Enhanced trellis based vector quantization for coordinated beamforming

This paper presents two techniques to boost the quantization performance of trelli-based beamforming vector quantization schemes [?], [?]. It is well known that tail-biting trellises exhibits performance superior to truncation without sacrificing coding rate. The drawback for using a general tail-biting trellis is its high decoding complexity. We present a special trellis suitable for the beamforming vector quantization application where efficient encoding and decoding can be performed by traversing the trellis twice. Trellis-based quantization is especially suitable for reporting channel state information in multi-base-station or base-stationrelay coordinated transmission schemes. Due to different path loss and shadowing from each transmission station, the receiving mobile experiences different long-term average signal powers from each transmission station. In the second technique, we introduce a mechanism for dynamically allocating quantization bits amongst different parts of the ...

• Type: Conference
• Year: 2010
• Where: ICASSP
• Authors: Chun Kin Au-Yeung, Shahab Sanayei
• Keywords: General Tail-biting Trellis, ICASSP 2010, Signal Processing, Trelli-based Beamforming Vector, Vector Quantization