Cat-M1, also known as enhanced Machine-Type Communication (and sometimes just called Cat-M), was designed to target IoT and M2M use cases with low cost, low power, and range enhancements. Cat-M1 was released in the 3GPP release 13 schedule. The design is an optimized version of the Cat-0 architecture. The single largest difference is that the channel bandwidth is reduced from 20 MHz to 1.4 MHz. Reducing the channel bandwidth from a hardware point of view relaxes timing constraints, power, and circuitry. Costs are also reduced by up to 33% compared to Cat-0 since the circuitry doesn't need to manage a wide 20 Mhz spectrum. The other significant change is in the transmit power, which is reduced from 23 dB to 20 dB. Reducing the transmit power by 50% reduces cost further by eliminating an external power amplifier and enables a single-chip design. Even with the reduction in transmit power, the coverage improves by +20 dB.
Cat-M1 follows other late 3GPP protocols that are IP-based. While not a MIMO architecture, throughput is capable of 375 Kbps or 1 Mbps on the uplink as well as downlink. The architecture allows for mobility and justifiably for in-vehicle or V2V communication. The bandwidth is wide enough to allow for voice communication as well, using VoLTE. Multiple devices are allowed on a Cat-M1 network using the traditional SC-FDMA algorithm. Cat-M1 also makes use of more complex features such as frequency hopping and turbo-coding.
Power is critical in IoT edge devices. The most significant power reduction feature of Cat-M1 is the transmission power change. As mentioned, the 3GPP organization reduced the transmission power from 23 dB to 20 dB. This reduction in power does not necessarily mean a reduction in range. The cell towers will rebroadcast packets six to eight times. This is to ensure reception in particularly problematic areas. Cat M1 radios can turn off reception as soon as they receive an error-free packet.
Another power saving feature is the Extended Discontinuous Receive (eDRX) mode which allows for a sleep period of 10.24 seconds between paging cycles. This reduces power considerably and enables the UE to sleep for a programmable number of hyper-frames (HF) of 10.24 seconds each. The device can enter this extended sleep mode for up to 40 minutes. This allows the radio to have an idle current as low as 15 uA.
Additional power mitigation abilities and features include:
- PSM, as introduced in Cat-0 and release 13.
- Relaxing neighbor cell measurements and reporting periods. If the IoT device is stationary or slow moving (sensor on a building, cattle in a feedlot) then the call infrastructure can be tuned to limit the control messages.
- User and control plane CIoT EPS optimization is a feature that is part of the RRC in the E-UTRAN stack. In normal LTE systems, a new RRC context must be created every time the UE wakes from IDLE mode. This consumes the majority of power when the device simply needs to send a limited amount of data. Using the EPS optimization, the context of the RRC is preserved.
- Header compression of TCP or UDP packets.
- Reduction of sync time after long sleep periods.