What features are needed by an IoT protocol?

For the Internet of Things (IoT) to become the reality as seen by forward thinking technologists, we need a network protocol that can: support thousands and hundred of thousands of devices, travel long distances, and reduce power requirements. Supermarkets wishing to deploy electronic shelf labels, will need a wireless protocol that supports over 100,000 devices. Farmers that wish to improve their crop output, will want a wireless protocol that can travel long distances. Homeowners will want a wireless protocol that allows battery operated sensors to operate for multiple years.

Why can't we use the current WiFi protocol for IoT?

The wireless protocol that is deployed in almost every office and home, is WiFi 802.11ac. The 802.11ac protocol operates in both 2.4 GHz and 5 GHz frequency bands and can deliver a single stream bandwidth of 866 Mbps. This is useful for streaming content, but falls short from an IoT perspective. Because of its high frequency, 802.11ac delivers high throughput, but has a limited range. Getting complete coverage for your home can sometimes represent a challenge. Aditionally, the frequency demands place a great power demand on battery powered devices. And lastly, the 802.11ac protocol cannot support thousands of devices.

What are the pioneering IoT protocols?

In the race to make IoT a reality, a variety of wireless protocols have been released. From a cellular perspective, there are two protocols of interest: LTE-M and NB-IoT. LTE-M (sometimes called Cat-M or LTE Cat-M1) is the low power wide area (LPWAN) protocol that is compatible with LTE technology. From an IoT perspective, this protocol has a throughput of 1 Mbps and reduced the power requirements from regular LTE. This technology will operate on all 2G, 3G, and 4G networks and is expected to roll out globally this year (2018).

The NB-IoT protocol was designed to further reduce radio signaling overhead, further improve battery life, support both IP and non-IP data, and support SMS. The NB-IoT protocol, while different from LTE, interoperates in a compatible way by using a single 180 kHz resource block within the LTE transmission range. The throughput of this protocol is 200 kbps. NB-IoT will start its global rollout this year (2018).

NB-IoT made major progress towards enabling battery powered sensors, but had to give up some key features of LTE, such as seamless handover. An NB-IoT device cannot do a handover in the connected state.

Another interesting protocol for IoT is Bluetooth Low Energy. This protocol works in the 2.4 GHz band with 3 advertising channels and 37 data channels, each 2 MHz wide. This protocol can transmit about 100 meters (with no barriers) and delivers a data rate of 1 Mbps. This protocol was designed to work in conjunction with a cell phone and enables applications such as: blood pressure monitors, exercise sensors, and location beacons.

Zigbee is a protocol that is similar to Bluetooth Low Energy. This protocol operates in the same 2.4 GHz band and can transmit a similar distance, about 100 meters. However, Zigbee has lowered the throughput to 250 kbps to provide better battery management than Bluetooth Low Energy. Additionally, Zigbee supports a mesh network of up to 65,000 devices. This protocol has found traction in the manufacturing market, enabling battery powered sensors.

Turning to long distance protocols, there are two that stand out: Sigfox and LoRa. Sigfox is a company that has designed an IoT protocol and operates a cloud based infrastructure for their connected devices. The protocol uses a frequency range of 902 MHz to 928 MHz in the US and was designed to solve two primary issues: enable long distance transmission and minimize battery drain. Uplink messages are very small (12 bytes) and I/O throughput ranges from 100 bps to 600 bps. The protocol was designed for a sensor network with devices that only communicate a few times a day. Devices are allowed a maximum of 140 uplinks messages (12 bytes) and our downlink messages (up to 8 bytes) per day. The number of devices that can be supported by one tower range from 8 devices, transmitting continuously to 800 devices transmitting at a 1% of the time. In contrast to the low data rates and limited number of devices supported by a single tower, Sigfox devices can transmit from 10 km to 100 km depending on the topography and since Sigfox devices are idle 99.x% of the time, batteries can last multiple years.

LoRa is an open IoT wireless network standard run by a collection of companies belonging to the LoRa Alliance. This protocol uses the 902 MHz to 928 MHz frequency range in North America. LoRa is a bidirectional IoT wireless network protocol that provides better throughput than Sigfox, up to 50 kbps, however with a smaller range of 15 km to 25 km. In order to accommodate IoT devices with differing needs, LoRa has defined 3 classes of devices: Class A -- low power, Class B -- deterministic latency, and Class C -- lowest latency, where Class A devices have a multi-year battery life similar to SigFox and Class C devices have the best performance but use the most power.

How does 802.11ah fit in?

The 802.11ah protocol, also referred to as HaLow, had the benefit of being designed after 802.11ac, Zigbee, and LoRa. As such, the designers of 802.11ah could leverage the lessons learned from these protocols.

