Skip to Main Content
We consider a signature quantization scheme for a group of users in a reverse-link direct sequence (DS)-code division multiple access (CDMA). Assuming perfect channel knowledge, a receiver selects the set of signatures that maximizes an average signal-to-interference plus noise ratio (SINR) from a random vector quantization (RVQ) codebook, which consists of independent isotropically distributed unit-norm vectors. The quantized signatures are relayed from the receiver to users via noiseless rate-limited feedback channels. Previously, we have proposed to organize entries of RVQ codebook into a tree structure (TS) to speed up a search for the optimal entry. Here we extend the TS scheme for a multiuser signature quantization. Numerical results show that for a given performance, a TS-RVQ codebook can be an order of magnitude less complex than an RVQ codebook.