How to correct range migration in presence of doppler ambiguity?

In radar signal processing, range migration becomes a challenge when a target moves too fast or the coherent processing interval (CPI) is insufficient. The keystone algorithm can correct range migration effectively as long as the target's velocity remains within the radar's maximum unambiguous speed, i.e., there's no Doppler ambiguity.

However, when the target's velocity exceeds this limit, Doppler ambiguity occurs, and the keystone algorithm alone is no longer sufficient.

My question is: How can range migration be corrected in such cases where Doppler ambiguity is present due to high target velocity? Are there alternative algorithms or processing techniques that can handle this scenario?

Answer 1

[Satyam]

Hi @Muhammad,

There are several methods existing for resolving Doppler ambiguity:

Multiple PRF (Pulse Repetition Frequency) Processing

Transmit and process the same scene with different PRFs. Each PRF gives a different ambiguous velocity. The true velocity can be reconstructed using the Chinese Remainder Theorem (CRT) or similar algorithms.

Process:

• Collect data at two or more PRFs.
• Measure the ambiguous Doppler at each PRF.
• Use CRT to solve for the true Doppler frequency (and thus true velocity).

Time-Frequency Analysis / Joint Estimation

Use algorithms that jointly estimate the ambiguity number and the Doppler frequency, such as:

• Maximum Likelihood Estimation (MLE)
• Sparse Recovery / Compressive Sensing (if the scene is sparse in velocity)
• Ambiguity function analysis: Search for the strongest peak across ambiguity numbers.

Range-Doppler Migration Correction with Ambiguity Search

For each possible ambiguity number (fold), apply the keystone correction, and check which result yields a focused target signature (e.g., maximum energy or entropy minimization).

[Muhammad]

I really appreciate your detailed response. However, I believe the approach you mentioned involves a search algorithm that explores multiple ambiguity numbers to find the best fit, which could be computationally intensive. My concern is — what if, within a single dwell, there are two targets with different ambiguity numbers? In that case, how can we accurately identify and separate the correct ambiguity number for each target? and then apply keystone or phase correction algorithm

[Satyam]

I agree that a global ambiguity search can be expensive and problematic when multiple targets have different ambiguity numbers. In practice, you’d first separate the targets in the Range Doppler domain (e.g., via clustering or detection) and then estimate and apply the ambiguity correction individually for each target before performing keystone or phase correction. This per-target approach avoids excessive global searching and ensures accurate correction.