The 802.11ah protocol was designed to reach the following goals:

support for many devices positioned within close proximity of each other;
enable high bandwidth two way communication;
communicate over long distances;
transmit through barriers; and
enable sensors to run on batteries for multiple years.

In summary, the 802.11ah protocol operates in the 902 MHz to 928 MHz frequency range in North America. It has a variable encoding rate which can be tuned to deliver the best bandwidth adjusting for distance and obstacles. From a device perspective (single spatial stream), the bandwidth ranges from 150 kbps to 86 Mbps. The data throughput rate of 802.11ah greatly exceeds the throughput of any of the other IoT protocols. The range of 802.11ah varies depending on the bandwidth used. At the 150 kbps bandwidth, 802.11ah can reach over 1 km. Learning from other IoT protocols that only use a pseudo random backoff mechanism to recover from collisions, 802.11ah has a variety of different mechanisms to organize communication between devices. The result is that a single Access Point can support over 8,000 devices. From a power management perspective, 802.11ah enables battery powered devices to operate for multiple years. Lastly, unlike LTE-M, NB-IoT, and SigFox, 802.11ah does not require the payment of network usage fees.

802.11ah Details -- PHY Layer

The primary features at the PHY layer were adapted from 802.11ac to operate effectively in the sub-1 GHz frequency range. 802.11ah uses Orthogonal Frequency Division Modulation (OFDM) to efficiently use its allocated bandwidth range. OFDM allows for multiple simultaneous transmissions on different sub-carriers without introducing interference with each other.

In the US, 802.11ah has a frequency range from 902 MHz to 928 MHz. Using OFDM, this frequency range is divided into separate channels. The separate channels can be used by different devices simultaneously, thereby increasing the overall bandwidth of the network or the channels can be combined to increase bandwidth for a specific device.

802.11ah Details -- Terminology

When discussing 802.11ah there are a few key useful terms:

IoT Device -- Is the IoT wireless equipment that captures data and transmits this data upstream into the 802.11ah network. Some examples of an IoT Device are: a temperature sensor, a wireless camera, or an asset location beacon. The term IoT Device refers to the data capture portion of the equipment and the radio and MAC portion of the equipment.
Station (STA) -- Is the portion of the equipment that implements the radio and MAC for the 802.11ah protocol. Each IoT Device has a Station component that implements the radio and MAC.
Access Point (AP) -- Is the wireless equipment unit that manages IoT Devices. It provides a connection point so that IoT Devices are connected to the Internet or some other network backbone. Each Access Point has a Station component that implements the radio and MAC.

802.11ah Details -- Key Features

The MAC layer implements many of the key attributes of the 802.11ah protocol. These key features make the 802.11ah protocol a great choice for an IoT network: supporting a high number of devices in a high density configuration; transmitting long distances; minimizing power requirements and thereby enabling multi-year IoT Device battery life; and delivering the best data bandwidth of any of the existing IoT protocols. In this section we will discuss the key features of 802.11ah and show how these features help 802.11ah reach its IoT goals.

Reduce the amount of data that is sent over the air
The MAC layer has been optimized to minimize the number of bits that are sent over the air. 802.11ah uses a short / compact MAC header format. Since the overhead of the MAC header may be considerable when compared with the size of the IoT payload, the MAC header was reduced from 30 bytes to 18 bytes. 802.11ah also uses a Null Data Packet (NDP) as a placeholder for some simple control packets, such as; CTS, ACK, PS-Poll frame.

Mitigate negative effects of signal fluctuation
To mitigate the negative effects of signal fluctuation, 802.11ah added the Subchannel Selective Transmission (SST) feature. This feature allows devices to quickly switch between a set of defined subchannels to adjust for short term fading conditions or interference. This allows a device to start out sending data using the 16 MHz channel and then scale back to lower channel bandwidths as needed.

Hierarchical Association Identifiers
In most of our marketing material, for simplicity we say that a single Access Point (AP) can support over 8,000 IoT Devices. In fact 802.11ah can support 2^13 devices (8,191). 802.11ah has introduced a way to refer to each of the IoT Devices in a organized, efficient, and hierarchical manner. The Association Identifier (AID) is assigned by the AP during the association process and has a 3 level hierarchy: page, block, and sub-block. This hierarchical ID structure simplifies communication among many devices, by using a bitmap to identify a set of devices. For example, the AID can be used to broadcast data to a hierarchical group of devices (GroupAIDActivated).

Restricted Access Windows
The purpose of the Restricted Access Window (RAW) is to allow IoT Devices to transmit data upstream while reducing the chances of a collision. The AP is responsible for defining the RAW channel, time of use, and assigning sets of devices to separate RAWs. A device with data to send can power sleep until its RAW becomes available.

