This paper proposes adaptive control of the number of surviving symbol replica candidates, Sm (m denotes the stage index), based on the minimum accumulated branch metric of each stage in maximum-likelihood detection employing QR decomposition and the M-algorithm (QRM-MLD) in orthogonal frequency-division multiplexing with multiple-input-multiple-output (MIMO) multiplexing. In the proposed algorithm, Sm at the mth stage (1 ≤ m ≤ Nt, Nt is the number of transmission antenna branches) is independently controlled using the threshold value calculated from the minimum accumulated branch metric at that stage and the estimated noise power. We compared the computational complexity of QRM-MLD employing the proposed algorithm with that of conventional methods at the same average packet error rate assuming the information bit rate of 1.048 Gb/s in a 100-MHz channel bandwidth (i.e., frequency efficiency of approximately 10 bit/s/Hz) using 16QAM modulation and turbo coding with the coding rate of 8/9 in 4-by-4 MIMO multiplexing. Computer simulation results show that the average computational complexity of the branch metrics, i.e., squared Euclidian distances, of the proposed adaptive independent Sm control method is decreased to approximately 38% that of the conventional adaptive common Sm control and to approximately 30% that of the fixed Sm method (Sm = M = 16), assuming fair conditions such that the maximum number of surviving symbol replicas at each stage is set to M̂ = 16.
- Broadband radio access
- Maximum-likelihood detection (MLD)
- Multiple-input-multiple-output (MIMO)
- Orthogonal frequency-division multiplexing (OFDM)
- QR decomposition