-
Notifications
You must be signed in to change notification settings - Fork 0
Status and Roadmap
Wiseung Jang edited this page Jul 7, 2026
·
1 revision
MMSE_CPP is currently successfully handling the core channel equalization and control symbol recovery. To achieve a complete end-to-end L1 PHY Downlink receiver, the following roadmap is proposed:
- SoA IQ Data Grid Memory Management
- CRS-based Channel Estimation & Noise Variance (
$\sigma^2$ ) Tracking - 2x2 MIMO PDSCH MMSE Equalizer (AVX2 & CUDA implementations)
- PCFICH extraction and CFI decoding
- PDCCH SFBC Demapping
- Soft Demapper Module: Convert Equalizer output (
x_hatandsinr) into QPSK, 16QAM, and 64QAM Log-Likelihood Ratios (LLRs) using Max-Log-MAP approximations. - Descrambling Integration: Fuse the Gold-sequence descrambling directly into the Equalizer/Demapper AVX kernel to save memory bandwidth.
- PDCCH Blind Decoding: CCE De-interleaving, Wrap-Around Viterbi Decoder (WAVA), and RNTI/CRC masking.
- Rate Dematching & HARQ: Circular buffer reconstruction and LLR soft combining.
- Integration with FEC: Handoff standard LLR blocks to external Turbo/LDPC decoder libraries.
🗺️ Wiki Navigation: 🏠 Home (主页) | ⬆️ Back to Top (回到顶部)
🔗 Project Links: 💻 GitHub Repository | 🐛 Report an Issue | 🛡️ Security Policy
MMSE_CPP is a high-performance LTE PHY baseband engine optimized for AVX2 and CUDA. > Released under the open-source license. Contributions are welcome!