In LTE a cell was, near enough, a transmitter. In NR a cell is a synchronisation signal block and a physical cell identity — and several transmission points can broadcast the same one. That single change is why a device can move across a building without generating the handover signalling it once would have.
A short known sequence the device correlates against blindly, with no prior knowledge of timing or frequency. Finding it establishes symbol timing and part of the physical cell identity.
Transmitted alongside, it completes the physical cell identity and resolves frame timing. Together the two identify which cell has been found and where its boundaries fall.
Carried in the same block, it holds the master information block — a very small payload whose main job is to describe where to look for the next piece.
Found using what the master block described. This carries what a device actually needs to attempt access: how to random access, what is permitted, and how to find anything further.
Remaining system information is delivered on demand rather than broadcast continuously, because a message repeated forever costs energy and creates interference whether or not anyone needs it.
An earlier generation broadcast a good deal of system information continuously, on the reasonable assumption that a device arriving at any moment should find what it needs immediately.
The cost of that assumption is permanent. Every always-on transmission consumes power at the base station, occupies resource that could carry traffic, and radiates interference into neighbouring cells — whether one device is listening or none.
NR inverts the default. Broadcast the minimum needed to be found and to ask a question; deliver everything else in response to the question. The result is a carrier that is quieter when idle, which matters more as networks densify and cells increasingly serve small numbers of devices.
At higher frequencies a usable link needs antenna gain, and gain means narrow beams. A synchronisation block radiated in one direction would only ever be found by devices in that direction.
So the block is transmitted repeatedly in different directions, each transmission carrying an index. A device finds whichever it hears best and reports that index, which tells the network which direction to use for it.
Beam management then runs continuously beneath the connection. As the device moves or the channel changes, it measures alternatives and the network switches beams. This is physical and medium-access layer work — fast, frequent and cheap, and invisible to the layers above.
Multiple transmission points can broadcast the same synchronisation block and the same physical cell identity. To the device they are one cell, so moving between them is a beam change rather than a handover.
| Beam switch | Handover | |
|---|---|---|
| Happens at | Physical and medium-access layers | Radio resource control — layer 3 |
| Signalling cost | Measurement reports and a switch | Full procedure with core network involvement |
| Time scale | Milliseconds | Tens of milliseconds and upward |
| Interruption | Effectively none | Brief, and occasionally visible |
| Triggered by | Movement within one cell identity | Crossing into a different cell identity |
Which is why deployments increasingly place several transmission points under one identity: the same physical coverage, with mobility handled at the cheap layer instead of the expensive one.
The primary and secondary synchronisation signals. The primary is a short known sequence a device correlates against blindly to establish symbol timing and part of the physical cell identity; the secondary completes that identity and resolves frame timing. Together they let a device find and identify a cell with no prior knowledge.
Because always-on transmission is a permanent cost — power at the base station, occupied resource, and interference into neighbouring cells, whether or not anyone is listening. NR broadcasts the minimum needed to be found and to ask a question, then delivers the rest on request.
The synchronisation signal block carries the synchronisation signals and the broadcast channel together. At higher frequencies a usable link needs narrow, high-gain beams, so the block is transmitted repeatedly in different directions with an index per transmission. A device reports which it heard best, telling the network which direction to use.
Yes. Points broadcasting the same synchronisation block and physical cell identity appear to a device as one cell. Moving between them is then a beam switch rather than a handover, which is handled at the physical and medium-access layers instead of by a full radio resource control procedure.
Beam switching happens at layers 1 and 2, takes milliseconds and causes effectively no interruption. Handover is a layer 3 procedure involving the core network, takes considerably longer, and can be briefly visible. That difference is why deployments group transmission points under one identity.
They share a scheduler and a resource grid, so aggregate capacity is capped compared with independent cells. A fault at one point also presents as coverage degradation rather than a clean outage, which is harder to localise operationally.
Written for engineers. Share it with one.
The layers and the RU/DU/CU split.
TECHNOLOGYWhere synchronisation is implemented.
KNOWLEDGEConnecting the transmission points.
KNOWLEDGECell selection in non-standalone.
PRODUCTThe transmission points.
TECHNOLOGYWhy beams exist at all.
Tell us the specification, the constraint and the deadline. Programmes that cross silicon, radio, embedded and AI are where Faststream is strongest.