Market Insight

Operators face an (over)abundance of options

October 27, 2015

Sam Lucero Sam Lucero Senior Principal Analyst, IoT

Want to learn more?
Have an expert contact you.

Low Power Wide Area Network (LPWAN) technologies are a key area of focus for operators in building their overall M2M/IoT strategies. In the past year, a number of new LPWAN technologies have become available and even more are being considered for standardization by 3GPP. This plethora of LPWAN alternatives poses a challenge for operators as they seek to choose the best option going forward. This article examines the evolving LPWAN technology landscape and suggests an approach for operators in selecting an LPWAN technology.

LPWAN technology options: an overview

LPWAN represents a solution to the performance gap between wide area cellular technologies, like GPRS, WCDMA and LTE, and local area short range wireless (SRW) technologies like Wi-Fi and ZigBee. LPWAN is meant to provide connectivity at lower price points than cellular by optimizing M2M/IoT use cases with low data rate requirements (along with other benefits to be described below.) At the same time, LPWAN enables long-range, wide area, point-to-multipoint connectivity at distances of 10 km to 30 km, providing a simpler deployment model than Wi-Fi over long distances by removing the need for complicated mesh topologies and extensive wiring to SRW access points. (Please note that there is discussion in the 3GPP organization to modify the 3GPP LPWAN proposal [discussed later in this article] to include multi-hop capabilities as a means of range extension. While this may not be an option for Release 13, it should be available by the time 5G is standardized, which is projected to be around 2020.)

IHS groups LPWAN technology and its vendor landscape into three broad segments:

  • Proprietary technologies – Some LPWAN technologies will remain held closely by their creators, whether by choice or circumstance. Some of these, such as SIGFOX and Ingenu’s RPMA technology, will form the basis of a public network offering. Others, such as Telensa’s Ultra Narrow Band, will be used as the basis for the creator’s own end-to-end M2M/IoT application offering.
  • De facto standards – Other LPWAN technology creators are seeking to establish de facto standards in the market. Semtech is trying to replicate the success of Bluetooth SIG with its creation of the LoRa Alliance to promote its LoRaWAN technology. NWave has offered its LPWAN technology as the basis for the Weightless SIG’s Weightless-N standard. As with the “proprietary” segment above, these “de facto” standards utilize unlicensed spectrum—typically in the 900 MHz band, but in some cases also in the 2.4 GHz and 433 MHz bands.
  • Cellular standards – Finally, efforts are underway to develop a formal 3GPP LPWAN standard. A key focus of this article is to explore the progress, including the options becoming available to operators through the 3GPP process. The overall goal is to standardize a technology that offers the benefits of the LPWAN variants listed above, but uses licensed spectrum for increased reliability and security and integrates more fully with operators’ existing core networks and services (e.g. mobile broadband).

3GPP LPWAN standard

In 2014, the 3GPP GERAN committee initiated a Study Item to develop a licensed LPWAN standard for the cellular industry. Such a standard will have the benefit of being fully open to the 3GPP ecosystem, with global scale and multivendor sourcing availability. Furthermore, it will reuse operators’ existing (and very expensive) core network and spectrum license holdings.

There were originally five major technology proposals made to the 3GPP GERAN committee for consideration to become the 3GPP LPWAN standard. However, two of these options—SIGFOX and LoRaWAN—did not move forward in the 3GPP process to become part of the eventual 3GPP LPWAN standard. Nevertheless, they are still available as technology options to operators and are described below:

  • SIGFOX – SIGFOX is an Ultra Narrow Band (UNB) technology developed by SIGFOX (HQ: Labège, France). SIGFOX received a massive $115 million funding round in early 2015 and is rapidly deploying its technology to provide public LPWAN access services for M2M/IoT applications in a number of countries throughout Europe, with deployment also projected to start in the United States in late 2015. While key operators have invested in SIGFOX, including Telefonica and NTT DoCoMo, it is working mostly through alternative connectivity service providers (CSPs) such as Aerea and Arqiva. Current deployments of SIGFOX technology are using unlicensed spectrum, but the company proposed SIGFOX as the basis of the 3GPP LPWAN standard.
  • LoRaWAN – LoRaWAN is a narrow band digital spread spectrum (DSS) technology developed by Semtech, a semiconductor maker. Semtech has gained significant market traction for the use of LoRaWAN for both private network and public network deployments through the creation of the LoRa Alliance in early 2015. Major technology vendors, such as IBM and Cisco, and operators, such as Bouygues Telecom and KPN, have joined as members. Semtech is currently the only LoRa chip producer, but says it will license the technology to other chip makers. While LoRaWAN public networks are currently being deployed in unlicensed spectrum by both operators and alternative CSPs, Semtech has proposed LoRaWAN as the basis of the 3GPP LPWAN standard.

