Each tower's coverage is computed with the log-distance path-loss model used to plan real cellular networks: signal falls off with the n-th power of distance, where n≈2 is unobstructed free space and n≈3–4 is a cluttered urban environment with buildings in the way. The device's received power from every tower is calculated independently from this formula — nothing is faked or interpolated from a texture. The device always tries to camp on whichever tower currently gives it the strongest real received signal, but switches (a "handover") only once a competitor beats the current serving tower by more than the hysteresis margin, exactly as real networks avoid rapid ping-pong handovers near a cell boundary.
PL(d) = PL0 + 10·n·log10(d/d0) [dB]
RSRP_i = TxPower_i − PL(d_i) [dBm]
interference = Σ (power of every other tower, linear mW)
SINR = signal_mW / (interference_mW + noise_mW)
throughput ≈ BW · log2(1 + SINR) [Shannon capacity, Mbps]
- Path-loss exponent n — how fast signal decays with distance; raise it to simulate a dense urban canyon, lower it for open ground.
- Tower Tx power — transmit power of every tower in dBm; higher power extends coverage but also raises interference for neighbours sharing the channel.
- Handover hysteresis — the dB margin a neighbouring tower must beat the serving tower by before the device switches; 0 dB ping-pongs at every crossing, higher values trade responsiveness for stability.
- SINR — signal-to-interference-plus-noise ratio, computed from every other tower's real signal contribution at the device's real position — the same quantity that determines real 5G data rates via Shannon capacity.
Real-world relevance: this is the same reasoning a radio-planning engineer runs (in far more detail) when siting cell towers — path loss sets the coverage radius, and same-channel interference from neighbouring cells, not distance alone, is usually what limits real-world 5G throughput.