An emitter fusion recognition algorithm based on multi-collaborative representations

Emitter signals modulation recognition based on discriminative projection and collaborative representation

“…In order to verify comprehensive performance of DP‐CRC, the following seven recognition methods were selected for comparison: KNN [16], RBF‐SVM [19], LC‐KSVD [28], SRC [22], random projection (RP) and compressively collaborative representation (RP‐CCR) [25], DP‐CRC‐L1 with l 1 ‐norm constraints, and CNN [8]. Among them, KNN and RBF‐SVM are classic machine learning classification methods; LC‐KSVD, SRC, and CCR are traditional dictionary learning classification methods; CNN is a deep learning classification method.…”

“…In [24], the idea of collaborative representation was introduced, while a collaborative representation classification (CRC) method was proposed with the l 2 ‐norm regular constraint. In [25], it was introduced into the field of emitter signals recognition. On this basis, through the introduction of various types of optimisation discriminant items, a series of reinforced classification algorithms such as label consistent K‐SVD speaking, feature engineering emphasises feature stability, while the classifier focuses on recognition capacity and timeliness [26–30].…”

Ensemble convolutional neural networks for automatic fusion recognition of multi‐platform radar emitters

“…Multi‐sensor fusion of emitter signals can be implemented by combining supplementary and contradictory information from multiple systems. Multi‐collaborative representations and Dezert‐Smarandache theory were proposed to address the problem of fusion recognition.…”