Parity bit selected spreading sequences for spread spectrum and code division multiple access systems

Abstract

Systematic block codes compute (n-k) parity bits for an information block of k bits, and append the parity bits to the end of the information bits to form an n-bit codeword. This redundancy provides certain ability for the receiver to detect or correct a limited number of errors caused by noise and interference on the channel. In direct sequence spread spectrum (DS-SS) communication systems, the information is spread over a large bandwidth by multiplying the information bits with a high rate pseudonoise (PN) sequence. In this thesis, these two techniques are combined to create a new technique, named parity bit selected direct sequence code division multiple access (parity bit selected DS-CDMA). The proposed scheme uses the (n-k) parity bits to select one of 2 (n-k) mutually orthogonal PN sequences. Then the selected PN sequence is used to spread each bit in the information block of k bits. But the (n-k) parity bits are not appended to the end of the information bits, and are not transmitted to the receiver. The receiver can obtain the parity bits by using 2(n-k) matched filters to determine which PN sequence is used for an information block of k bits. Then the errors can be corrected by employing a decoding algorithm. Block codes with different lengths are considered in our simulation in order to find which one is best for the proposed scheme at a given Eb/No. The simulation results in a multi-user environment show that the proposed parity bit selected DS-CDMA system can provide better BER performance than both the uncoded DS-CDMA and the block coded DS-CDMA systems.