As is well known, one of the pillars on which the success of the IoT rests is its requirement for a multitude of devices to be connected to the rest of the world via the internet. As such, it’s safe to assume that wireless communication will be the backbone that underpins this. The question is which technology is most suitable? There are multiple options, each of which offers different methods for establishing connectivity for IoT applications. Of these, Wi-Fi, Bluetooth and cellular are the most well-known today. Which is used will depend on the application, range, data throughput and battery life, and it’s even possible that a combination of technologies is necessary.
This post is focussed on reasons why Wi-Fi will likely emerge as one of the most popular IoT technologies for the smart/connected home. We’ll look at the different applications of Wi-Fi in sensor-based networks and the requirements of Wi-Fi silicon for smart home applications.
The charts below show that Wi-Fi is already a ubiquitous connectivity solution, with Wi-Fi shipments expected to hit close to 4 billion units by 2021. Dialling this down to market shipments for IoT use cases, the shipments are expected to hit about 500 million by 2021 (see Market Shipments requiring Low Power Wi-Fi, below).
The main segments that require low data-rate, low-power Wi-Fi are smart home, wearable and healthcare and commercial and industrial segments. The smart home segment is expected to take the largest share of this pie.
Smart home and IoT
There are several reasons for the popularity of Wi-Fi in smart home applications over other technologies. There’s its ease of set-up, its native IP network support, the strength of the network management tools available and the general familiarity people have with it. Wi-Fi also offers much greater range compared to its main technology competitors.
However, Wi-Fi traditionally is a more power hungry technology compared to the alternatives, so to ensure its practicality for IoT, steps must be taken to ensure that it is designed in as power-efficient a manner as possible.
But what are the requirements of Wi-Fi silicon for IoT applications? Any silicon IP designed to meet low power consumption requirements for IoT applications would, in general, need to have the following characteristics:
- Cores designed ground up for IoT
- Low power consumption metrics
- Low die area and bill of materials (BOM)
- Low-cost solution
- Easy to integrate into SoCs
- Flexible partitioning and interfaces
- Robust performance
- To meets or exceed standard IEEE requirements
- Field-proven solutions
- Pass interoperability and qualification requirements for Wi-Fi certification
The smart home promise – power consumption as a key driver
The devices in the smart home (controlled devices) fall into three categories (from the perspective of power consumption).
|Type of devices||Power consumption sensitivity|
Since battery-powered devices have high power sensitivity, it could be argued that Wi-Fi isn’t a suitable option for them. Therefore, to understand Wi-Fi power consumption it’s useful to dive deeper into different power consumption scenarios by analysing different message exchanges between a controller and controlled devices.
The flow chart below takes three different usage scenarios; a thermostat, a smoke alarm detector and a smart door lock.
In all three cases, the controlled devices would need to exchange:
- Initialization/authentication messages – a onetime message exchange at power-up.
- Keep alive messages (k) – messages transmitted once approximately every 30 seconds to maintain the association.
If configured for power saving, while associated, the controlled devices will spend most of its time in “connected standby” state
- The devices would awaken at regular intervals (called DTIM to receive [a] beacon (b)
- The devices will simply receive [a] beacon and go back to sleep.
The messages in the active state could be one more of the following type:
- Periodic data transmission – a thermostat sending recorded temperatures every 60 minutes.
- Event triggered data transmission – a smoke alarm detecting a smoke or a fire and sending the alarm.
- Smart door lock – a smartphone (controller) sending a command through an app to lock/unlock the door.
A generalised flowchart of these message transactions are shown below
In order to achieve the lowest power consumption and thus lengthen battery-life, it is necessary to optimise not only the Transmit and Receive power consumption but also the connected standby power, (which includes power consumed in the boot-up and sleep states).
Generally, existing Wi-Fi cores are designed to transmit and receive large amounts of data quickly, and are not optimised to achieve the lowest power to lengthen the lifetime of battery-powered devices as required for IoT applications. In contrast, Imagination’s Whisper wireless communications cores have been designed from the ground-up to meet the requirements of the IoT.
Some examples of the design choices made inside Whisper cores are:
- A power centric RF design, trading noise and a reduction in transmit power for lower power consumption.
- Lowering the clock frequency of the PHY and reduced bit widths in the data path etc. to lessen dynamic and leakage power consumption.
- Flexible software partitioning by offering Thin, Full and Fully Embedded MAC models.
Whisper radio processing unit and power consumption metrics
The power consumption of Whisper in two usage scenarios is shown in the figures below:
- Transmit Event – smoke alarm transmits the health of the smoke sensor once a day
- Receive Event – a smart door lock gets a command once a day to lock or unlock a door
For more information on the Whisper architecture and its unique low power design, please refer to the previous blog post on this topic.
To summarise, we think it likely we will see Wi-Fi widely adopted for IoT applications, with low power consumption as the key driver. However, repurposing existing Wi-Fi cores for IoT is not an ideal solution. Imagination’s Ensigma wireless communications cores offer a complete end-to-end, low-power offering for IoT that consumes half the receive power compared to all available solutions. It offers thin MAC, full MAC and fully embedded MAC options, with a hard-macro option to ease integration. Ensigma’s Wi-Fi for IoT solutions are offered with compliance validation and are fully interoperability labs tested so customers can be sure the parts will do what they need and enable them to go to market with confidence.
If you’d like to listen to a webinar on ‘How low-powered Wi-Fi-enabled sensors will power the future of IoT’ given by Narayanan Raman, the senior business development manager for Ensigma, then you can register for it here.