As mesh networking standards like Zigbee and Bluetooth Low Energy continue to trickle into the mainstream, it will become increasingly important for consumers and device manufacturers alike to understand when “traditional” IP connectivity remains necessary.
A mesh network is a communications configuration in which radio node devices connect to each other in a scalable web, but do not necessarily connect to the broader internet. Most mesh technology requires additional devices or software to achieve internet connectivity, though there are some protocols that are automatically internet-enabled.
Beyond merely maintaining technological continuity — which, to be clear, is not nothing — IP connectivity can often elevate the functionality of devices that are integrated into a mesh network. While internet-independence is valuable insofar as it enables devices connected over mesh networks to continue to communicate even in the event of an internet (or power) outage, it also precludes users from accessing and controlling their devices remotely. In other words, by integrating their mesh network with IP, a consumer gains the immense benefit of, say, ensuring their front door is locked from work.
For device developers, this integration enables the transmission of diagnostic or usage data from meshed devices to the cloud. Among other things, the ability to transmit such data is the cornerstone of the increasingly popular edge computing paradigm, wherein minor workloads are handled by devices’ inbuilt processors and major workloads are ferried to a remote data center for processing (i.e. the cloud).
For instance, while a smart thermostat may be equipped with enough computing power to regulate a home’s temperature according to preset parameters, it likely won’t have enough power to run the calculations necessary to improve the home’s overall energy efficiency. Accessing enough compute to support sophisticated analyses like these typically entails connecting to a gateway device that not only possess the requisite processing power, but serves as a bridge to the internet.
Understanding IP Connectivity for Mesh Networks
To achieve IP connectivity within the confines of a mesh network, the meshed devices must be able to communicate not only with each other, but with the broader internet. As one might intuit, this communication occurs at the network layer of a networking stack.
Every smart device is comprised of a hardware layer (including a radio that transmits signals), a network layer (through which devices communicate with each other), and an application layer (through which end users hand down commands), and while each has an important role to play in facilitating connectivity and interoperability, the middle layer holds the greatest bearing over the present discussion.
Broadly speaking, there are two ways a meshed device can connect to the internet. One, it can feature a natively IP-based networking stack that enables data to move seamlessly from device to device and from a device to the internet. Two, it can connect to a separate IP-enabled gateway device or controller that serves as a conduit between the non-IP device (and often, the mesh network as a whole) and the internet.
To choose the right approach for integrating IP connectivity into your mesh network, it’s important to understand the ins-and-outs of the various networking standards that you have at your disposal. For the purposes of this article, we fill focus on three major mesh networking technologies: Bluetooth Low Energy (BLE), Zigbee, and Thread.
Bluetooth Low Energy
Bluetooth Low Energy uses the General Access Profile (GAP) to enable highly energy-efficient communication across mesh networks. When it comes to integrating devices on a BLE mesh network with IP, however, things become a little less straightforward. Fortunately, BLE device developers have a few options at their disposal.
First, a series of standard RESTful APIs make it possible to discover and manage BLE devices from the internet. These APIs can interface with GAP and the Generic Attribute Profile (GATT), but require a gateway component like a router to facilitate remote access. Similarly, the HTTP Proxy Service (HPS) empowers users to set up their mesh devices to communicate with the cloud, but again, doing so requires a gateway HTTP client device.
But in many cases, your best bet for IP connectivity will be the Internet Protocol Support Profile (IPSP). This provides Bluetooth Smart 4.2 devices with actual IPv6 addresses, which both enable IP capabilities over the low-power network and allow BLE devices to interface with non-BLE IoT devices that use IP. Again, this requires a gateway device, usually a Bluetooth IPv6/802.11 router.
The Zigbee protocol is already an established global IEEE 802.15.4 standard for creating low-power, low-latency mesh networks in residential and commercial environments. Newer specifications define an internet protocol that provides users with more control over their devices, as well as a host of new device capabilities — including the ability to connect to the internet. Zigbee Smart Energy V.1 (and the newer V.2) relies on Zigbee IP, a network layer that, like BLE, uses IPv6 over Low-Power Wireless Personal Area Networks (6LoWPAN) and header compression to route IPv6 packets over IEEE 802.15.4.
Zigbee IP is the first open standard for an IPv6-based full wireless mesh networking solution, and it allows devices to easily establish an internet connection without consuming a great deal of power. Generally speaking, Zigbee IP mesh networks include several device types: a coordinator that controls network security and formation, routers to extend network range, and hosts for specific sensing or control functions. In a home, for instance, smart appliances can be both routers and end devices. A border router can then be used to integrate this system with the internet.
Additionally, thanks to Dotdot, the Zigbee Alliance-developed common language for IoT devices, Zigbee devices can now interface with Thread, an IP-based networking stack that enables IP-based addressability and functionality over WiFi.
Unlike BLE and Zigbee, the Thread protocol only exists at the network layer. It emerged from an industry-wide effort to standardize the capabilities of 6LoWPAN, and it supports both mesh and IP connectivity for devices right out of the box. Thread’s IPv6-based networking protocol supports both device-to-device and device-to-internet connections, and uses IEEE 802.15.4 for power efficiency. Thread also uses edge routers, or “Border Routers,” to facilitate secure joining.
While occasionally integrated into smart home devices, Thread is geared more toward commercial applications like smart sensors and building IoT devices. Again, with the introduction of Dotdot, Zigbee and Thread can be integrated in a way that brings the impressive range of Zigbee devices together with the IP capabilities of the Thread network layer.
Is IP Connectivity Necessary?
This is a complicated question. On the one hand, relying too heavily on IP connectivity can lead to serious issues. For instance, if a homeowner builds a smart lighting system around a mesh network with IP connectivity, they risk losing contact with their smart device gateway in the event of an internet outage. This could result in their inability to turn off the lights, even from within their own home! In short, because the internet has intermittent connectivity issues (and is comparatively power-hungry), if your foremost need is simply for your devices to communicate with each other effectively, a non-IP mesh network will likely be sufficient.
On the other hand, however, IP connectivity is clearly necessary to accomplish certain tasks. Integrating mesh networks with IP functions can allow for remote access that could, for instance, allow a consumer to control their home’s thermostat or security devices from work. Similarly, for industrial applications, a factory’s IoT-enabled machinery may need IP access to be able to send data to a cloud platform for inventory management or some other business application.
Ultimately, these latter sets of circumstances are defined by end users’ needs to remotely access and manage an array of smart devices, many of which must be capable of transmitting the data they collect over considerable distances. These circumstances also introduce new security concerns to the equation, as network nodes that are assigned IP addresses are vulnerable in different — and potentially greater — ways than those that are sequestered on a localized mesh network.
That said, sometimes, integrating IP into a mesh network is absolutely necessary. In cases in which your personal or business goal can only be accomplished via an IP-specific capability, using one of the technologies outlined above will likely be the most effective way to achieve the kind of connectivity you need to succeed.