Despite considerable success, as evidenced by the plethora of innovative solutions for consumers, companies and society, there’s a fundamental issue that the M2M industry, particularly the MNOs, chooses to ignore. Today’s wireless networks are not truly fit for an intended purpose: they do not and cannot enable the optimum delivery of M2M and IoT applications.
Cellular networks, for example, were designed for voice and high-volume data traffic, not intermittent payloads of under 30 bytes only requiring a throughput of 100 bps. Moreover the communications overhead can be 500 to 600 bytes – well over an order of magnitude higher.
That particular issue has led to the development of lightweight protocols like MQTT (Message Queue Telemetry Transport) and CoAP (Constrained Application Protocol). But protocols are only part of the problem. The real issue is intrinsic: in order to take M2M and the IoT to the next level and enable a cost-effective, long-term future, we need a new kind of network.
The Optimal Network
This is a summary of the shopping list proposed by SIGFOX:
- Low cost and low energy consumption – the latter is needed to increase battery life
- Ease of use – including device management and integration with IT systems
- Frequency independence – thereby facilitating cost-effective world-wide coverage
- Embedded subscriber identification – no SIM cards, no need for subscriber lifecycle management
The latter requirement indicates that regular cellular technology cannot deliver the requisite functionality. The new network has to be designed from the bottom up using a relatively new technology and a new business model.
It has to be dedicated to M2M and the IoT.
SIGFOX employs a patented radio technology that is currently based on UNB (Ultra Narrow-Band), which employs the license free ISM bands. However, the solution is not dependent on ISM, and other frequencies such as white space could be used in future when it’s available and usage is legal.
UNB enables the cost-effective transmission of data over a very narrow spectrum to and from connected objects. The devices have outstanding sensitivity, which in turn minimizes the number of antennas (base stations) needed for wide area coverage.
SIGFOX deployed a nationwide network in France using 1000 antennas, which the company financed for a very modest 3 million Euros.
Currently, each base station can handle up to a million battery-operated devices. Cell density is based on an average rural range of about 30-50 km (18–30 miles); 3-10 km (0.5-6 miles) for urban areas.
Energy consumption is 200 to 600 times lower than an equivalent cellular network, and the ratio of payload data to protocol data is normally between 20 to 50 percent. In cellular networks, it would be 1 percent or less.
Furthermore, each of the modems used in the network (to emit and receive data) are significantly less energy intensive than devices using competitive technologies. Network performance for M2M traffic is therefore more efficient and communication costs are much lower, basically because the network is fit for its dedicated purpose.
In a nutshell, the networks of MNOs are colossally over-engineered for almost all M2M applications.
The wireless modems for cellular networks are sophisticated, multi-functional products: UNB modems don’t need all that functionality, e.g. multi-band air interfaces and SIM cards. Therefore, they cost less to produce and are easier to deploy; they’re operated via an AT command set. UNB modems are available from qualified partners, like Silicon Labs, and SIGFOX indicates that the major chip manufacturers are adding support for their protocol, which will encourage more vendors to take the technology on board and in turn reduce costs even more.
UNB: A Game-Changer?
There is clearly a huge momentum behind current M2M technologies, but the business case for cost-effective and energy-optimized connectivity is compelling, and not limited to specific sectors. That said, there are obvious targets such as fire detection and automatic meter reading (AMR).
The latter is a low-throughput, low-data rate application that involves collecting data millions of home and businesses, and cost is the gating factor.
The favored architecture involves RF modules embedded in the meters, which are connected in a neighborhood area network to a concentrator that communicates to the utility company via a cellular network. Therefore, there is no need to have SIM cards inside every meter.
UNB realizes the same objective, but does so in a much simpler and more effective way.
For example, Plextek has designed a cost-optimized base station for AMR, each of which can accommodate up to 5k meters. Four base stations connect to a controller unit and the traffic is backhauled over ADSL or another IP-based network.
The company’s protocol features forward error detection and correction, automatic retransmissions and power control.
SIGFOX has deployed a fire detector application where the smoke detector object transmits daily heartbeats to ensure the alarm will work when most needed, and in the event of a fire, sends alert messages. This is a third- party application that connects to the company’s managed service through a standards-based API in order to receive the messages.
Meters and smoke detectors don’t move around, so roaming isn’t relevant, but it will be required if UNB is to take off and deliver its full potential.
As indicated, building a network is not expensive, and there are many applications that only need citywide access, and that, according to SIGFOX, can be financed by revenue from the business project.
One can thus foresee international networks being formed via alliances.
There’s a lot to like about this development, and time will tell how fast and how far it will go. In the meantime, in common with other game-changing technologies, we can expect the concept to be trashed by the competition. Most MNOs were late to the M2M table, primarily because the ARPUs were so low compared to voice and data traffic; they chased the big consumer numbers but now those numbers apply to apps such as AMR so that particular revenue stream is under threat and more will follow.
It’s imperative that we watch this space.
Edited by Braden Becker