oneM2M: Devising an interoperability framework for IoT

The Internet of Things will not be able achieve the necessary scale or flexibility it needs to meet multiple use cases unless an interoperability framework can be established. Wireless editor James Atkinson talks to Ultan Mulligan at oneM2M, the body working to produce an interoperability enabler for IoT

oneM2M: Devising an interoperability framework for IoT

The vision of the Internet of Things (IoT) is to share data originating from individual verticals across a wide range of different applications and end users. There is a huge amount of work going in to developing M2M/IoT connectivity solutions, devices and applications, along with attempts to standardise these efforts.

But if the vision of the IoT is to be fulfilled there also has to be some kind of interoperability solution that enables applications to communicate with, co-operate with, and reuse one another’s data or components.

This is the goal of oneM2M – to create an interoperability enabler. As oneM2M’s January 2015 white paper, The Interoperability Enabler for the Entire M2M Ecosystem puts it, what is required is ‘a framework for all these machines to integrate and interact with one another efficiently and effectively’.

oneM2M’s goal ‘is to develop technical specifications which address the need for a common M2M service layer that can be readily embedded within various hardware and software and relied upon to connect myriad of devices in the field with M2M applications servers worldwide’.

A new framework
This kind of framework is needed to deliver the scalability and flexibility the market needs to enable M2M/IoT to maximise its potential, oneM2M argues.

At the core of oneM2M’s approach are two main tenets: providing an interworking framework; and enabling re-use of what is already available as much as possible. What it is not trying to do is provide a standard across the whole M2M/IoT environment of networks, applications and devices.

oneM2M is made up of eight major world standardisation bodies including: ARIB, ATIS, CCSA, ETSI, TAI, TSDSI, TTA and TTC, and it also works with the likes of The Broadband Forum, Continua, Global Platform, Home Gateway Initiative, New Generation M2M Consortium (Japan) and the Open Mobile Alliance (OMA).

Speaking to Wireless, Ultan Mulligan (pictured), Marketing Committee Vice-Chair at oneM2M and Director of Communications at ETSI (European Telecommunications Standards Institute), says: ‘All the analysts’ projections and predictions are for massive deployments of M2M/IoT devices even though they have been paring back on the numbers a bit over the last year.

‘But if you put yourself in the shoes of someone looking to invest in M2M now, may be working with a technology developer, and you want to bring forward an M2M application or invest in one, you’ve got to deal with multiple industrial groupings, protocols and brand names – it is a huge environment to navigate,’ he observes.

‘So, who should you back? Which technology should you adopt? Maybe you don’t gamble yet because you can’t see a winner. Or, you just invest for the short term so it doesn’t matter if the solution you choose is a dead end in five years time, as you got your money back within two years,’ says Mulligan.

IoT confusion
He points out that this kind of confusion abounds at all levels of IoT at the moment and the situation is not going to improve in the near term, especially with the proliferation of low power wide area network protocols emerging; some narrow band solutions from independent vendors or organisations, along with different cellular derived options progressing through the standards body 3GPP.

Mulligan says: ‘There are more and more use cases emerging and many more to come and with many different radio interfaces closely aligned to specific use cases such as: radios for long range, low throughput; radio interfaces tailored for short range environments; interfaces to meet very high reliability requirements; high security and authentication for industrial use cases; up to 4k wireless video streaming applications; and we are even seeing infrared interfaces being used for things like tags that display prices in retail environments.’

Which layer of the stack?
Mulligan says he has difficulty seeing convergence in terms of the air interfaces. ‘We may see one or two dominate some particular applications or sets of applications, but it won’t happen for a long time yet.’ That means standardisation to enable interoperability has to take place elsewhere within the M2M/IoT stack.

Hence, the need for oneM2M to find an interoperability solution that enables all these different air interfaces to co-operate without forcing anyone to change their technology.

‘If we are to see the growth of M2M as predicted by all these analysts we really need to see some economies of scale, so we are trying to enable this at the service layer; that interoperability layer between the silos if you like of each activity technology,’ says Mulligan.

He notes that at the moment there is not much access outside of the individual technology silos that will allow applications to make use of different connectivity technologies and understand what those different technologies are.

‘At oneM2M we are trying to provide a platform to enable app to app communications and of course ensuring that it is all done securely and that privacy demands are taken into account,’ says Mulligan.

