A mesh network may be the answer to connecting the Internet of Things without needed the internet itself or access to the cloud
Everything we use on a daily basis is increasingly becoming connected. We already see it with our cars, thermostats, wearable devices, and appliances. However, these devices often require direct connections to the Internet to operate which means slower connections, a lack of privacy, and higher overall cost.
In things like cars, we see the need for cellular connectivity which is expensive. The chips themselves require certifications that cost millions of dollars, making the cost a barrier for new companies. In addition, when people buy a connected device they often aren’t willing to purchase a data plan along with it. This means companies swallow the cost of the data plan price themselves. In cheaper devices, like thermostats and wearables, this additional cost is too high to bear, so they are often only packaged with Wi-Fi or Bluetooth capabilities.
In addition to being expensive, there are major privacy and security risks that come with allowing the data from devices to be transmitted to and stored within the cloud. We’ve seen examples over and over of data breaches – security is a cat and mouse game and the hackers always seem to gain an advantage and expose data and users.
In terms of network efficiency, moving all of the data from these devices to the Internet or the ‘cloud’ is extremely inefficient. Often the people and other devices are located nearby – so sending everything to and from the Internet is a waste.
When it comes to ease of use, adding cellular connectivity is easy to navigate for users, however it is not a feasible long-term solution due to high cost in many of the future devices which will be small and cheap.
Additionally, when technology like Wi-Fi and Bluetooth are used, it often requires user configuration. Many people today find it difficult to set up and use Wi-Fi in their own homes with devices like smart TVs and mobile phones, so it will likely be more difficult for them to connect devices which have no screens and complicated ways to connect to networks with security and encryption.
The solution to these problems is autonomous mesh networks built from a collection of underlying wireless technologies. Most mobile phones currently support Wi-Fi, Bluetooth, and Cellular connectivity, along with some more specialized beacon technologies, Wi-Fi direct, Wi-Fi p2p and Multipeer. All of these can be stitched together so the devices that support the most technologies can form the backbone of the mesh, allowing cheaper IoT devices, which support only one or two of these technologies such as Bluetooth or Wi-Fi, to communicate automatically and easily with nearby devices.
The right mesh technology is coming
Right now, a mesh technology doesn’t exist that can stitch together all of these underlying wireless protocols in a way that doesn’t require expert users and lots of user intervention, but this technology is coming. Soon it will be possible for mobile phones to form the backbone of mesh networks, automatically select which networks to use and make use of multiple networks simultaneously.
Companies have approached this problem with special-purpose mesh routers and extra equipment, but people have enough boxes in their home. Mobile phones are already something most people have – and they are the way that people will be interacting with IoT devices anyway – so why not build the capability of the mesh into the phone in the first place?
Using autonomous local meshing, the efficiency of the network is improved; data from IoT devices like thermostats, appliances, and connected devices in the home can now be sent without the Internet to each other and to the homeowners mobile phone when they are nearby. When they are not home, these devices could fall back to Internet mode, or in the case where a phone mesh becomes widespread – it could continue to send through the mesh hopping from phone to phone until it reaches the homeowner at some farther away geographic location.
Information sharing between cars is most likely useful in vehicles located within relatively close proximity to one another. There’s no need for every car within a city to know the entire state of every other car – but as two cars become closer, more information may be useful to avoid collisions and to plan routes for each car effectively.
Security-wise – since data is not being stored in the cloud – it has less permanence. The risk of a user’s phone being compromised or the encryption of the data being broken along the path is still there, however, this can still happen with the Internet-enabled version. The mesh also has a unique ability to merge and aggregate data from multiple users – so perhaps this data could still be sent to the cloud, without compromising individual users identities, while still building value for companies who are interested in using this data to improve services in some way.
The critical benefit of mesh networks built from a variety of underlying wireless protocols is that each has its own strength. Some work over long distances, some work with low delay, low energy, high throughput – and all can be combined together to build something that is more valuable than the sum of its parts – just like the people it will be connecting. Which means in the future, the Internet of Things may wind up just being Things.
Dr. Jason Ernst is the Chief Networking Scientist at YO!