BLAST
Understanding BLAST
Before BLAST, multiple antennas were used only for diversity (improving reliability). BLAST showed that with sufficient scattering (rich multipath), multiple antennas can create parallel spatial channels, each carrying an independent data stream. The capacity scales linearly with min(Nt, Nr) rather than logarithmically with SNR alone.
V-BLAST uses zero-forcing or MMSE detection with successive interference cancellation: detect the strongest stream first, subtract it, then detect the next. D-BLAST uses space-time diagonal coding for better performance but higher complexity. Modern systems use MMSE-SIC or maximum likelihood detection.
C = ∑i=1min(Nt,Nr) log2(1 + λi·SNR/Nt)
λi = eigenvalues of HHH
High-SNR scaling:
C ≈ min(Nt,Nr) × log2(SNR/Nt)
BLAST Variant Comparison
| Variant | Encoding | Detection | Diversity | Complexity |
|---|---|---|---|---|
| V-BLAST | Per-antenna | ZF/MMSE-SIC | None (multiplexing) | Low |
| D-BLAST | Diagonal ST | ML/SIC | Partial | High |
| Alamouti | Orthogonal STC | Linear | Full (no mux gain) | Low |
Frequently Asked Questions
V-BLAST vs D-BLAST?
V-BLAST: one stream per antenna, SIC detection, simpler. D-BLAST: diagonal coding across antennas, better diversity, higher complexity and latency.
How does it multiply throughput?
min(Nt,Nr) independent streams in the same bandwidth. Capacity scales linearly with antenna count at high SNR.
Still used?
The name is historical; the spatial multiplexing principle underlies LTE (8 layers), 5G NR (8 layers), Wi-Fi 6/7 (16 streams). Detection methods evolved to MMSE-SIC and ML.