Work begins to standardize 'internet-of-things' protocol

Work begins to standardize 'internet-of-things' protocol

Summary: OASIS launches technical committee that promises to have a widely accepted machine-to-machine connectivity protocol available in about a year.


OASIS has announced a new technical committee is being formed to formalize a standard protocol for machine-to-device-to-sensor-to-refrigerator-to-other-machine-somewhere-else-on-the-network interactions, otherwise known as the "internet of things.' 

Network-Lattice photo by Joe McKendrick

The protocol, "MQ [Messaging Queue] Telemetry Transport," or MQTT, is described on the site as a "machine-to-machine (M2M)/Internet of Things connectivity protocol." The protocol, designed as an "extremely lightweight publish/subscribe messaging transport," is intended to facilitate "connections with remote locations where a small code footprint is required and/or network bandwidth is at a premium." Examples include "sensors communicating to a broker via satellite link, over occasional dial-up connections with healthcare providers, and in a range of home automation and small device scenarios."

MQTT is also well-suited for mobile applications, proponents say, due to "its small size, low power usage, minimized data packets, and efficient distribution of information to one or many receivers."

Having a universally accepted and adopted M2M protocol will help get new devices and systems to market faster, since they often are built using many variations of hardware and software platforms, device types, and networks, OASIS notes. 

MQTT was invented by Dr Andy Stanford-Clark of IBM, and Arlen Nipper of Arcom (now Eurotech), back in 1999. It also has been called the “SCADA protocol,” the “MQ Integrator SCADA Device Protocol” (MQIsdp), the “WebSphere MQTT” (WMQTT).

OASIS' MQTT Technical Committee will work with MQTT as its base document to "define an open publish/subscribe protocol for telemetry messaging designed to be open, simple, lightweight, and suited for use in constrained networks and multi-platform environments."  OASIS has scheduled the first, in-person  meeting to be held in Boston on Monday, 25 March 2013, to be hosted by IBM.  A working specification will be completed by March 2014, OASIS says.

The MQTT TC is intended to complement previous work by the OASIS AMQP Technical Committee, which released a specification that provides for transaction and publish & subscribe messaging between autonomous businesses, departments and applications using an open protocol for enterprise middleware. The MQTT specification adds a means "by which sensors, control systems, embedded systems and mobile devices can publish and subscribe low-level, technically-orientated data," OASIS says. "There is natural affinity to bridge MQTT with AMQP, so as to connect telemetry with enterprise applications."

(Photo: Joe McKendrick.)



Topics: Cloud, IT Employment, Enterprise 2.0

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • What's wrong with TCP/IP?

    Sounds very much like a wasteful reinvention of the wheel - what's wrong with TCP/IP?
    • What's wrong with TCP/IP

      My understanding is MQTT runs on the application OSI layer, not network or transport layers. I think that distinction is critical, it implies that you could choose other lower level mechanisms for networking and transport that are better suited to the environment that it lives. Maybe these live in ad-hoc networks, possibly your car briefly talking to adjacent cars about their telemetry to better anticipate collisions; or maybe sensor packs in a jet turbine reporting temperature data to better direct maintenance before a failure. Clearly, TCP/IP might be overkill in a multitude of scenarios.
    • Absolutely not a reinvention!

      MQTT is optimised to deal with networks that are low-bandwidth and potentially unreliable. As the other commenter pointed out, it runs on top of the transport/network layer (in the case of MQTT, it actually runs on top of TCP/IP - the related MQTT-S protocol doesn't need TCP/IP). Think of it more like a push protocol for small devices, sensors, etc.
  • 6lowpan

    "not a reinvention"
    IPv6 already includes a specific layer for sensors that are "low bandwidth and potentially unreliable", so I'm still unclear as to what MQTT is bringing to the table. Will MQTT work seamlessly with 6lowpan or is it a competitor in many use cases?
    • MQTT is reliable pub/sub messaging for mobile

      Runs atop TCP, so complementary to 6LoWPAN. On 3G Android is 10x less power, 8x less bandwidth overhead and 93x faster throughput than HTTPS. And has QoS built-in.
      Joe Speed