Deterministic Subspace Based Blind Channel Estimation for Doubly-selective Channels

In this paper, a new deterministic subspace based blind channel estimation algorithm for doubly-selective channels is introduced. It operates on timeand frequency-shifted versions of the received sequence, and requires only two receive antennas. Furthermore, due to its deterministic nature, it needs only a short data record to work reliably. The proposed algorithm unifies and extends previously proposed deterministic subspace based blind channel estimation algorithms. Performance results also show an improved performance over an existing stochastic subspace based blind channel estimation algorithm, which also requires only two receive antennas.