Spectrum Requirements and Technology Developments for M2M

With machine-to-machines communications set to grow rapidly in the coming years, this will cause pressure on the frequency bands, especially below 1GHz.

M2M traffic in wireless networks is expected to quadruple between 2015 and 2022. This is especially the case as machine type communications expand to different market segments. Energy, transportation, banking, healthcare, and consumer electronics are just some sectors in which the technologies are being advanced.

At the M2M workshop in Mainz in March, experts from across Europe gathered to discuss spectrum issues surrounding M2M.

Spectrum regulation is increasingly coming to the fore as regulators, industry and administrations try to meet the challenges and address potential issues created by the introduction of a multitude of M2M technologies into the market.

3GPP

The 3rd Generation Partnership Project has defined solutions for M2M. They include extended coverage GSM for Internet of Things (EC-GSM-IoT). Long Term Evolution cellular system and its variant LTE-M is another solution, which could carry a lot of the traffic for M2M communications.

Likewise, Narrowband Internet of Things (NB-IOT) is flexible and enables easy implementation, providing operational coverage and support for thousands of devices. Based on 200 kHz channel width it can easily be inserted in GSM, UMTS (universal mobile telecommunications systems) or LTE bands.

Interesting times lie ahead as 3GPP's Release 15 “pre-5G” will include M2M solutions. This is aimed for June 2018, with Release 16 coming by the end of 2019.

With a harmonised spectrum for mobile/ fixed communication networks (MFCN) already used for M2M, the question of whether there will be an impact on GSM switch off was posed at the workshop.

Dedicated spectrum – 400 MHz bands

A number of speakers expressed interest in a dedicated spectrum for LTE M2M networks for utilities. Such spectrum in the 400 MHz bands was emphasised, taking into account the propagation characteristics of these bands.

Colin Chandler, chair of the Standards and Regulatory Working Group within 450 Alliance, said 450 MHz is best positioned for critical connectivity. The 450 MHz spectrum should, he said, be exclusive, offer cost effective coverage, low latency and robust mobility among other things.

“Physical propagation properties of 450 MHz allow building and operation of wireless networks for critical M2M applications with low investment and operational costs,” said Mr Chandler.

Dr. Bernd Sörries, Chairman of the Information and Telecommunication Committee of the German Association of the Chambers of Industry and Commerce, said “from a technical viewpoint, lower frequency spectrum enables wider area coverage and better penetration deep into buildings”.

The ECC/WGFM was called upon to consider whether Long Term Evolution studies in these bands could potentially take into account narrowband IoT. Other technologies for dedicated networks for M2M applications were also named such as LTN (Low Throughput Network), RPMA (Random Phase Multiple Access), Wireless M-bus, and IEEEE 802.15.

Dealing with increased congestion

One major issue facing the influx of M2M technologies is the increased congestion on the networks. More traffic will be carried over cellular networks, as well as low power local and wide area metropolitan networks, causing pressure on the current use in frequency bands - especially below the aforementioned 1GHz.

While spectrum demand is predominantly for spectrum below 1GHz, it may be complemented in the future with higher frequency bands.

Dr Michael Sharpe of the European Telecommunications Standards Institute (ETSI) said it was important to ensure that M2M devices do not send their data all together and congest the network. He referred to some solutions like that in 2G, where the Core Network (CN) can notify the Radio Access Network (RAN) of a CN overload situation.

“An implicit reject mechanism can be used by the network to reject all devices configured for low priority access,” he said.

With the majority of M2M devices sending or receiving only small amounts of data, their power consumption shall be as low as possible. This, said Dr Sharpe, should mean that the use of network resources “remains as efficient as possible”.

The bands

From a regulatory point of view, commercial mobile bands may be suitable for providing many M2M services if certain conditions are met. The commercial networks may need to be 'hardened' or made more robust. For example, the radio link and network availability may need to be increased over time. In some cases dedicated networks may be necessary – especially when used for mission critical purposes such as smart electricity grids.

However, at the M2M workshop it was agreed that the M2M solutions will demand both, generally authorised spectrum as well as individually licensed spectrum (MFCN, Private Mobile Radio Systems (PMR), and Public Access Mobile Radio (PAMR)) as they are suitable for specific M2M requirements.

Not all solutions need to be treated the same in frequency regulation: solutions for short range / long range communication, including mesh network and ad-hoc network solutions may need to be treated differently. Similarly, some very asymmetric network solutions can also be treated contrarily.

Generally authorised spectrum

Demand for generally authorised spectrum was also discussed. There is a strong request for more harmonisation in 870-876/915-921 MHz (Seven ETSI System Reference documents have been created and many presentations at the workshop addressed this topic) for technologies such as 802.11ah, BTLE, and 802.15.

Dr Benoît Ponsard, Director of Standardization in SIGFOX, said his organisation believes dedicated networks for IoT should “remain on license-exempt network bands”.

