NB-IoT vs Cat-M1 vs 4G
for Tracking

NB-IoT occupies a 180 kHz channel, the width of a single LTE resource block, and that narrow channel is the source of all its strengths and limits. Concentrating transmit energy into a narrow band buys link budget. The 3GPP target is a 164 dB maximum coupling loss, roughly 20 dB better than legacy GSM, which translates into coverage that reaches into basements, deep inside buildings, and underground utility pits where other radios fail. The cost of that link budget is throughput. Real-world NB-IoT gives you on the order of tens of kilobits per second downlink and uplink, and at the edge of coverage a single small message can take several seconds to deliver as the modem repeats transmissions to claw back the link.

The power story is what makes NB-IoT a multi-year-battery technology, and it rests on two features. Power Saving Mode (PSM, 3GPP Release 13) lets the device tell the network it is going dormant for a defined period, during which the modem draws single-digit microamps while staying registered, so it does not pay the heavy energy cost of a fresh attach when it wakes. Extended Discontinuous Reception (eDRX) stretches the paging cycle so the device only listens for downlink at long, agreed intervals instead of every 1.28 seconds. Together they mean a tracker can report once or twice a day and spend the other 23-plus hours in deep dormancy. The catch is latency. A device in PSM is unreachable from the network until its own timer fires, so NB-IoT is not for anything you need to command on demand.

NB-IoT also has no soft handover. The device camps on a cell, and if it physically moves to a new cell it must reselect, which involves re-synchronizing and often re-attaching. For a sensor bolted to a wall or a seal on a parked container that is fine. For anything driving down a highway it is a non-starter, because the asset will spend its energy budget re-attaching instead of reporting. That is why NB-IoT suits fixed and slow-moving assets: utility meters, environmental sensors, parked-container seals, and the kind of low-duty trackers covered under NB-IoT asset tracker development.

When the product needs continuous live tracking, large data, or video, the LPWA world is left behind in favor of full LTE. Cat-1 gives around 10 Mbps down and 5 Mbps up and is the workhorse for mains-powered or vehicle-powered trackers that report every few seconds. Cat-1bis is the same data class but reduced to a single receive antenna, which simplifies the RF layout and shrinks the bill of materials, and it has become the default for compact vehicle trackers that need real-time behavior without the antenna burden of a two-antenna design. Cat-4 pushes to 150 Mbps down and is the tier for a device that carries an AI dashcam or uploads video clips.

The reason 4G is not a battery technology is power. These radios do not implement the aggressive PSM and eDRX dormancy of the LPWA classes in any practical low-current way, and they keep a heavier protocol stack active. Average current is measured in tens of milliamps even when idle, and transmit peaks reach hundreds of milliamps. On a coin cell or a small lithium primary that is a death sentence within days. So Cat-1 and Cat-4 belong on vehicles with ignition power, on devices with a generous rechargeable pack and solar, or anywhere a wall outlet is available. For connected-vehicle and OEM telematics programs that ride on vehicle power, this is the right tier, and it is the foundation of much telematics and GPS tracking platform work.

There is a real-time dimension beyond raw throughput. Full LTE keeps the device in connected mode or short-cycle DRX, so round-trip latency is tens of milliseconds and the cloud can command the device instantly. That responsiveness is what makes 4G the right choice for high-value cargo that needs immediate alerts and on-demand position queries, the pattern behind container and cargo tracking solutions.

For two decades the default tracker radio was 2G GSM/GPRS, because it was cheap, ubiquitous, and good enough for an SMS or a tiny TCP payload. That era is ending. Operators across Europe, North America, and Asia are refarming 2G spectrum for LTE and 5G, and many networks have already switched it off. A tracker built on a 2G modem today is a tracker with a built-in expiry date, and fleets that standardized on 2G are now facing forced hardware swaps across thousands of devices.

The migration target is exactly the LPWA pair this guide covers. NB-IoT and LTE-M were specified by 3GPP as the long-life replacements for 2G machine traffic, with better coverage and far better power behavior. The practical guidance is to migrate moving and on-demand devices to LTE-M, migrate fixed low-duty devices to NB-IoT, and use full LTE only where the application genuinely needs the bandwidth. Designing the migration so a single new tracker SKU covers most of an existing mixed fleet is where most of the cost saving lives, and it is a core part of the approach to cellular connectivity design.

India is on its own timeline. 2G coverage remains broad here and operators have signaled they will keep it running for basic voice longer than Western markets, but new machine-to-machine deployments should still not be built on it, because the long-term direction is clear and a five-to-ten-year tracker will outlive the 2G plan. Building new on LTE-M or NB-IoT now avoids a forced redesign mid-life.

The decision collapses to three questions asked in order. First, does the asset move while reporting? If it is fixed or only relocated occasionally, NB-IoT is on the table and usually wins for battery life. If it moves while tracking, NB-IoT is out because of reselection cost, and you choose between LTE-M and full LTE. Second, does the product need live, on-demand, or high-bandwidth behavior? If yes, you need full LTE. If it can report on a schedule and tolerate sleep, LTE-M wins and keeps your multi-year battery. Third, what power source is available? Battery-only pushes you toward NB-IoT or LTE-M; vehicle or mains power frees you to use Cat-1 or Cat-4.

Worked through real products, that framework lands as follows. A parked-container seal or a utility-pit sensor that wakes twice a day goes NB-IoT. A two-wheeler tracker, a livestock collar, or a cargo tracker on a small rechargeable pack goes LTE-M, because it moves but still needs to sleep. A commercial-vehicle telematics unit on ignition power, or an AI dashcam uploading clips, goes Cat-1bis or Cat-4 because power is not the constraint and real-time matters. High-value, high-theft cargo that needs instant alerts and on-demand position usually justifies full LTE even on battery, paired with aggressive duty cycling. When the call is genuinely close, the resolution is to prototype on a dual-mode BG95 and measure, because a power profiler settles arguments that spec sheets cannot. That same evidence-led approach carries into every asset tracking engagement and into the broader IoT product development process.