CAT-M Evolution in 3GPP Rel. 14

It’s been a while since I talked about LTE based technologies for the Internet of Things (IoT) or Machine to Machine (M2M) communication. Slowly but steadily, we are now seeing NB-IoT and CAT-M1 technology deployed in practice even though I would have thought things would move faster. 3GPP has not slept in the meantime, however, and has continued to evolve both the standards for NB-IoT and also for CAT-M. In this post, I’d like to describe some interesting additions to the CAT-M system that has resulted in specifications for CAT-M2 devices.

In case you don’t remember, CAT-M1 devices are basically stripped down LTE devices that do not have to listen to the full 20 MHz channel of an LTE carrier but are limited to a channel bandwidth of 1.4 MHz with a maximum theoretical throughput of 1 Mbit/s in uplink and downlink. Also, they are usually operated in half-duplex mode, i.e. they can only transmit or receive at a particular time. While this requires a whole new set of signaling channels, the advantage of CAT-M1 devices is that they can be built with less complexity, and power consumption might also be lower than that of full LTE devices. If capacity of a single 1.4 MHz CAT-M1 channel is not sufficient, several CAT-M1 channels can be configured in a cell and activated and deactivated as required. The quick activation and deactivation is important, as additional 1.4 MHz CAT-M channels reduce the bandwidth that is available for LTE devices. To increase the range of CAT-M transmissions in both the uplink and the downlink direction, coverage enhancement mode A with levels 0 and 1 can be configured for re-transmitting data many times to increase redundancy. While this sounds great, it obviously reduces overall cell capacity significantly.

In Release 14, 3GPP has now specified a number of additional features. In addition to the CAT-M1 channel bandwidth of 1.4 MHz, CAT-M2 defines a channel bandwidth of 4×1.4 MHz, i.e. 5 MHz with a maximum theoretical data rate of 4 Mbit/s in uplink and downlink. Again, several CAT-M2 channels can be activated in a cell should that be necessary from a capacity point of view. But since CAT-M1/M2 has to share capacity with LTE devices, I suppose that few operators would be keen to actually take away that much capacity from broadband Internet services for that. But that’s an assumption at this point in time as there are only few CAT-M1 deployments in practice so far.

To further increase the coverage range, coverage enhancement mode B with coverage enhancement levels 3 and 4 has been introduced. These feature an even higher number of configurable repetitions, this time without power control. This makes sense at higher repetition factors as each transmission has to use the highest power level anyway to keep the number of repetitions as low as possible. It sounds nice on paper but repeating data several dozen times totally kills cell capacity.

Surely, these are interesting ideas that have gone into Release 14, so lets see if and when they will find themselves in real networks. Judging by the current state of NB-IoT and CAT-M deployments, I don’t think we’ll see this anytime soon.

P.S.: If you google for CAT M2, you will still find a lot of articles that state that CAT M2 is NB-IoT. This has been changed by 3GPP in the meantime however, with NB-IoT now being referred to as CAT-NB1 devices!