“It's important to keep them on license-exempt spectrum because of business issues. We don't want to have fully dedicated spectrum for one technology, this is not the spirit of sharing the spectrum. But we consider that in the long run if we wanted to be very effective in the spectrum policy it's better to have technologies sharing the same time frequency of occupation.”

Hamid Reza Karimi director at Huawei Technologies' Corporate Strategy Department, said his firm believes “sufficient harmonised spectrum is available for licence-exempt use by the IoT”. “In Europe this includes the well-established 862-870 MHz, 2.4 GHz, and 5 GHz bands, and the new 870-876/915-921 MHz bands which are in a process of harmonisation right now.”

Current regulatory framework conditions with regard to M2M, such as considering separately licensed and license-exempt bands, were not questioned.

Individually licensed, but non-exclusive, spectrum access may be an option (especially in 400 MHz, 870-876/915-921 MHz). So far, some demand has also been expressed for license-exempt usage of the 1900-1920 MHz band (DECT community, and some SRDs). It could also be an alternative for some more critical wireless industrial applications.

Harmonisation

There was a strong request for more harmonisation in 870-876/915-921 MHz bands. Some 10 speakers addressed the topic. The 870-876 and 915-921MHz bands for SRDs, which is being underused in most European countries (see ECC Report 189), have been the subject of intensive studies within CEPT, such that now, new entries have been made on CEPT Recommendation 70-03 encouraging its release.

Figure: Current ERC Recommendation 70-03 – 870-876 MHz

One such speaker was Dr Joachim Sachs, Principal Researcher with Ericsson Research, who said harmonisation was important.

“If you really want to have a successful market that will enable economy of scale, it's really important that there is harmonisation of the rules within the EU and also on a global basis,” he said.

CEPT announced that an Addendum to CEPT Report 59 (6th update of EC Decision on SRDs) is planned for the end of this year.

For this, there is also the need to find a balance between national flexibility and the right level of EU harmonisation.

There are also some studies on-going and changes in the future may occur for the regulations, for example the introduction of more flexibility for the duty cycles.

Duty cycles

The suitability of duty cycle definitions for M2M was taken into account. Some devices may normally transmit at very low duty cycles but have a need for a higher duty cycle when a specific event happens. For example, in cases of emergency or when an alarm goes off.

PMR solutions were discussed, notably in the 400 MHz band, in particular for mission-critical M2M applications and platforms. Likewise, they could be presented where there is a need for a larger degree of customisation of the network to the needs of the M2M network provider.

Different models, dedicated and shared IoT networks, also hybrids, are possible, and may be driven by business case needs. This could also trigger the need to find synergies for common national platforms.

Shared spectrum access

Another option could be the appropriate authorisation or notification for shared spectrum access (especially in 400 MHz, 870-876/915-921 MHz). This could, it was suggested, achieve coexistence between M2M and the incumbent application, as well as between various M2M networks (see also ECC Report 132 for these options).

Over the course of the two days we heard from speakers from a variety of sectors. It was clear from the presentations that building, home automation, smart metering, intelligent transport systems, and wireless industrial sectors will be key areas when it comes to M2M. Even new market sectors could be developed by M2M applications such as remote healthcare and smart agriculture.

With this in mind, extra high voltage and high voltage smart grids will require enhanced communications and resilient systems. A low voltage smart grid would be appropriate for licence-exempt M2M spectrum.

Dr Simon Dunkley from Silver Spring Networks talked about metropolitan and rural area networks which use the Wi-Sun mesh network and how a smart grid can lead to a smart city. He looked at examples like Glasgow where sensors gather data and implement a dimming operation on street lighting based on pedestrian footfall, vehicular traffic, noise measurement and environmental sensors.

He called for consideration of the requirements for mesh-networks (self-organisation, indoor/outdoor frequency agile systems), as well as spectrum requirements: a minimum of 3 MHz would be required based on his experience.

Constraints and needs

There are several challenging needs for some of the M2M applications. Massive deployment is one issue, i.e. the deployment will see up to several thousands of devices per square metre in metropolitan areas. But also in some cases – most notably when it comes to high voltage smart grids - very high reliable and very low latency communications will be needed.

Some questions still need to be answered for future road safety related intelligent transport systems. For example, how could the interoperability/coexistence of different technologies be ensured (ITS-G5, LTE-V2X) in the band 5875-5905 MHz?

In this regard, the views from the automotive industries are certainly relevant. Hybrid solutions with vehicle sensor platforms will also play an important role in future developments in road safety.

Interoperability between M2M devices should be achieved, and spectrum for Wireless Industrial Applications in the range 1.4 GHz to 6 GHz (about 80 MHz, e.g. 2 x 40 MHz) may need some review, by taking into account global requirements.

One thing's for certain lots more discussion will be needed as the IoT and M2M communications become ever-more commonplace in our society.