The search functionality is under construction.
The search functionality is under construction.

Keyword Search Result

[Keyword] second order statistics(2hit)

1-2hit
  • Filter Design for IBI Suppression in OFDM Based Filter-and-Forward Relay Beamforming

    Satoshi NAGAI  Teruyuki MIYAJIMA  

     
    PAPER-Wireless Communication Technologies

      Vol:
    E99-B No:9
      Page(s):
    2072-2080

    In this paper, we consider filter-and-forward relay beamforming using orthogonal frequency-division multiplexing (OFDM) in the presence of inter-block interference (IBI). We propose a filter design method based on a constrained max-min problem, which aims to suppress IBI and also avoid deep nulls in the frequency domain. It is shown that IBI can be suppressed completely owing to the employment of beamforming with multiple relays or multiple receive antennas at each relay when perfect channel state information (CSI) is available. In addition, we modify the proposed method to cover the case where only the partial CSI for relay-receiver channels is available. Numerical simulation results show that the proposed method significantly improves the performance as the number of relays and antennas increases due to spatial diversity, and the modified method can make use of the channel correlation to improve the performance.

  • Blind Source Separation Based on Phase and Frequency Redundancy of Cyclostationary Signals

    Yong XIANG  Wensheng YU  Jingxin ZHANG  Senjian AN  

     
    PAPER-Digital Signal Processing

      Vol:
    E87-A No:12
      Page(s):
    3343-3349

    This paper presents a new method for blind source separation by exploiting phase and frequency redundancy of cyclostationary signals in a complementary way. It requires a weaker separation condition than those methods which only exploit the phase diversity or the frequency diversity of the source signals. The separation criterion is to diagonalize a polynomial matrix whose coefficient matrices consist of the correlation and cyclic correlation matrices, at time delay τ= 0, of multiple measurements. An algorithm is proposed to perform the blind source separation. Computer simulation results illustrate the performance of the new algorithm in comparison with the existing ones.