The number of IoT connected “things” is expected to grow dramatically over the next few years, with estimates ranging as high as 50 billion connected devices by 2020. Indeed, this has caused somewhat of a stampede to try to offer broadband connectivity options to attach all of these devices to the cloud.
Current networks supporting 4G/LTE offer connectivity but at a price that most IoT things can’t afford. Its not only a connection cost issue, but also one of power requirements (standard LTE modems are relatively power hungry), and size (many LTE modem chips are quite large given the size of many of the IoT devices). And most IoT devices need relatively low speed and low amounts of data transmission requirements, which are not the design center for modern LTE networks. So alternatives to standard wireless broadband networks are needed.
Two players in this gold rush to connect all the devices coming on line are Sigfox and LoRa. Both are targeting the low power, low bandwidth requirements that most “things” need. And both are now deploying networks based on their own standards (Sigfox is a proprietary network owned by the company and LoRa is an alternative open network proposed by a consortium). But does either have a chance of success longer term? The answer, in my opinion, is no.
This is WiMax all over again
The standards body controlling the specs on broadband wireless connectivity (3GPP) was slow to react to the needs of low power and low bandwidth connections. This is not uncommon given the nature of various constituencies represented in the standards committee. Most standards bodies take time to do their work.
But recently, 3GPP approved the Narrowband LTE spec (NB-LTE), also known as NB-IoT, technically designated as Cat M1 (1 mbps data rate) and Cat NB1 (40 kbps data rate) in Release 13 of the 3GPP LTE standard.
The fundamental success factors of any network over the long term are investment to create scale, support by the major players, and supporting standards versus being proprietary. On all fronts, NB-LTE wins out.
First, current LTE networks around the world can be upgraded with minimal investment to support NB-LTE. Indeed, many carriers have already announced their intention to do so (e.g., Verizon, AT&T). This will make it nearly ubiquitous within the next 1-2 years. Proprietary networks like Sigfox and LoRa need to be built out and therefore require major investments. LTE already has a huge footprint that can be leveraged, whereas the alternatives start from a green field deployment which can take years to fully deploy.
Second, the largest vendors of modem chips (e.g., number 1 supplier Qualcomm, as well as major suppliers like Intel and others) have announced their endorsement of NB-LTE standards and have stated they will make system on a chip and standalone modem chips available to this standard within the next several months. And major carrier infrastructure vendors (e.g., Ericsson, Nokia, Huawei) have also announced support. The other networks can’t claim this level of support, and will have relatively few chip and infrastructure suppliers as they go where the standards (and volumes) lead them.
Finally, NB-LTE is a worldwide standard and will therefore be implemented the same way current smartphone based networks are, allowing cross-carrier roaming and universal access. This scale is a major advantage to service providers, commercial users of Internet of Things/Enterprise of Things, and mobile “things.” Scale is important in any connectivity scenario as users will not deploy IoT with spotty or intermittent coverage.
There is one other point that’s critical. NB-LTE has a direct pathway to the roll out of 5G networks. While 5G is still 3-4 years away, it will be critical to future uses. As a result it’s likely that NB-LTE will be around for a long time. Neither Sigfox or LoRa can claim such integration into the future networks.
The bottom line is fairly simple: NB-LTE low power ultimately wins, as internationally deployed standards win out over proprietary green field networks. The chip companies and carrier infrastructure vendors ultimately will move to standards in supplying products. The major carriers will modify their existing LTE networks which will overwhelm the proprietary ones in sheer scope and capability.
Current investments being made in Sigfox and LoRa (e.g., Vodafone) are only stop gaps to fill the void but it will be hard for them to grab enough scale to be successful longer term. Within 3-4 years, I expect all will be mothballed, or at least well on the way towards obsolescence. If there’s one thing we’ve learned from WiMax, it’s that is very hard to go against the grain when so many vested interests are at stake, even if you do have a potentially better, or at least more rapidly deployed solution.