We propose new adaptive tree search algorithms for multiple-input multiple-output (MIMO) systems based on path metric comparison. With the fixed number of survivor paths, the correct path metric may be temporarily larger than the maximum path metric of the survivor paths under an ill-conditioned channel. There have been also adaptive path metric algorithms that control the number of survivor paths according to SNR. However, these algorithms cannot instantaneously adapt to the channel condition. The proposed algorithms accomplish dynamic adaptation based on the ratio of two minimum path metrics as the minimum is significantly smaller than the second minimum under good channel conditions and vice versa. The proposed algorithms are much less complex than the conventional noise variance-based adaptive tree search algorithms while keeping lower or similar error performance. We first employ the proposed adaptive tree search idea to K-best detection and then extend it QRD-M MIMO detection.

We develop an optimum code allocation scheme by investigating the peak to average power ratio (PAPR) characteristic of a down-link multi-carrier (MC)-CDMA system using Walsh-Hadamard code. It is shown that PAPR of a MC-CDMA system is highly dependent upon the selection of code combination. Based on this fact, we develop the allocation method which minimizes PAPR according to the number of active users. In addition, an efficient suboptimum code combination search scheme is also proposed for near minimum PAPR.