The interoperability issue
In the white paper cited above, oneM2M take the example of the mining industry where there are a number of suppliers of heavy plant like JCB and Caterpillar. They make use of M2M technology to gather data on the performance of their machines to help with product improvement. But the mine operators have no access to that data; or, if they want to get it, they have to install their own proprietary systems.

‘Each manufacturer will have their own dedicated M2M management platform, so the data is collected in a different way and in a different format dictated by each manufacturer and you cannot easily compare plant performance between vendors or across different mining sites,’ explains Mulligan.

oneM2M does not prevent each manufacturer using their own dedicated M2M solution, although they could use oneM2M’s platform interfaces if they so choose, but it enables them to access and compare data from different platforms and devices.

That way the mining company gets information on productivity such as which machines work best in which situations, which are most effective and which are the most durable. The plant manufacturer uses the data to enhance and develop its products and receives information that enables it to provide timely upgrades and preventative maintenance.

‘We do this by trying to deploy commonly used industry protocols or management platforms for M2M/IoT devices by mapping our solutions onto communications protocols already in use. There is a lot of legacy M2M already out there, especially in industrial environments that needs to be supported,’ says Mulligan.

Release One specifications
oneM2M published its Release One in January 2015, which comprises a set of 10 specifications covering: requirements, architecture, API specifications, security solutions and mapping to common industry protocols such as CoAP, MQTT and HTTP. It also makes use of OMA and Broadband Forum specifications for device management capabilities.

Release 2, due out in the middle of 2016, will focus specifically on the home and industrial environments and will also deal with security issues. It will also look in particular at the semantics of interoperability; i.e. how you identify the data that you are managing and how you put meaning into the data.

Mulligan explains: ‘Different apps will want to expose different types of data and send it to different end users, so we need to look at the security and privacy issues. To enable the reuse of data by different apps or enable one device system to be used by another you need semantics. You need to be able to understand what this data actually is. To do this we are making use of other standards that are already out there.’

Dealing with security
He adds that security is a key consideration for oneM2M. A security architecture is defined in Release 1 with security functions covering identification, authentication, authorisation, security association, sensitive data handling and security administration. There is a large variety of deployment scenarios to take into account, involving any type of device.

So, oneM2M must offer multiple authentication and provisioning options and modular security according to the needs. The first release provides hop-by-hop security over each communication link. End-to-end sensitive data handling will come in later releases.

‘We are building in support for oneM2M specifications with other organisations. For example, the Open Mobile Alliance’s Device Management along with the Broadband Forum’s device management protocols are integrated in Release 1 already. Interworking with OMA’s Lightweight M2M protocol is being included as part of Release 2.

‘We are also mapping down to technologies such as CoAP (used to standardise the way network edge end devices hook up to web applications via gateways) and MQTT (a lightweight publish/subscribe messaging transport protocol, used a small code footprint is required and/or network bandwidth is at a premium).

Industry co-operation
‘We are working with organisations like the AllJoyn open source project (managed by the AllSeen Alliance) and the Open Interconnect Consortium (OIC), both of which are driving work to enable device interoperability,’ continues Mulligan.

‘We held a work day with them earlier this year and are actively working with them on interoperability at the device ecosystem level. So, we realise that technologies like AllJoyn and OIC can work well with oneM2M and present a coherent approach to interoperability issues.’

Looking ahead, Mulligan says: ‘It is quite clear that there is no shortage of possible standards, but it will be a while before we begin to see the real winners and losers. This industry is still in its infancy and different parts of the M2M/IoT stack are at different levels of development.

‘We can certainly see that there is a lot of competition at the air interface level, but oneM2M is not trying to pick winners and losers or trying to align itself with anyone. Instead, we are trying to federate what is there and build things for valid business reasons, so that we have interoperability solutions that can handle legacy M2M systems and the new ones coming along in the future.

He concludes: ‘The industry is realising it must take the lessons of interoperability, security and privacy on board or you will not last. That is true of platforms, devices, applications and so on. How you handle security will sort the men from the boys. But we must have this interoperability platform, because if we don’t M2M/IoT will not fly as high as it can or reach its potential.’

For more information on oneM2M visit: http://www.onem2m.org/

 

 

Leave a Comment