Target Wakeup Time
The purpose of Target Wakeup Time (TWT) is to allow the AP to schedule the activity of different IoT Devices at different times to minimize communication contentions. Devices that don't register a TWT, must wake up every beacon frame time.

Data Buffering
The AP will buffer data for a device while a device is in a power sleep mode, subject to space availability in the AP. This allows a device to power sleep, wake up at the appointed time, and not lose data.

Traffic Indicator Map
The Traffic Indicator Map (TIM) identifies the devices for which the AP has pending and buffered data. The TIM packet is transmitted on every Beacon frame. This information allows devices not involved in pending transmissions, to go back into power sleep until their next wake up time.

One Hop Relay
To simplify network configurations, 802.11ah supports single hop relay. This feature allows an IoT Device A to connect to another IoT Device B which is connected wirelessly to the Access Point (AP). So from the perspective of IoT Device A, it hops over one IoT Device to communicate with the Access Point.

Precise Location Information
By integrating technology from the IEEE 802.11mc standard, 802.11ah can deliver the same location accuracy possible with 802.11ac (1 meter), with the additional benefit of the extended coverage range of 802.11ah.

Performance, Configurations, and Use Case Analysis

Performance data provided in this section was derived either: by mathematical modeling, or by using the knowledge that engineers and radio scientists have learned over the years; or by testing on physical equipment; or by running 802.11ah code on a network simulator. To make our information clear we have put the tags: (Math), (Test), and (Sim) to denote performance information derived via: mathematical modeling, physical testing, and simulation respectively. Note that a free discrete-event network simulator (ns-3) is available from nsnam.org and is licensed under a GNU GPLv2 license.

Transmission Distance
Our marketing material says that 802.11ah can transmit data 1 km with a data bandwidth of 150 kbps. This is a simplified and conservative number. The 802.11ah specification states that data transmitted with an MCS 10 encoding scheme can reach 1 km with up to 166 kbps, when using the short guard interval. The mathematical modeling behind the 802.11ah specification used a power level of 200 mW. However, in the US we can transmit in the 802.11ah frequency spectrum using up to 1 Watt (30 dBm) . Using mathematical models to predict the transmission range of 802.11ah using 1 Watt of power, we arrive at a distance of 1.5 km. In conclusion, for planning purposes, one is safe using the conservative range value of 1 km.

Number of IoT Devices per Access Point
Our marketing material says that one 802.11ah Access Point can support over 8,000 IoT Devices. This is a simplified number. The 802.11ah specification states that a single AP can support up to 8,191 IoT Devices. However in reality, the maximum number of IoT Devices supported by a single AP depends on data transmission requirements of each IoT Device and their distance from the AP.

Using the ns-3 simulator, it was shown that 6,960 IoT Devices can be connected at a distance of 1 km and each transmitting a message every 60 seconds with no loss of data.

Flexible Bandwidth
One of the key features of 802.11ah is the tradeoff it provides between throughput and range. The 802.11ah specification says that the protocol can deliver a throughput of 150 kbps at a distance of 1 km and closer to the AP it can deliver a throughput of 86 Mbps.

Using the ns-3 simulator, it was shown that:

6,960 IoT Devices can be connected at a distance of 1 km with each transmitting a message every 60 seconds with no loss of data. Each upstream payload was 100 bytes and each upper level acknowledgement was 90 bytes. A 2 MHz channel was used with a MCS 0 encoding.

20 continuously streaming IP cameras, at a distance of 1 km, can reliably transmit data at 160 kbps.

10 continuously streaming IP cameras, at a distance of 200 m, can reliably transmit data at 255 kpbs.

Access Point Configurations
The number of Access Points you need and the configuration of Access Points and IoT Devices is dependent on the problem you are solving. Some framing question are:

How many IoT Devices do you need to support?
What is the IoT Device density?
What are the bandwidth requirements of your IoT Devices?

The following link gives a detailed configuration analysis of 3 possible configurations: a single AP, a single ring of APs, and a double ring of APs. In summary:

A single Access Point can support over 8,000 IoT Devices at a distance of over 1 km, which is 0.62 miles and encompasses an area of 490 acres.

A single ring of Access Points (7 APs) can support 50,000 IoT devices at a distance of 1.5 km, which is 0.93 miles and encompasses an area of 1,100 acres.

A double ring of Access Points (19 APs) can support over 150,000 IoT devices at a distance of 2.3 km, which is 1.46 miles and encompasses an area of 2,700 acres.

802.11ah Use Case Overview

As you can see from the foregoing information, the designers of the 802.11ah protocol did a great job of reaching their design objectives.

Objective: multi-year battery life for IoT Devices
The 802.11ah protocol, through a variety of techniques, delivers the highest throughput rates of all the IoT protocols while reaching the long battery life objective. The ability to have IoT devices that operate for multiple years on batteries is very beneficial because now IoT sensors or controls can be easily added without the additional complexity of supplying wired power to these devices.

