Această secțiune este disponibilă doar în limba engleză.
The start of this chapter listed the four steps of ESM: intercept, measure, identify, respond. Behind those words is a chain that processes thousands of pulses every second. Whoever understands how that chain works also understands why an RWR sometimes warns late, wrongly or not at all, and what an LPI radar or a new waveform does about it.
An ESM receiver hears a cacophony of radars all at once. For every pulse it writes a short card: when, which frequency, how long and from which direction. It sorts those cards per emitter and then looks up in its library who that is.
In practiceOver a front line an aircraft hears dozens of radars at once, friendly and hostile. Sorting takes time and can fail if two radars look alike, or if a radar does something that is not in the library. Then an unknown symbol appears, or nothing at all.
Receivers: listening wide or listening sharp
No receiver covers the whole band at once, with high sensitivity and fine frequency resolution. Every ESM architecture is therefore a compromise between probability of intercept (POI) and sensitivity:
| Receiver | How | Strong | Weak |
|---|---|---|---|
| Crystal video | detects the whole band at once, without tuning | simple; almost always catches strong pulses | insensitive; measures frequency poorly or not at all (classic older RWR) |
| IFM (instantaneous frequency measurement) | measures the frequency of a single pulse directly, by phase comparison | wideband and a frequency per pulse | loses track when two pulses arrive at the same time |
| Superheterodyne (tunable) | sweeps a narrow, sensitive receiver across the band | sensitive and accurate | low POI: only hears what transmits on the frequency it is tuned to at that moment |
| Channelised / digital | splits the band into many parallel channels, nowadays digitally after fast sampling | wide and sensitive; separates simultaneous signals | complex, expensive and computationally heavy; standard in modern systems |
Modern RWRs and ESM suites usually combine a wideband digital channel (high POI) with a more sensitive, tunable channel that is set on interesting signals. An LPI radar (Frecvenţa modulată continuă Radarul valurilor) aims precisely at that weak spot: its power per frequency is so low that the wideband channel does not see it, and its bandwidth so large that the narrow channel hears only a small part of it.
Pulse descriptions and de-interleaving
For every detected pulse the receiver creates a pulse descriptor word (PDW): a short digital record with the time of arrival (TOA), the frequency (RF), the pulse width (PW), the amplitude and the angle of arrival (AOA). Modern systems add features of the modulation within the pulse. Over a busy area a system processes hundreds of thousands to millions of them per second.
Those pulses arrive mixed together, from all radars at once: an interleaved pulse train. The processor first has to sort them per emitter, the de-interleaving:
- Coarse sorting on features that hardly change per emitter: angle of arrival (a radar does not jump bearing within a fraction of a second), frequency band and pulse width.
- Fine sorting on timing. Within each group the processor looks for a repeating pattern in the arrival times. A histogram of the time differences between pulses shows peaks at the PRI of each emitter; pulses that fall on such a fixed rhythm belong together.
- Leftover pulses that fit nowhere are kept. They may be a new emitter, a radar that has just started transmitting, or noise.
This is also why PRI jitter and stagger (Modularea PRI) work not only against jammers but also against ESM: a radar without a fixed rhythm is harder to sort. And this is where the "ghosts" on an RWR display come from: if two radars work on the same bearing and frequency, the processor may merge them, or split one radar in two.
Identification: the library, and where it comes from
A sorted emitter has a set of measured parameters: frequency (range), PRI and its pattern, pulse width, scan period and type (Caracteristicile antenei), and sometimes modulation within the pulse. That fingerprint is compared with the emitter library, in an RWR the mission data file. The result is rarely a hard yes or no: the processor gives the best match with a confidence, and sometimes several candidates. Two radars with overlapping parameters are hard to tell apart. A good library therefore also contains the modes of each radar and the combinations that belong to one weapon system: a search radar and a fire control radar on the same bearing say more together than each one alone.
Where do those parameters come from? From ELINT (COMINT vs ELINT): the patient collection and technical analysis of radar signals by specialised platforms and stations, often already in peacetime. ESM and ELINT use similar receivers, but with a different purpose:
| ESM | ELINT | |
|---|---|---|
| Purpose | respond immediately: warn, evade, counter | build knowledge: how exactly does this radar work? |
| Time | fractions of a second | hours to years of analysis |
| User | crew and weapon system | analysts and library managers |
| Product | warning, cue, track | parameters, modes, the library itself |
Together they form a loop: ELINT fills the library, ESM uses it, and whatever ESM cannot place in practice goes back to the analysts. A radar that switches on a never-used mode in wartime (a war reserve mode, Introducere) slips straight through that loop, until the library has been updated.
Direction finding and location
An RWR finds direction coarsely. It compares the signal strength between four antennas in the quadrants (amplitude comparison) and achieves an accuracy of a few to more than ten degrees. Enough to know from which angle the threat comes and to aim countermeasures, but too coarse to locate a site.
More accurate systems measure the phase difference between closely spaced antennas (interferometry) and achieve fractions of a degree. That makes location possible:
- Cross-bearing (Triangulare): two or more bearings from different positions intersect at the emitter. A single aircraft can do this alone, by taking several bearings along its route.
- TDOA and FDOA (Intercepție, DF și geolocație (TDUA/FDOA)): several receivers measure the difference in time of arrival and in Doppler shift of the same signal. Together they yield a position, without the need for accurate direction-finding antennas.
A located, identified emitter is a target. It can be avoided, jammed or, in SEAD (SEAD & MOARTE), attacked. That is why radars switch off under threat, relocate after transmitting and use decoy emitters that mimic their signal.