An analysis of the CDMA capacity using a combination of low rate convolutional codes and PN sequence

Abstract

In wireless personal communication services systems, all end users access the communication channel by sharing a common frequency bandwidth. Therefore efficient multiple access techniques must be implemented in order to allow all users in a given area to access the communications network facilities. Of particular interest is the code division multiple access scheme which allows random asynchronous multiple access by the users while offering a capacity-performance trade off, that is, accepting a larger number of users with a lower error performance. Therefore one of the most important factor when evaluating CDMA or any other multiple access technique is the system capacity. The shortage of optimal very low rate convolutional codes in the literature has led to a class of new quasi-optimal very low rate codes, called nested convolutional codes. These codes which are easy to construct provide a free distance that is very close to the Heller bound, therefore producing a substantial coding gain. Especially important for CDMA, these coding gains can be translated into system capacity improvements. We present an analysis of the CDMA capacity for the reverse, or uplink channel, from the mobile user to the base station. Using a constant overall bandwidth expansion, we obtain the best sharing of the CDMA bandwidth between the error correcting code and the PN sequence and show the improvement in system capacity that can be obtained over more traditional CDMA systems where almost all the bandwidth expansion is due to the PN sequence spreading.