Cellular, satellite, and WSN seem like three different worlds, but they share the same underlying physics and the same engineering pressures.
- Path loss. Friis applies everywhere. Cellular: tens to hundreds of dB. Satellite: 200 dB to GEO. WSN: tens of dB at short range. The differences in transmit power, antenna gain, and noise figure are how the engineering closes the link in each case.
- Capacity. Shannon-Hartley sets the limit: . Cellular squeezes more capacity from the same band by reusing frequencies in space. Satellite uses spot beams to do the same (and adaptive coding to fight rain). WSN trades data rate for range and battery (lower rate -> more processing gain -> longer link).
- Latency. Speed of light, distance, and protocol overhead. Terrestrial is bounded by light-time of fiber routes. GEO has the unavoidable 250 ms penalty. LEO matches fiber. WSN has its own battle: duty-cycled MACs introduce sleep latency.
- Energy. Cellular phones recharge daily. Satellites have solar panels and 15-year fuel budgets. WSN sensors must run on coin cells for years.
Understanding these tradeoffs is the foundation. Specific systems (Starlink, NR mm-wave, ZigBee Thread mesh) are particular points in the design space, chosen for the application's demands.
The hardware-security perspective adds one more thread: every wireless system is a radio, every radio is eavesdroppable by anyone within range, and every protocol is only as secure as its weakest link. GSM showed us that air-interface crypto is not enough if you don't have mutual authentication. GPS shows us that even passive one-way systems can be spoofed. WSN shows us that weak crypto on millions of cheap devices becomes the world's largest DDoS botnet.
The next chapters (network security, side-channel attacks, fault injection) will dive into the offensive side. The defense always begins with understanding how the system was designed and what assumptions it was built on. Read the standards. Look for the assumption that did not survive contact with reality.