In TECER’s Phase 1, a TVWS kit designed for links of 10 to 15 km did not go beyond 400 meters in the first field test. This page explains where the range of a radio link comes from, which factors reduce it and which of them were confirmed, or not, in the project’s tests.
How is range calculated?
A radio link works like a calculation of additions and subtractions, called the link budget. The signal leaves the transmitting radio with a certain power. Along the way, it gains and loses strength:
- It loses strength in the cables and connectors between the radio and the antenna.
- It gains strength in the transmitting antenna, which concentrates the signal in one direction.
- It loses strength on the way through the air. The loss grows with distance and also when passing through vegetation, terrain and buildings.
- It gains strength in the receiving antenna.
- It loses strength again in the cables and connectors on the receiving side.
What remains needs to stay above the receiver’s sensitivity (the minimum signal level it can understand) and above the noise and interference in the region. Gains and losses are measured in decibels (dB). Every 3 dB of loss corresponds to about half the power; every 10 dB, to one tenth.
For this reason, losses that seem small at each stage can, when added together, greatly reduce the range. Simulation programs, such as Radio Mobile, do this calculation using terrain data, but they assume that cables, connectors and antennas are in good condition.
What factors reduce the range?
After the test in Mosqueiro, the technical team listed six hypotheses. The table shows what each one affects in the link budget and what the following tests revealed.
| Factor | What it affects | Status at the end of Phase 1 |
|---|---|---|
| Loss in the coaxial cables | Stages 1 and 5 | Confirmed in the UFPA/LASSE laboratory |
| Oxidation on the RF connections | Stages 1 and 5 | Found when a base station was opened; it was not among the initial hypotheses |
| Antenna height and orientation | Stages 2, 3 and 4 | Influence observed in the field; changing the orientation changed the range |
| Interference from other signals | Noise at the receiver | Occupied channels identified in the region; effect not confirmed |
| Spectrum saturation | Noise at the receiver | Not confirmed by the instruments used |
| Chip incompatibility or firmware error | Operation of the radio itself | No answer; it would require tests the team had no way of carrying out |
Cables and connectors
A coaxial cable loses part of the signal with every meter. This loss grows with frequency and with poorer material quality. Cables made in the field, with poorly fitted connectors, lose more. In the laboratory, the cables made in Mosqueiro lost much more signal than those sent by the manufacturer. The measurement procedure is in Measure the loss of a coaxial cable with the NanoVNA.
Oxidation has a similar effect. An oxidized connection creates resistance and disturbs the passage of the signal: part of it turns into heat, part of it goes back toward the radio. See Inspect and restore oxidized RF connections.
Antennas
Sector antennas concentrate the signal within a range of directions. Outside that range, the gain drops sharply. In Mosqueiro, signal quality at 50 meters was 80% with the antennas facing each other, 60% with the receiving antenna turned sideways and 35% with it turned backward. When the antennas were turned toward the interior of the island, the range went from 200 to 400 meters.
Height also matters: the higher the antenna, the fewer obstacles there are in the signal’s path, even on links without LOS (line of sight).
A signal that is too strong
The opposite problem also happens. On very short links, the signal can arrive too strong and saturate the receiver. In Tururi, a 450-meter link only became stable after an attenuator was installed, a part that reduces the strength of the signal by a known amount.
What remained open?
Fixing cables and connections improved the results, but did not solve every case. In the UFPA tests, two of four points at the Guamá Science and Technology Park remained unstable. In Tururi, all the points served by one of the base station’s antennas failed. The hypotheses of interference, component incompatibility and firmware remain unconfirmed.
The lesson about method is the same in every case: measure one part of the path at a time, before attributing the problem to the technology as a whole.
Sources
- Project facts: Phase 1 logbook, Instituto Nupef, 2024 (internal document).