Definition
LTE-M (LTE for Machine-Type Communications, also known as eMTC) is a 3GPP R13 defined LTE-based cellular IoT technology, operating in 1.4MHz narrowband, coexisting with LTE broadband on the same carrier. LTE-M supports handover, roaming, VoLTE and, like NB-IoT, is a mainstream LPWAN technology. LTE-M is suitable for applications requiring mobility, medium data rates (peak about 1Mbps), and voice capabilities, differentiating it from NB-IoT (static, ultra-low power) and Cat.1 (medium-high rate).
Technical Features
- 1.4MHz narrowband: Coexists with LTE, flexible deployment, can reuse existing sites
- Mobility: Supports handover, reselection, roaming, suitable for vehicles and wearables
- VoLTE: Supports voice, suitable for wearables and emergency calls
- Power saving: PSM, eDRX extend battery life
Coverage and capacity
LTE-M supports CE mode (Coverage Enhancement) for improved coverage, with repeated transmission enhancing link budget. Compared to NB-IoT, LTE-M has slightly weaker coverage but better mobility and speed. Both can be deployed together, with operators choosing according to scenarios.
Typical Use Cases
Wearable devices, vehicle tracking, mobile payments, smart meters (mobile), industrial IoTmobile terminals, etc. SimtoLink provides LTE-M IoT SIM cards, forming a complete product matrix with NB-IoT and Cat.1.
LTE-M is an important technology choice for SimtoLink in low-power IoT scenarios. For applications requiring mobility and voice capability such as wearables, vehicle tracking, and mobile payments, SimtoLink provides LTE-M-based IoT SIM cards and connectivity solutions. The SimtoLink cloud platform supports lifecycle management of LTE-M devices, forming a complete cellular IoT product matrix with NB-IoT and Cat.1.
1.4 MHz Still Uses LTE Mobility
In the specifications, LTE-M is called eMTC. It narrows the device radio to 1.4 MHz so a cheaper crystal and power amplifier can be used, but cell search, random access, security mode, and bearer setup still follow the LTE procedure. When it shares a carrier with ordinary LTE devices, the scheduler sets aside narrowband resource blocks for it, and wideband users keep the rest. “Narrowband” here means device capability and scheduling granularity. It is not a separate IoT private network. Coverage enhancement buys link budget by repeating transmissions. Once the repetition count rises, the airtime of one small packet grows noticeably.
Compared with NB-IoT, it keeps the measurements needed for connected-mode handover and cell reselection, and it keeps the bearer types VoLTE needs. Peak rate stays on the order of about one megabit per second: enough for location, status, and a short stretch of voice, not enough for video. Writing the peak from the module datasheet into a service commitment falls apart at the cell edge. The edge is often on repetition and low-order modulation.
It Pays When the Device Moves and Sometimes Sends a Medium Payload
The conditions are specific. The device moves across cells, or it needs voice, SMS, or more timely downlink, and it should use less power than Cat.1. Shared-bike locks, pet or personal trackers, and mobile payment terminals are in this group. A meter, a manhole sensor, or a smoke detector that barely moves and reports a few dozen bytes a day does not use the extra handover capability. Narrower NB-IoT, or even a non-cellular link, is often a better fit. The operator must enable LTE-M on the target band, and the roaming network needs the same capability. A subscription only on the home core, with no 1.4 MHz scheduling on the visited network, makes the device fall back to wider LTE or lose service entirely.
Power saving still depends on PSM and eDRX, but the listen cycle has to be read together with power control and paging delay. If an unlock command must arrive within a few seconds, eDRX cannot be stretched to tens of minutes. A voice call also requires an IMS and QoS subscription. A module that supports VoLTE does not mean this SIM’s APN has IMS open. When the device moves between areas, a reselection priority in the mobility policy that pushes it onto a carrier with no LTE-M produces “there is signal, but attach fails.”
Copying Power-Saving Parameters onto Narrowband IoT
A common mistake is to drop NB-IoT PSM timers straight into an LTE-M project. Coverage level, paging cycle, and the TAU timer do not share the same ranges. An AT command written into the module may be rewritten by the network, or rejected. Another mistake is an antenna still designed for a Cat.4 wideband device. Efficiency falls when 1.4 MHz sits at the band edge: the rate looks normal nearby, and the link drops repeatedly in a garage. Turning handover off “to save power” means that once the device enters a basement or an interchange it can only reselect, and downlink commands are lost for the whole reselection.
Frequency support is also easy to judge from the module and not from the SIM. The module may claim many bands. If the SIM subscription or the roaming agreement opens only some of them, the device locks onto a crowded carrier and repetition mode eats cell capacity. Before go-live, record the attach carrier, coverage level, handover count, and the wake delay of one downlink along the real route. Do not stop at peak throughput in the lab.
Once coverage-enhancement repetitions rise, even a very small status packet holds the narrowband resource for a long time, and the number of devices the cell can serve at once falls. If mobile payments or location reports are dense, first reduce the share of points that must be locked at the highest repetition level. Do not lock every point there. A voice call also takes guaranteed-bit-rate resources and competes with pure data devices on the same carrier. If the SIM subscription has not opened that bearer, the voice switch in the module menu means nothing. A route test should at least keep the carrier, the coverage level, whether a handover happened, and the time from sending one downlink to the device’s response. A lab test that uses only a wideband antenna overestimates narrowband efficiency at the band edge. Garage and in-vehicle results will be worse. If the voice bearer is not open in the subscription, the voice switch on the module will not place a call. The failure is a core-network reject, not the coverage level.