1. The FCC Greenlighting CBRS for Private LTE
The FCC recently greenlit
the opening of the Citizens Broadband Radio Service (CBRS) 3.5 GHz band for commercial use. Previously, the 3.5 GHz band had been the domain of DoD members, namely the U.S. Navy’s radar systems. The FCC’s move is opening exciting avenues
for the development of private LTE networks and more through the FCC’s new three-tiered spectrum sharing license model. While the DoD and other incumbent users will retain priority access to the 3.5 GHz band, commercial network operators can now request Private Access Licenses (PALs) or the lower tiered General Authorized Access (GAA) to use the band in 10 MHz increments for ten-year terms. This is a major win for wireless carriers, who had been pushing the FCC to greenlight the CBRS band since the rules were set in 2015, in order to create interoperable 5G IoT networks. While some smaller wireless and industrial IoT providers pushed back on the move toward counties over census tract zones, wireless carriers are happy. And the word on the street is that the open CBRS band could fill a missing spectrum gap for IoT solutions providers working on use cases like indoor asset tracking
. On November 6th, CBRS Alliance announced
that the FCC approved Sierra Wireless’ new AirPrime class of CBRS-ready embedded modules. The best may be yet to come for CBRS-enabled IoT, but I’m glad to see that the FCC is working closely with carriers and providers to empower the next wave of connectivity solutions!
2. The Rise of the Edge
While the opening of the CBRS band solves numerous issues for applications requiring high-bandwidth and low-latency, we know that prefiltering and streamlining data before using up bandwidth is often smart. Edge computing is helping IoT solutions providers reduce bandwidth needs while increasing real-time decision making capability by pushing ML modeling to the edges of IoT networks. Fifty years ago, computers filled a room. Thirty years ago, PCs became all the rage. Then cloud computing brought us back to centralized, elastic computing in data centers. Now, we’re moving back to the edge
because mobile devices can handle the load and also because providers are listening to customers’ needs to run AI models trained in the cloud on edge devices so that those devices can efficiently make snap decisions in mission-critical situations. Before the rise of the edge, those devices would have to use valuable power and time to transfer data back to the cloud, process it through a few models, and transmit conclusions back to the edge device so that it can act upon a given scenario. Take self-driving cars for instance. AVs need to be able to run numerous operational and mission-critical safety processes without relying fully on data centers. As Eric Simone argues
, “Edge computing wins out over cloud processing when time-sensitive events are happening.” The cloud giants have heard the call. While Google is a bit later to the game
than Amazon’s AWS Greengrass and Microsoft’s Azure IoT Edge, Google’s ability to draw on talent both internally and through its TensorFlow and Kaggle open source communities may allow it to gobble up AI stack market share
from the Amazon and Microsoft. Whoever ends up on top of the AI stack, IoT networks will reap the benefits.
3. The Rollout of LTE-M and NB-IoT
Narrowband-IoT (NB-IoT) and LTE Category-M1 (LTE-M) rollouts are ramping up and serving new use cases. For the better part of a decade, cellular IoT applications
were more or less focused on use cases with plenty of available power, fat budgets for carrier license fees, and specific needs for high-bandwidth and low-latency connectivity. With 5G on the horizon, the big telcos and 3GPP want to position cellular IoT as an all-encompassing IoT connectivity solution. In 3GPP and the wireless carriers’ vision, 5G will serve the high-bandwidth, low-latency requirements; CBRS will facilitate private LTE networks; last but certainly not least, NB-IoT and LTE-M will serve the majority of other IoT needs, specifically LPWANs—at least while they’re deploying and refining 5G through the early 2020s so that it can potentially surpass its forerunner protocols. LTE-M’s big advantage for LPWAN networks
stems from capping the maximum system bandwidth at 1.4 MHz, honing in on that classic LPWAN trope of transmitting small data packets at low transfer rates to prolong battery life. What’s more, LTE-M networks can link up with existing LTE infrastructure with a simple software patch whereas other LPWAN solutions operating in unlicensed spectrum often require building out proprietary network infrastructure like LoRaWAN gateways
and so forth. Compared to LTE-M, NB-IoT has an even lower maximum bandwidth (200 KHz compared to LTE-M’s 1.4 MHz), which makes it a great option for truly low-power LPWANs that can operate outside of the LTE band. In closing, it’s great to see the increasing pace of NB-IoT and LTE-M deployments worldwide. As you can see on this list of recent and ongoing deployments
, both NB-IoT and LTE-M are spreading across the world through a healthy ecosystem of telcos and providers. We’re beginning to see an increasingly-rich cellular IoT solutions landscape emerge. In the end, just as with edge computing, whichever players rise and fall, on the whole, IoT will still see huge gains from the competition-driven innovation.