The shift to the smart home has begun although it is at its earliest stages. As with previous technology revolutions, the smart home is driven by the falling cost of silicon. But this is only part of the picture: increasing energy efficiency, both in terms of compute power and communications, is key to enabling smart devices that can support lifestyles. Technological shifts of this kind take a decade to manifest themselves but once they take hold the ramifications are enormous.
The term ‘smart home’ obscures the characteristics that will make its implementation attractive to most consumers. A number of the core technologies required to make smart-home devices have been available for more than a decade. And keen DIY enthusiasts have been able to build digital lighting controls and other forms of automation into their homes using devices that support, for example, the Zigbee protocol.
Although Zigbee is effective at interconnecting smart devices such as intelligent light bulbs and temperature controllers, it is focused on islands of automation, supporting smarter heating and lighting rather than delivering the smart home in its entirety. Zigbee does not readily support the concept of interconnection between different networks that lies at the heart of the internet of things (IoT).
The real smart home of the future will be an ecosystem that constantly learns. It will be aware of your movements and preferences, with the ability to adapt to changes in them over time with the help of information from your wearables and scores of fixed and mobile sensors connected to the IoT (Fig. 1).
Figure 1: Example of smart home connectivity through IoT
Different parts of the network will cooperate at different times to enable specific applications, such as transmitting video of a visitor at the door to the TV or a tablet watched by the homeowner. There may be a heating system that reacts to people moving around their house. If it’s unoccupied, the system may activate in response to the time they’re likely to arrive home, that time being calculated to take account of real-time traffic conditions.
The key to enabling this future is one where interoperability is paramount and where vendors of systems for this environment recognise that the smart home will be built in stages. A consumer may fit one or two rooms with smart lighting or install a smart thermostat in order to make immediate fuel savings before integrating them into the full smart-home environment. They will not want to replace those systems simply to make them compatible with the future ecosystem. Fortunately, they are unlikely to have to make that choice.
The current wireless communications environment for smart-home systems looks confusing – and this is likely to have been a factor in slowing adoption rates up to now. Security cameras will often use WiFi interfaces for high video transfer rates. A set of smart light bulbs is likely to use Zigbee’s lighting profile. A user may interact with a smart thermostat using Bluetooth.
A unifying force is version 6 of the Internet Protocol (IPv6). The exponential growth in the number of sensors and connected devices that will need to be deployed in smart homes around the world, IPv6 is an ideal protocol because of the large address space it provides (Fig. 2). Access to such a large address space without the use of address-translating gateways will be instrumental in enabling the ecosystem around the smart home.
Figure 2: Table to show difference between IPv4 and IPv6
There are further advantages to IPv6. It has tools to support stateless address autoconfiguration, a facility that is highly suited to setting up sensor network nodes that have very limited processing power. The original IPv6 protocol incurs significant overhead on low-power wireless links. The IPv6 version of the Low-power Wireless Personal Area Networks (6LoWPAN) standard handles a subset of IPv6 functions with discovery optimised for smart-home and IoT networks.
IPv6 and 6LoWPAN will help drive the consolidation of IoT and smart home protocols around three key standards. Wifi already supports IPv6. The IEEE 802.15.4 wireless protocol that underlies Zigbee is able to run IPv6. A continuing programme of enhancements mean Bluetooth will also be made IPv6-compatible through the adoption of the 6LowPAN protocol stack.
Integrated into millions of smartphones, tablets and other portable devices, Bluetooth is already the gateway to the smart home’s IoT. Further enhancements that support the migration to IPv6 will make it the smart choice for a much wider range of devices. The introduction of multi-hop mesh networking ability, the focus of another Bluetooth SIG working group, will let Bluetooth Smart be used in an even wider range of situations.
A wireless mesh network is a network built on a number of fixed and mobile nodes that make it possible for communication from one node hop across number of others on the way to a destination. The destination may be a peer on the network or a gateway providing access to the wider internet. It is a highly reliable means of providing connectivity to nodes within a network as, if one node that is used to relay data to another fails, it is generally possible to substitute another (Fig. 3).
Figure 3: Example of a wireless mesh network
The mesh is an important feature of low-power networks, which have short transmit range compared to technologies such as WiFi and cellular – which gain their longer reach at the cost of higher energy consumption. The multi-hop strategy offered by mesh networks overcomes the range limitation by allowing devices that are out of range of each other to send packets to much closer nodes that then use short hops to complete delivery to the destination.
In smart-home networks, mesh networking provides another advantage when dealing with multiple devices. For example, to turn the lights on in a conventional IoT network, the master needs to send a message to each smart bulb. A mesh network can use a technique called ‘flooding’ in which the device sending the ‘lights on’ command to the bulbs closest to it. Those bulbs then relay commands to their neighbours until the room is lit.
Although meshes offer some distinct advantages for smart home and IoT scenarios they have not grown as quickly as Bluetooth or Wifi in the market. The development of mesh capabilities for Bluetooth is likely to coincide with much stronger uptake. Bluetooth Smart Mesh has the advantage of being developed later than existing mesh protocols: its developers have been able to use the experiences of earlier work on mesh technologies to guide progress on a forthcoming standard.
Bluetooth standards development has 15 years of experience in creating protocols that offer consumers ease of use and interoperability across many different types of device, using the results of billions of systems in the field to fine tune the standards to the state they are in today. Bluetooth Smart Mesh will ensure that the mesh capabilities of the network will not result in a break in compatibility.
The groundwork has already been laid. Version 4.1 of Bluetooth Smart introduced the concept of the Scatternet, an addition to the existing Piconet approach used by the protocols. The Piconet uses a conventional master-slave configuration where one master is logically connected to a set of slaves, with the master initiating communication. The Scatternet makes it possible for a master node with a logical connection to one or more slave nodes to also be a slave of another node. This ability to swap modes makes it possible for a Scatternet to implement multi-hop connectivity without forcing a complete change away from the pure master-slave Piconet architecture.
The result of these developments will make Bluetooth not only a major player in the smart-home ecosystem, it is likely to become the smart choice for low-power networking whatever the device.