We love phone numbers at Neustar, of course we do. In fact, Tom has shown in Part 1 of this series, Telephone Numbers Are For People, Not Machines, that in the particular case of Machine to Machine (M2M), using phone numbers to address those devices and machines (and not people), can be short sighted. My colleague, James, has explained in Part 2, How to Name Your M2M Device, why the industry currently promotes phone numbers for M2M devices and what we’ll need to incorporate standard naming practices going forward.
It’s not enough to explain what is not working or why it won’t work in the future, if we don’t come up with something that does! Accordingly, I will explore some traits of M2M devices, and consequently derive the characteristics of a future-looking addressing plan for them.
Following that, I will show some contemporary internet-based addressing solutions that seem to comply with those requirements, and show how phone numbers, and the International Mobile Subscriber Identity (IMSI,) fall short. Lastly I’ll briefly discuss how some industry forums are seeing the future of M2M device addresses.
The Future of Addressing
M2M devices are extremely versatile. They do use voice, text and data to communicate with back-end servers, peer machines, or hubs but that’s roughly where the resemblance to a human using a phone ends:
- Some devices are expected to work unattended for up to few years, e.g. those used in industrial areas or in long underground piping systems. Replacing a battery or SIM card is not an option.
- Additionally, SIM may be embedded to lower chip costs or prevent fraudulent usage (taking the SIM of the meter for some long-distance calls).
- Some devices are happy with very short messages, using 2G SMS, with very infrequent usage (e.g. monthly telemetry average report for a sensor,) while others may heavily bring real-time voice and video, like surveillance equipment.
- The number of such devices is expected to be 25-50B by 2020. Even those that believe it is still proportional to the earth populations (like perhaps telephone usage), see a ratio of x4-x8 per person, once you consider the various industries (utilities, motor, health, home appliances, environment and more.)
That’s a lot of machines.
Please do not rage against the machines.
Here’s the thing. Machines, unlike people, do not have to “dial” a number or have a relatively “short” representation in order to remember. Furthermore, there is no need for a human facing MMI / UI (such as a small keyboard with numbers only,) so a small character set is not mandatory.
Meaningful and readable addressing with full text may assist in operational and maintenance administration, as much as HTML and XML assisted software developers in understanding what happens under the hood, as a clear plain English identifier can be associated with the device.
Will you understand an SMS from 150035832795 saying “1″, or do you prefer dishwasher@m2m.myhouse sending “Finished, 183min?”
Clear Communication
Furthermore, as M2M is inclusive of Machine to Human (M2H) and Human to Machine (H2M,) people will appreciate non-cryptic identification. Speaking of clear plain English, as communication is global and there is a clear shift of power to China, i18n (internationalization) can be applied in the same way that former English DNS has been adapted to include other languages with International Domain Names (IDN). Also, with respect to DNS and the Internet in general, we are all more used to identifiers like email and web addresses such as URLs.
A system that will allow us to address an enormous and fast-paced growing number of machines is required. “IoT”, the internet of things, visualizes a future with every device connected. This is inspiring. And challenging. Phone numbers were not designed for that. They will exhaust and betray the real people who need them for phone calls: You and me!
A clear separation of the communication service provider (wireless carrier, cables operator, or future technologies such as weightless) will also provide architectonic robustness, and benefit end users allowing easier portability and switch to a better or a cheaper service. As telephone numbers and IMSIs are tied to carriers, this degree of freedom can only be achieved by using intermediary identifier, such as Uniform Resource Identifier (URI), URL, Network Access Identifier (NAI), or user@domain.
Note that industry forums such as standardization bodies (ETSI), the European Committee for Communications, and others, recognize that Internet identifiers, such as IPv6, would be used for connected devices going forward. They also encourage this flexible intermediary, as commonly used in internet today with DNS; devices have IPv4 and DNS allows a readable representation to be mapped.
Lastly, as M2M technologies may go well beyond current known mobile protocols. What can be predicted (foolhardy as it is) is that SS7 is moving away, and IP is more prevalent. Using a mobile identifier for addressing, such as IMSI, creates a fragmented addressing space where anything that relies on tested and proven IP is a better prospect.
Clearly some form or URI / URL / NAI such as m2m://home.fridy.com/meters/electricity/british-gas/1 together with IPv6 connectivity, can provide portability, endless space, readability, internationalization, robustness architecture, distributed handling, and much more…
In the next installment of this series, Mark will show specific vertical segments of M2M solutions and how they can be successfully implemented with the vision presented in this post. Modern URI based Internet-pace addressing to allow versatility, portability, and sustainability. So you have that to look forward to!
