First, let’s define a few terms. I’ll assume you know what low power wide access networks (LPWANs) are, and how the IoT uses them. But the interaction of LPWANs and the sharing economy bears examination.
Most LPWANs use unlicensed spectrums, which anyone (or more to the point, anything) can access freely. Organizations can set up gateways or base stations anywhere, which can, in theory, reach out to a shifting network of IoT devices across many square miles.
However, if there are too many gateways, they will begin stepping on each others’ signals and make the entire system inefficient.
One solution is to use a shared gateway model, rather than a proliferation of private gateways. The gateways are publicly accessible, and the costs are borne by the operators of the gateways. The costs consist of bandwidth and power used, and of course, buying and maintaining the hardware.
Gateway operation costs can be recouped in several ways. In time, the organizations who field IoT devices will contract with gateway operators to send the data through their devices. The cost per data packet will be much lower than with the current system, where data users maintain their own private gateways.
But, if your devices are using a shared gateway, who do you pay? The gateway operators are forced to work that out between themselves, becoming a kind of de facto public data grid. Contracting with any gateway operator would, in theory, give you access to the entire shared gateway network.
Aye, there’s the rub. What happens when the inefficiency is not caused by too many gateways, but too few? What happens when you start losing vital packets because the gateways are spread too thinly, or because there are just too many IoT devices accessing the system at once?
The short answer is that wait times grow from a few milliseconds to several minutes, and the networks become increasingly difficult to use.
MXC’s MXProtocol seeks to reinvigorate the shared LPWAN gateway model by adding a very interesting feature — Smart Bidding. IoT data users can place a maximum bid on the priority of their data. All the data will still be allocated time — it all gets where it is going eventually — but systems which require data to be delivered regularly, quickly, or with the maximum of reliability can bid higher.
This system allows anyone to use an already established network. They always have the option to purchase and maintain their own network systems. It’s just that if you have mobile systems you would have to cover the entire city with your own gateways. That can be impractically expensive and complex. In MXProtocol’s shared gateway system when a device being tracked goes out of range, it can communicate through someone else’s gateway. This does mean paying the gateway owner for access, but that is for access that the original operator would never have otherwise. Instead, it uses device resources that the owner of the next gateway would normally not need, and would be wasted.
Like Airbnb, the gateway operator can open up its unused bandwidth for bidding, and recoup some small amount of money for it. If the guest user bids high enough, the gateway owner might even delay their own low-priority use in favor of the payment.
Some users might be willing to pay as many as a few fractions of a penny per packet. They would be virtually guaranteed of getting the first crack at the system resources no matter what gateway they were using. Other users might not wish to bid very much at all. This would cause their data to be delayed slightly at times or places of peak traffic, but this will be perfectly acceptable to many users.
The sharing economy is about making more efficient use of resources. It means making access less expensive, by making smarter choices about how resources are used.
That’s easy. Everyone gets the same ‘premium’ service, only paying to access another network if theirs is out of range of the device in question.
That’s simple too. The customers who are using the capacity for low priority tasks and really don’t care about a little extra latency continue on like nothing is happening. They might have to resend a few packets, at the very worst.
But the customers whose devices are operating high priority devices or doing something else — anything else — that they feel is critical, have registered their willingness to pay a little extra to use another network’s capacity. They don’t have to make the decision when the congestion occurs — the system knows their preferences, does all the work invisibly, transferring any payments due directly between the users involved them at the end.
They enable users to accommodate more sensors to existing LPWANs without adding extra capacity, by making the prioritization of data automatic, transparent, and customer-led.
The biggest objection anyone can have is that it costs money, and it costs so very little money that it barely matters.
This article was originally posted here.