Definition
5G (Fifth Generation) is the fifth-generation mobile communication technology, supporting scenarios such as eMBB (enhanced Mobile Broadband), uRLLC (ultra-Reliable Low Latency Communications), and mMTC (massive Machine Type Communications), suitable for applications like ultra-high-definition video, low-latency flight control, and V2X.
Technical Features
5G adopts key technologies such as millimeter wave (mmWave), massive MIMO, beamforming, and network slicing. Peak rates can reach 20 Gbps, end-to-end latency can be reduced to within 1 ms, and connection density can reach millions of devices per square kilometer. Dual-band deployment with Sub-6 GHz and millimeter wave balances coverage and capacity, with NSA (Non-Standalone) and SA (Standalone) coexisting, gradually evolving towards SA.
Three Major Application Scenarios
eMBB (enhanced Mobile Broadband): Supports high-bandwidth services such as 4K/8K ultra-high-definition video, AR/VR, and cloud gaming, with typical downlink speeds exceeding 1 Gbps. uRLLC (ultra-Reliable Low Latency Communications): Targeted at industrial control, remote surgery, autonomous driving, and drone flight control, requiring millisecond-level latency and 99.999% reliability. mMTC (massive Machine-Type Communications): Supports massive IoT terminal access, with connection numbers reaching millions per square kilometer, suitable for scenarios such as smart cities and smart metering.
Applications in communication modules
5G modules are widely used in CPE, industrial gateways, in-vehicle terminals, drones, and high-definition cameras. Through 4G/5G dual-mode or 5G single-mode design, they meet different requirements for bandwidth, latency, and power consumption in various scenarios. Applications with high low-latency demands, such as vehicular networks (V2X), smart factories, and remote healthcare, rely on the highly reliable connections of 5G modules and networks.
Engineering Practice Supplement
Cross-team collaboration for "5G": Hardware provides impedance and power budget; RF provides spurious and sensitivity margin; embedded systems define thread models and storage lifespan; cloud defines SLA and rate limiting. For scenarios like NB-IoT, the business model needs to match wake-up cycles, packet length, and downlink capacity; for vehicular and mobile scenarios, session continuity during handover and roaming must be considered. At project closure, conduct a "red team" style review: check if defined behavior exists for weak networks, power loss, expired certificates, clock jumps, and full storage.
Firmware constraints of standalone and dual connectivity
A 5G module has two network paths. Non-standalone operation keeps the control plane on LTE and can put the user plane on the new radio, so the terminal must maintain two radio states. Only standalone operation registers with the 5G core alone. If firmware assumes non-standalone while the live site has opened standalone only, the module finds a cell and then fails to attach. The reverse is also true: a build that enables only standalone has no service where an LTE anchor is still required. Selection should record the operator's deployment stage, not only whether the module cover says 5G.
Dual connectivity also raises power and scheduling complexity. When the anchor cell is poor, a strong secondary cell cannot keep the link alive by itself. An IoT gateway that uploads video at full power for long periods must be checked for amplifier heat and enclosure temperature rise, or it will reduce power first and then lose rate. Low-latency services cannot treat air-interface delay as end-to-end delay; the core, firewalls, and platform queues all count. Massive-connection use should not pick a module optimized for wide bandwidth; the signaling overhead and standby current will not fit.
Choosing millimeter wave or mid and low bands
Mid and low bands reach farther and pass through walls better, which suits most outdoor and factory IoT terminals. Millimeter wave offers wide bandwidth but poor diffraction. The antenna must stay aligned, and blockage or rain fade can drop the link suddenly. A handheld or vehicle-mounted device that moves is unlikely to hold a millimeter-wave beam. Fixed cabinets, indoor distribution, and line-of-sight backhaul are the cases worth evaluating. Module support for millimeter wave does not mean the antenna, phased-array calibration, and mechanical parts are ready.
A common mistake is to treat a lab peak as the delivery figure, and to ignore behavior after fallback to 4G. At the coverage edge the device changes generation often. If the session is not kept, the upper layer sees periodic interruptions. If antenna count and isolation were designed for a single antenna, a multi-antenna rate promise cannot be met; see antenna isolation and MIMO. Export projects should freeze the bands before talking about certification. Band and certification planning should come before tooling the enclosure.
A secondary cell cannot stand in when the anchor weakens
In non-standalone operation, the secondary cell's user plane depends on the anchor. When anchor quality falls, a strong secondary cell will not keep the service up alone. The terminal loses the whole connection and then reselects. Treating secondary-cell signal strength as proof that 5G is still usable misreads coverage. Logs should keep both the anchor and secondary cell identities; otherwise the only record is an unexplained interruption.