompting us to check mail in the morning before brushing our teeth. Without a doubt, the continued desire to automate everything that is possible will radically change the economy, politics and personal life. Analysts believe that the next revolution in ICT will be connected with the Internet of things - an ecosystem of billions of autonomous devices: sensors, controllers, machine tools, devices, etc. The scale of the revolution is evidenced by the fact that, according to forecasts, the total number of devices in such a network will reach 50 billion by 2020. Obviously, the best way to connect so many devices is with a wireless network.
While futurists predict our lives in 2020, engineers and researchers are asking if modern wireless technologies can withstand such a huge number of devices, most of which are powered by rechargeable batteries.
Although today there are already a number of technologies for personal wireless networks based on RFID, ZigBee, Bluetooth that support low-power devices, their capabilities are limited by the number of devices on the same network, bandwidth, range and other parameters. On the other hand, urban and regional network technologies, such as today's WiMAX and LTE, are also not suitable for the Internet of things due to their high energy consumption and relatively high cost of use. That's why 3GPP, IEEE and other international organizations today are busy trying to adapt their technology to the needs of the Internet of things.
Consider, for example, IEEE and IEEE 802.11 networks known as Wi-Fi. This wireless technology was originally created for high-speed connection of a limited number of stations located in a room at a short distance between each other. Therefore, it is not effective for transmitting short messages by a large number of IoT devices remote from each other. To meet the requirements of the Internet of Things, the IEEE 802 LAN / MAN Standard Committee (LMSC) created the IEEE 802.11ah Task Group ah, TGah, which aims to expand the scope of IEEE 802.11 networks by developing an energy-efficient protocol that allows thousands of stations located in indoors, and outside it, work in the same frequency-spatial region.
In this series of articles, we will look at the activities of TGah and its tasks, as well as the main mechanisms that will be included in the IEEE 802.11ah (.11ah) add-on. Work on .11ah should be completed at the beginning of 2016, which means that in about a year the first devices using the new technology will appear on the market.
In 2010, having studied the properties of the range <1 GHz (S1G), the Committee defines this range as promising for Wi-Fi devices operating outdoors. However, due to the lack of available spectrum, S1G does not allow the use of wide bands, especially> 20 MHz, introduced in .11n and .11ac. However, signal-code constructions (CCMs) developed in .11ac and adapted to narrow .11ah channels can provide hundreds of MB / s if the channel conditions are good enough.
At the same time, S1G has definitely better radio wave propagation characteristics in street scenarios compared to the 2.4 and 5 GHz traditional Wi-Fi range, which increases the radius of the coverage area to 1 km with a standard transmission power of 200 mW. Effective CCMs and favorable propagation characteristics of radio waves allow using S1G to build sensor networks that are superior to ZigBee, Bluetooth, NFC, etc. in throughput and coverage, while remaining very energy efficient.
So the main group of use cases for .11ah is as follows:
In all of these applications, an access point (AP) spans hundreds or even thousands of devices — sensors or controllers that transmit short packets from time to time. A huge number of stations fighting for access to the channel leads to collisions, and the use of standard packets for sending short messages increases the overhead caused by the relatively long packet headers. Despite the fact that the required aggregate throughput in the considered scenarios does not exceed 1 Mbit / s, all these features reduce the efficiency of using channel resources. An equally important problem is the need to reduce energy consumption, since the sensors are mainly powered by batteries.
IEEE 802.15.4g wireless devices widely used in industry can run on battery power for a long time, but the range and available data rates are very low. Therefore, the second group of scenarios is the construction of a transport communication network between IEEE 802.15.4g sensors and remote servers. IEEE 802.15.4g routers collect data from devices and relay them to servers over a .11ah network. In other words, .11ah extends the network coverage of .15.4g. In addition, since IEEE 802.15.4g speeds are insufficient for transmitting video streams, IEEE 802.11ah can also be used in these scenarios to transmit images from surveillance cameras.
High bandwidth and a large coverage area make S1G attractive for increasing the coverage area of a Wi-Fi access point and for reducing the load of mobile networks (mobile traffic offloading), which is a successful solution to the problem of telecom operators arising from the ever-increasing volume of mobile traffic. Although .11n and .11ac networks have comparable data rates to LTE (or even higher) data rates, these technologies can hardly be used to offload mobile networks outdoors because of the small radius of radio visibility. In contrast, IEEE 802.11ah will be very useful, especially in countries with a wide available S1G channel, for example, the USA.
The table below contains the basic requirements of the use cases described above:
Having studied S1G regulation in various countries, TGah ran into two problems.
The first problem is that the bands available in S1G for industrial, scientific and medical communications vary from country to country. The current version of the draft standard defines which channels should be used in the United States, Europe, Japan, China, South Korea and Singapore.