The three technologies that are moving forward in the 3GPP process are GSM Evolution, LTE Evolution, and “Clean Slate”. As described below, GSM Evolution continues in the 3GPP GERAN committee as an enhancement of GSM, while LTE Evolution and Clean Slate have just recently effectively been joined together to form the basis of the evolving NB-IoT standard in the 3GPP RAN committee.

  • GSM Evolution – GSM Evolution is an effort to reuse underlying GSM technology in conjunction with a CDMA overlay to get lower power consumption and deeper coverage, albeit at the cost of greater complexity. While it coexists with current GSM network infrastructure, it requires modification of the remote device radio and is not a simple firmware upgrade. Ericsson is a key proponent of this approach. (IHS has adopted the GSMA’s term for this technology—“GSM Evolution”—although it is also known as “EC-GSM” or “Extended Coverage GSM”.)
  • LTE Evolution – LTE Evolution was an effort to reuse underlying LTE technology, with further refinement of its features and functionality for M2M/IoT applications beyond those currently envisioned for LTE-M and to enable LTE to operate in 180 MHz spectrum bands. Ericsson, Nokia Networks, and Intel were among the key proponents of this approach among a large number of supporters, including many operators. (IHS has adopted GSMA’s term for this technology—“LTE Evolution”—although it is also known as “NB-LTE”, or Narrow Band LTE.)
  • “Clean Slate” (CS) – The Clean Slate (CS) proposal was first made by Vodafone and essentially became an integration of technologies originally proposed by Huawei and Qualcomm. As the name suggests, CS moved away from adapting current standards for M2M/IoT and sought to create an optimal LPWAN standard by integrating fundamental technological building blocks. (IHS has adopted GSMA’s term for this technology—“Clean Slate”—although it is also known as “NB-CIoT”, or Narrow Band Cellular IoT.)

Features of the new “NB-IoT” proposed standard

In September 2015, 3GPP was able to harmonize the “LTE Evolution” and “Clean Slate” proposals into a single “NB-IoT” (Narrow Band IoT) proposal that moves from a GERAN Study Item to a RAN Work Item and should be finalized as a standard with Release 13 due in March 2016.

The NB-IoT proposal offers the following features and functionality. These are analogous to the technical goals of most LPWAN technologies, both for licensed and unlicensed spectrum use, so the description below will give a good overview of the technical capabilities of LPWAN technology in general.

  • UNB (~200 kHz) wide area connectivity, initially re-farming licensed GSM channels and in-band LTE carriers and potentially reusing LTE guard band spectrum.
  • Support for a dramatic increase in the volume of nodes per cell site (targeting >50,000 nodes per cell site in an area of 0.86 square kilometers).
  • Extremely low power consumption targeting >10-year battery life supporting one 200-byte message per day using a 5 Wh battery.
  • Enable device costs comparable to or lower than current GPRS devices costs.
  • Provide extended coverage for low data rate services into challenging locations, such as electricity meters located in basements or water meters located underground.
  • Latency insensitive transmission of less than 10 seconds; NB-IoT will not be applicable for latency-sensitive, real-time applications.
  • Limited data rate and small payloads, up to about 30 kbps, making NB-IoT most suitable for sensor monitoring applications.
  • Limited mobility based on cell reselection rather than inter-RAT handover, so not applicable to applications that rely on seamless mobility.
  • Unlike some other LPWAN options, NB-IoT will not require an overlay core network. Instead, operators can deploy NB-IoT leveraging existing infrastructure developed for EPC with optimizations.

