Microchip Technology is a part of the LoRa Alliance, and industry organization formed in January 2015 to promote semiconductor vendor Semtech’s LoRa technology and evolve the LoRaWAN protocol as an open industry standard. Other Alliance members to date include IBM and Cisco as well as operators Belgacom, Bouygues Telecom, Fastnet, KPN, Singtel and Swisscom. KPN has announced that it will deploy LoRa infrastructure in 2015.
LPWAN technology for M2M and IoT applications is seeing increasing hype due to its key purported advantages compared to current cellular technologies, namely: lower costs, longer battery life, longer range, greater node density and (in contrast to low power local area networks) a simplified network topology.
The LoRa Alliance is only one of several efforts to promote LPWAN technology for M2M and IoT applications. SIGFOX just announced a $115 million funding round in February 2015 and also has the support of a number of operators, including: NTT DoCoMo, SK Telecom, Tele2 and Telefonica. Likewise, the Weightless SIG adopted NWave’s LPWAN technology in September 2014 as the basis of its forthcoming Weightless-N open standard in the Industrial, Scientific and Medical (ISM) unlicensed frequency bands. Finally, there is a longer-term ETSI effort, called “Low Throughput Networks,” to define a LPWAN standard.
However, LPWAN does face a number of challenges in gaining market acceptance; these challenges ultimately dissolved a previous effort in the mid-2000s to standardize the Wavenis LPWAN technology as “Wave2M” under the auspices of the now-defunct Wave2M Community. These challenges include:
- Module costs are still comparable to current cellular GPRS pricing. As the Microchip Technology announcement states, their LoRa module will sell for $10.90 in volume, which is comparable to current GPRS pricing. IHS believes pricing for SIGFOX modules is also in this range. While pricing could decline significantly (down to perhaps $2 per unit at high volumes), it will take some time for this to happen, during which LWPAN will see little cost benefit relative to the much more widely deployed cellular technologies.
- Unlicensed spectrum is problematic for long range, public network access. One factor that killed the muni-Wi-Fi movement in the 2000s was the use of unlicensed spectrum as a means of provisioning long range, public internet access. Generally, unlicensed spectrum works best in local and personal area networks where the risk of interference is minimized and most devices connecting to the network are under the direct control of the network owner (e.g. a wireless industrial automation system used in a plant environment). Operators are better able to provide quality of service at long range because they use their very expensive licensed spectrum.
- The market fragmentation resulting from four major industry efforts to develop LPWAN technology will lead to fear, uncertainty and doubt among potential adopters. It’s risky and expensive to design in a technology that not only won’t be as nearly as widely supported on an infrastructure basis as cellular for a number of years, at least, but also may not even survive the coming shakeout as the market determines which LPWAN technology will be the “winner”.
- The mainstream cellular industry is already developing M2M/IoT-focused technologies in the form of LTE CAT-0 and LTE-M. LTE CAT-0 will reduce module costs by roughly 50% compared to current base LTE module pricing and will see commercial deployment in the 2015 – 2016 timeframe. More importantly, the LTE-M specification should reduce module pricing down to nearly current GPRS levels (~$10), improve batter life to 5 – 10 years, improve coverage, and vastly increase the number of devices able to connect to a single base station. LTE-M should be commercially available in the 2017 – 2018 timeframe. Altair Semiconductor announced its LTE CAT-1 and LTE CAT-0 chipsets in February 2015.
- M2M applications tend to grow more complex, and therefore need higher speeds, over time. The general trend in the market is that applications are initially deployed with relatively minimal functionality, such as automotive safety and security applications. However, over time, as the developer gains more experience and invests in further development, the functionality of the application becomes more complex and requires more bandwidth, as in the case of automotive infotainment applications. While this is a very general trend, not applicable to all situations, it does mean that LTE CAT-0 and LTE-M, with download speed of around 1 Mbps, will outstrip current LPWAN technologies, which range from less than a Kbps to 300 Kbps download speeds.
- Operators generally prefer low-risk and simplicity and don’t want to manage multiple networks. Although the current operator support for LoRa and SIGFOX is encouraging (and necessary, in IHS’s view), the long-term prospects for widespread operator support are still problematic. Supporting a LPWAN technology will mean that an operator has to place a risky bet on which LPWAN variant will win in the market, and manage the LPWAN network alongside its mainstream infrastructure. As LTE CAT-0 and LTE-M become widely available, IHS things many operators will opt to use these technologies to support M2M and IoT applications.
IHS believes that LPWAN can be an important addition to the M2M and IoT ecosystem, especially if it becomes formally standardized by ETSI and 3GPP. However, for the reasons cited above, the window of opportunity is short for some LPWAN variant to displace LTE as the mainstream cellular/wide area connectivity technology for M2M. If a majority of operators, and especially leading operators like AT&T, China Mobile, and Vodafone don’t adopt LPWAN by the end of 2017, the market opportunity will likely simply pass to LTE-M.