The second problem is the lack of free frequencies. Therefore, the channels used in .11ah are 10 times narrower than in .11ac (one of the latest additions to the Wi-Fi standard): 1, 2, 4, 8, and 16 MHz (Only 1 and 2 MHz are mandatory). Moreover, the physical layer of the IEEE 802.11ah standard is inherited from .11ac and adapted to the available S1G bands.
For channels> = 2 MHz, CCMs are 10 times slower CCMs of the .11ac standard, that is, OFDM symbols in .11ah are 10 times longer than in .11ac, while the number of subcarriers in .11ah channels is the same, as in the corresponding .11ac channels. For example, both channels - 2 MHz in .11ah and 20 MHz in .11ac - contain 64 subcarriers, of which only 52 transmit data. For a 1 MHz channel, the total number of subcarriers is two times lower, but only 24 of them (which is less than 52/2) transmit data.
.11ah inherits from the .11ac 10 CCM standard (called MCS0, ..., MCS9) with different speeds and reliability. To expand the channel range of 1 MHz, the standard defines a new CCM MCS10, which is nothing more than a modification of MCS0 with double repetition, which increases transmission reliability. Preliminary studies show that, thanks to MCS0, outdoor .11ah will allow data to be transmitted in the 1 MHz channel at 200mW for a distance significantly greater than 1 km.
The available data rates for different channels and CCMs are listed in the table below:
These rates can be improved by reducing the OFDM symbol duration and using multiple spatial streams (MIMOs).
So, the duration of the ordinary OFDM symbol in .11ah networks is 40 μs, 20% of which is a guard interval containing redundant data and preventing intersymbol interference. IEEE 802.11ah allows you to halve the guard interval, which increases the data transfer rate by 11%.
IEEE 802.11ah stations can use up to 4 spatial streams. As you know, N spatial streams increase the data transfer rate N times. Thus, the maximum data transfer rate in .11ah networks reaches almost 350 Mbps. However, 350 Mbps is the maximum data rate measured at the physical layer. In fact, if you do not change the data link protocols, then the data transfer rate, say, at the network level will be significantly lower.
Why this is happening and how TGah changed the link layer protocols to reduce protocol loss will be described in the next article.
In the early 2000s, social networks and mobile gadgets united people from all over the world: replacing personal communication, they presented us with a network that would be more correct to call the “Internet of People”.
Both phenomena, computers and the Internet, unpredictably changed even people's habits, prompting us to check mail in the morning before brushing our teeth. Without a doubt, the continued desire to automate everything that is possible will radically change the economy, politics and personal life. Analysts believe that the next revolution in ICT will be connected with the Internet of things - an ecosystem of billions of autonomous devices: sensors, controllers, machine tools, devices, etc. The scale of the revolution is evidenced by the fact that, according to forecasts, the total number of devices in such a network will reach 50 billion by 2020. Obviously, the best way to connect so many devices is with a wireless network.
While futurists predict our lives in 2020, engineers and researchers are asking if modern wireless technologies can withstand such a huge number of devices, most of which are powered by rechargeable batteries.
Although today there are already a number of technologies for personal wireless networks based on RFID, ZigBee, Bluetooth that support low-power devices, their capabilities are limited by the number of devices on the same network, bandwidth, range and other parameters. On the other hand, urban and regional network technologies, such as today's WiMAX and LTE, are also not suitable for the Internet of things due to their high energy consumption and relatively high cost of use. That's why 3GPP, IEEE and other international organizations today are busy trying to adapt their technology to the needs of the Internet of things.
Consider, for example, IEEE and IEEE 802.11 networks known as Wi-Fi. This wireless technology was originally created for high-speed connection of a limited number of stations located in a room at a short distance between each other. Therefore, it is not effective for transmitting short messages by a large number of IoT devices remote from each other. To meet the requirements of the Internet of Things, the IEEE 802 LAN / MAN Standard Committee (LMSC) created the IEEE 802.11ah Task Group ah, TGah, which aims to expand the scope of IEEE 802.11 networks by developing an energy-efficient protocol that allows thousands of stations located in indoors, and outside it, work in the same frequency-spatial region.
In this series of articles, we will look at the activities of TGah and its tasks, as well as the main mechanisms that will be included in the IEEE 802.11ah (.11ah) add-on. Work on .11ah should be completed at the beginning of 2016, which means that in about a year the first devices using the new technology will appear on the market.
Use cases
In 2010, having studied the properties of the range <1 GHz (S1G), the Committee defines this range as promising for Wi-Fi devices operating outdoors. However, due to the lack of available spectrum, S1G does not allow the use of wide bands, especially> 20 MHz, introduced in .11n and .11ac. However, signal-code constructions (CCMs) developed in .11ac and adapted to narrow .11ah channels can provide hundreds of MB / s if the channel conditions are good enough.