Based on the technical characteristics described above, it is clear that NB-IoT will not be suitable for all M2M/IoT applications. Rather, NB-IoT will form the basis of a 3GPP LPWAN standard that complements LTE-M and, indeed, other existing cellular standards by providing an option optimized for mostly stationary, very low data rate, latency-insensitive applications such as smart metering, remote sensing, and basic telemetry, especially in challenging environments. NB-IoT will form part of an overall portfolio of cellular technologies available for connecting M2M/IoT applications. This concept is illustrated below, in Figure 1.

Further work remains before the standard can be finalized, most importantly a final decision on the uplink technology the standard will use. The two current proposals are SC-FDMA (Single Carrier – Frequency Division Multiple Access) and FDMA with GSMK (Gaussian Minimum Shift Keying). SC-FDMA is the uplink protocol for LTE while FDMA, with GSMK a less processing-intensive technology. The 3GPP states it will “strive for a single solution” by the time Release 13 is published, and the uplink technology should be finalized in a December 2015 meeting.

Despite the intense political maneuvering over the past year between different camps to push through various proposals as the 3GPP LPWAN standard, it now appears that the mobile industry has settled on a single LTE LPWAN proposal for 3GPP standardization. (It is important to recognize that GSM Evolution / EC-GSM is still moving forward and will be relevant to the extent that GSM networks remain active. Towards the end of this decade, and increasingly into the next decade, GSM network shutdown announcements will become increasingly common.)

The different camps have been able to compromise to put forward a proposal that reuses the LTE RAN (Radio Access Network) and EPC (Evolved Packet Core) infrastructure currently being deployed around the world, while optimizing this infrastructure to support an array of IoT use cases, longer battery life, increased robustness, greater volumes supported per base station, and lower costs per device.

Recommendations for operators

The buzz around the LPWAN market has grown intense over the past year, and some operators worry about falling behind in a potentially important and fast-moving new market. However, IHS believes that operators have the time to be patient and let a formal cellular LPWAN standard mature through the 3GPP RAN Work Item process. IHS estimates that the total installed base of LPWAN connections—across all technology variants, including some not discussed above—totaled nearly 23 million connections in 2014 and will rise to about 191 million connections by 2019. These calculations are based on both a “bottom-up” estimate of actual deployed shipments and the TAM opportunity for LPWAN (and reasonable penetration rates) across a range of market segments, including industrial automation, smart energy, and smart cities.

This forecast represents very fast, impressive growth. However, the 2019 projection of 191 million connections is still relatively small compared to the 541 million cellular M2M/IoT connections forecast for that year. While IHS anticipates continued fast growth for LPWAN beyond 2019, and the potential is certainly there for LPWAN-based M2M/IoT connections to eventually outstrip connections using current 2G/3G/4G technologies, the point is that operators can afford to wait for a full 3GPP LPWAN standard to come to the market by mid-year 2016, with full commercialization in 2017.

Alternatively, nothing is precluding an operator from trialing a current proprietary LPWAN technology. If the operator has the resources to run such a trial, a number of useful insights may be gleaned around issues such as: how best to monetize services, which applications are the best candidates for LPWAN connectivity, and how LPWAN technology actually performs in the field. Using non-3GPP standard technology, these trials would still be practical because, while there are certainly nuanced technical differences between the options, in large part these technologies are trying to solve the same challenges and offer many of the same benefits.

However, the 3GPP LPWAN standards process offers specific benefits: integration with existing operator radio access and core networks, use of operators’ expensive licensed spectrum assets, the global scale of the 3GPP ecosystem, and the use of well-established security and management mechanisms.

Nevertheless, apart from the standards process, IHS believes that a number of different LPWAN technologies will continue to exist in the market, at least over the near to medium term. There is significant space in the market for LPWAN technologies that operate in unlicensed spectrum and enable both private and public network services that can be deployed by non-traditional actors, in addition to operators. Ultimately, the IoT will comprise a rich fabric of different connectivity technologies, targeting different use cases and operating using a range of economic models.

Share facebook Twitter Google Plus Linked In Add This Contact Us