原文地址：http://blog.m2mconnectivity.com.au/2015/01/29/cellular-networks-vs-custom-m2m-networks-who-will-win-in-the-iot-space/

Cellular Networks vs Custom M2M Networks – Who will win in the IoT space?

29/01/2015 by m2m connectivity Leave a comment

According to Machina Research over 20 billion devices worldwide will automatically relay information over the Internet by 2022. Industry pundits are divided as to which wireless communications will best serve the burgeoning Internet of Things (IoT) landscape. Those that see existing cellular networks as the best option typically promote 4G/LTE to facilitate the IoT. Other industry experts advocate the need for new networks that are M2M/IoT specific.

Traditionally, cellular communications were perceived as ill-suited to the low cost and low power characteristics of many M2M/IoT applications. However, technologies that leverage existing networks make sense and 4G/LTE is being promoted as the smoothest path to adoption among carriers who are eager to turn off their less efficient 2G networks. But are these cellular networks ever likely to be suitable for low cost low power M2M use?

Most M2M applications do not need the higher bandwidth of 4G/LTE with the bandwidth requirements per device being very small, even in a fully realised IoT system that generates huge amounts of data produced by billions of sensors. There are standards being developed and promoted within LTE that may well give LTE operators the ability to provide low data costs through efficient use of their existing LTE bandwidth for these IoT applications. If the operators choose then perhaps they will be able to reduce IoT user costs to a level to really help drive the market.

A key component here is the new LTE-M standard that may offer low power consumption (up to five years for a device running on AA batteries), easy deployment, interoperability, low overall cost, and excellent coverage.  The case for 4G/LTE networks for IoT is further supported by the availability of smaller, simpler LTE modules working on LTE-M that could, in the future, sell for as little as $5-10, with lower max data rates and implemented power saving techniques.

On the opposite side to the LTE for IoT argument we see new dedicated wireless networks being proposed and established with the aim of providing M2M connectivity in ways that carriers do not provide today. These companies will face competition from entrenched cellular operators, and their success will depend on having a compelling offer to the IoT market on pricing of hardware and service, better features such as power consumption, and providing the coverage required.

In a world of start-ups, Sigfox is a leader. This French company has already built nationwide networks for M2M in France and the Netherlands and a network for the UK is in the works.

Costs involved in building entire networks could seem prohibitive but Sigfox’s Spanish network is estimated to cost less than €15 million and will take about seven months to complete. Millions of home-security systems in Spain will soon connect to this nearly completed network.

Sigfox hopes to raise more than $70 million to build a national network in the US for the IoT. Purchasing the rights to use spectrum is a major expense of building a cellular network. Sigfox uses licence-free ISM bands instead then sells data plans for between $1 and $12 per device per annum. Batteries in these devices can last for years although there is a limit of 140 messages of 12 bytes per day, this is sufficient for many connected devices.

Competitors point out that the software of these start-up companies is proprietary. In the case of Sigfox, it has gathered support from a handful of companies, including the carrier Orange, to draft an ad hoc specification released through the ETSI standards group and is working on a formal standard.

US based start-up, On-Ramp Wireless has installations in the US, Africa, Europe, Asia, and Latin America. Most of these are private customer-run networks in the electricity, gas, and oil industries. As a founding member of the IEEE 802.15.4K task group, On-Ramp Wireless hopes to develop an open standard for low-energy critical infrastructure monitoring.

The Weightless Special Interest Group (SIG) is developing a wireless radio standard using TV white spaces spectrum. Weightless is being developed as a royalty-free ‘open standard’ and Version 1.0 of Weightless was published in April 2013. Weightless SIG has since been working on the test and certification process to ensure that chips and equipment from a mix of suppliers will interoperate – important for a budding networking technology.

Dedicated wireless networks for M2M have substantial potential and are gaining traction. These networks can be very applicable for industrial, agricultural, transport, and other sectors for a number of use cases. To date, the lack of wide coverage and an open standard hampers the wide adoption of these networks.

Meanwhile the number of M2M devices communicating on existing cellular networks is growing. Acquisitions and partnerships in the industry will continue with vendors augmenting their existing portfolios and developing new integrations between IoT platforms and core enterprise applications. Will 4G/LTE networks or custom M2M networks be the communication technology of choice for M2M deployments? The revenue stakes are high and the answer will become evident in the next 2-3 years.