At the same time, S1G has definitely better radio wave propagation characteristics in street scenarios compared to the 2.4 and 5 GHz traditional Wi-Fi range, which increases the radius of the coverage area to 1 km with a standard transmission power of 200 mW. Effective CCMs and favorable propagation characteristics of radio waves allow using S1G to build sensor networks that are superior to ZigBee, Bluetooth, NFC, etc. in throughput and coverage, while remaining very energy efficient.
So the main group of use cases for .11ah is as follows:
- smart meters (gas, water and energy consumption);
- smart energy systems (Smart-grid, for Russia with its bowels it is of little relevance, but in the West ...);
- monitoring of the environment and agricultural land (temperature, humidity, wind, water level, environmental pollution, animal condition, detection of forest fires, etc.);
- automation of production processes (extraction and processing of oil, ore, chemical and pharmaceutical industries, etc.);
- health systems / fitness system (remote measurement of blood pressure, heart rate, weight);
- care system for the elderly and newborns;
- smart House.
In all of these applications, an access point (AP) spans hundreds or even thousands of devices — sensors or controllers that transmit short packets from time to time. A huge number of stations fighting for access to the channel leads to collisions, and the use of standard packets for sending short messages increases the overhead caused by the relatively long packet headers. Despite the fact that the required aggregate throughput in the considered scenarios does not exceed 1 Mbit / s, all these features reduce the efficiency of using channel resources. An equally important problem is the need to reduce energy consumption, since the sensors are mainly powered by batteries.
IEEE 802.15.4g wireless devices widely used in industry can run on battery power for a long time, but the range and available data rates are very low. Therefore, the second group of scenarios is the construction of a transport communication network between IEEE 802.15.4g sensors and remote servers. IEEE 802.15.4g routers collect data from devices and relay them to servers over a .11ah network. In other words, .11ah extends the network coverage of .15.4g. In addition, since IEEE 802.15.4g speeds are insufficient for transmitting video streams, IEEE 802.11ah can also be used in these scenarios to transmit images from surveillance cameras.
High bandwidth and a large coverage area make S1G attractive for increasing the coverage area of a Wi-Fi access point and for reducing the load of mobile networks (mobile traffic offloading), which is a successful solution to the problem of telecom operators arising from the ever-increasing volume of mobile traffic. Although .11n and .11ac networks have comparable data rates to LTE (or even higher) data rates, these technologies can hardly be used to offload mobile networks outdoors because of the small radius of radio visibility. In contrast, IEEE 802.11ah will be very useful, especially in countries with a wide available S1G channel, for example, the USA.
The table below contains the basic requirements of the use cases described above:
Physical level
Having studied S1G regulation in various countries, TGah ran into two problems.
The first problem is that the bands available in S1G for industrial, scientific and medical communications vary from country to country. The current version of the draft standard defines which channels should be used in the United States, Europe, Japan, China, South Korea and Singapore.
The second problem is the lack of free frequencies. Therefore, the channels used in .11ah are 10 times narrower than in .11ac (one of the latest additions to the Wi-Fi standard): 1, 2, 4, 8, and 16 MHz (Only 1 and 2 MHz are mandatory). Moreover, the physical layer of the IEEE 802.11ah standard is inherited from .11ac and adapted to the available S1G bands.
For channels> = 2 MHz, CCMs are 10 times slower CCMs of the .11ac standard, that is, OFDM symbols in .11ah are 10 times longer than in .11ac, while the number of subcarriers in .11ah channels is the same, as in the corresponding .11ac channels. For example, both channels - 2 MHz in .11ah and 20 MHz in .11ac - contain 64 subcarriers, of which only 52 transmit data. For a 1 MHz channel, the total number of subcarriers is two times lower, but only 24 of them (which is less than 52/2) transmit data.
.11ah inherits from the .11ac 10 CCM standard (called MCS0, ..., MCS9) with different speeds and reliability. To expand the channel range of 1 MHz, the standard defines a new CCM MCS10, which is nothing more than a modification of MCS0 with double repetition, which increases transmission reliability. Preliminary studies show that, thanks to MCS0, outdoor .11ah will allow data to be transmitted in the 1 MHz channel at 200mW for a distance significantly greater than 1 km.
The available data rates for different channels and CCMs are listed in the table below:
These rates can be improved by reducing the OFDM symbol duration and using multiple spatial streams (MIMOs).
So, the duration of the ordinary OFDM symbol in .11ah networks is 40 μs, 20% of which is a guard interval containing redundant data and preventing intersymbol interference. IEEE 802.11ah allows you to halve the guard interval, which increases the data transfer rate by 11%.
IEEE 802.11ah stations can use up to 4 spatial streams. As you know, N spatial streams increase the data transfer rate N times. Thus, the maximum data transfer rate in .11ah networks reaches almost 350 Mbps. However, 350 Mbps is the maximum data rate measured at the physical layer. In fact, if you do not change the data link protocols, then the data transfer rate, say, at the network level will be significantly lower.
Why this is happening and how TGah changed the link layer protocols to reduce protocol loss will be described in the next article.