Data freshness, not speed, most important for IoT
The age of sensor data is more important than how fast it takes that information to travel around Internet- and Location-of-Things environments, say some experts. Scientists are, in fact, rethinking the network because of it.
“It’s not enough to transmit data quickly. That data also needs to be fresh,” says MIT in a news release.
The university has been working on better ways to ensure that sensors, which distribute readings for analysis, provide the most salient stuff. It’s not easy because you can’t just send everything at the same time (an obvious solution) — there isn’t enough bandwidth.
Drones, self-driving cars, and other IoT users need the most up-to-date data to work properly: A networked swarm of vehicles won’t work right with second-old position data from one, and newer data from another, for example — they’ll crash into each other. So, identifying what is stale information and what isn’t so you can compensate will become more important than the prior holy grail of minimizing the amount of time data takes to reach a destination, some say. The same idea applies to other IoT sensors. It’s a radical change.
“Only recently have researchers considered the age of the information. How fresh or stale information is from the perspective of its recipient,” the school writes.
MIT says its solution is to build an algorithm that creates an index for each node in the wireless network. The freshness of data, the quality of the channel, and a manually set node priority are all factors contributing to the index. The index fluidly prioritizes and optimizes traffic, and it does it without overloading the delicate radio setup.
The school’s simple network is designed to work with just one data receiver, but with several nodes (drones, say) providing the sensor data. It’s a realistic assumption that only one node can send over a wireless data channel at any one time, but that you want the data receiver to obtain the freshest data from all of the nodes.
Nodes with a higher index value thus get priority over the channel. The latest data along with the best radio propagation scores higher “moment-to-moment” than stale data over an iffy connection, but older data over a good connection scores higher than fresh data over a dodgy connection. A newer, faster drone, say, can be manually set to a high priority.
Eventually, the group plans to make their self-freshening network work with multiple base stations.
Securing wireless data
MIT has been making other strides in broad IoT-oriented wireless networking concepts, too. Unrelated to the freshness-of-data idea, a separate group within the school has been working on ways to secure machine wireless data better — something that will likely become increasingly important as the IoT takes hold.
The university suggests that the solution is to simply move data packets randomly between frequencies so fast that hackers can’t intercept them. It’s a form of frequency hopping, as is used now, but it shifts frequencies every bit in microseconds. Significantly faster and more densely, in other words.
The school’s “packet-level frequency hopping sends one data-packet at a time, on a single one-megahertz channel, across a range of 80 channels,” MIT says in another news release. It’s too quick for intruders, it claims. And a shared, confidential key is used to re-map the data at the receiver.