In the home: battery powered moisture detectors can be placed in the crawl space under the house; water leak detectors can be placed under sinks; the water leak detectors can trigger the battery powered emergency water shut off valve; and a battery powered mailbox alert can let you know when you have mail.

In a building / highrise: battery powered service request buttons can be placed in restrooms; battery powered water leak sensors can be placed under sinks and water heaters; and situational environment sensors, measuring temperature or air quality, can be placed in specific locations to understand environmental changes that are bothering the occupants.

On a farm: battery powered soil sensors can be easily installed in each needed area, enabling effective implementation of variable rate irrigation (VRI) and precise delivery of fertilizer; battery powered livestock sensors can monitor health and track location.

In manufacturing: battery powered sensors can be added to packing materials to provide an understanding of the environmental stresses placed on product packaging during shipping; sensors can be included with the product, allowing marketing to do A/B testing or track product quality based on consumer product usage patterns.

Objective: improved bandwidth over existing IoT protocols

In the home: WiFi phone coverage can be greatly extended; surveillance video cameras can be placed at greater distances, such as a video camera at a gate entry point.

In a building / highrise: cameras outside of elevator doors can be used to determine the expected elevator traffic, enabling more efficient scheduling of elevators;

On a farm: WiFi phone coverage can be greatly extended; drones can be controlled at longer distances while also providing video feeds, allowing examination of crop health;

In a hospital: smart gurneys could provide the same sensor information as a connected bed, allowing hospitals to scale their care more effectively during emergency situations.

In logistics: video cameras can be distributed throughout the yard to provide a better understanding of activity and traffic.

Objective: support many devices

In a supermarket: electric product shelf labels can be used on over 150,000 different products.

In farming: a greenhouse can use many soil sensors within a confined area, enabling effective implementation of variable rate irrigation (VRI) and precise delivery of fertilizer.

logistics and manufacturing: many assets can be tracked, with a single ring of Access Points over 50,000 assets can be tracked.

Objective: long transmission reach

In the home: WiFi phone coverage can be greatly extended; surveillance video cameras can be placed at greater distances, such as a video camera at a gate entry point; a battery powered mailbox alert can let you know when you have mail.

In a building / highrise: all heavy equipment, regardless of where it is located, can be enhanced with a IoT sensor that monitors their operation, thereby reducing maintenance costs and unexpected equipment breakdowns; battery powered sensors can be placed anywhere.

On a farm: with a double ring of Access Points, sensors can cover a 2,700 acre farm.

In logistics: many assets can be tracked within the depot, a single ring of Access Points can cover 1,100 acres.

Additional Benefit: Precise Location Information
By integrating technology from the IEEE 802.11mc standard, 802.11ah can deliver the same location accuracy possible with 802.11ac (1 meter), with the additional benefit of the extended coverage range of 802.11ah. Using packet timing triangulation from 4 access points, location information can be refined to include the floor number within a building.

In logistics: the location of all assets can be tracked; congestion patterns in the depot can be analyzed and reduced.

In a department store, big data analysis of the location patterns of shoppers can: help adjust display placement; identify which products are desired, but are too costly (correlate browsing time in one location with a purchase event); identify areas of the store that needs more sales people.

In a hospital: the location of patients and mobile equipment can be determined.

In a large warehouse store: location information can be used to identify an area of the store that needs more sales people, for example a customer on an aisle for more than 10 minutes may indicate that they need assistance.

In farming: livestock movements can be tracked; geofences can detect if any livestock is outside of the ranch boundary.

Choosing the right IoT networking technology

We are here to help you understand your IoT technology challenges. By engaging in conversation, we learn the issues relating to different IoT application environments and we can together determine if 802.11ah is the right technology for you.

In summary 802.11ah provides:

great distance coverage:
- 1 Access Point: 1 km, 0.62 miles, 490 acres;
- 1x Access Point ring: 1.5 km, 0.93 miles, 1,100 acres;
- 2x Access Point ring: 2.3 km, 1.46 miles, 2,700 acres;

the best device density per Access Point:
- 1 Access Point: 8,000 IoT Devices
- 1x Access Point ring: 50,000 IoT Devices
- 2x Access Point ring: 150,000 IoT Devices

the highest uplink and downlink throughput of any of the IoT wireless protocols:
- 150 kbps to outermost IoT device in AP rings
- 2 Mbps at a range of 500 m (single AP)
- 6.5 Mbps at a range of 250 m (single AP)
- 16 Mbps at a range of 100 m (single AP)

multi-year battery life for IoT Devices

and requires no network usage fees to be paid to wireless telecom companies.

We look forward to speaking to you about your IoT needs.