The performance of space-time block codes in providing transmit diversity is severely degraded when strong localized interference is present. This problem is addressed by investigating a recently proposed coherent space-time block code decoding algorithm for unknown interference suppression. The algorithm assumes a Gaussian noise and interference approximation and is based on a cyclic-based maximum-likelihood estimation technique (CML). In this thesis, simulations are done applying CML in a coherent system with unstructured interference to validate previous work. An extension of these results is obtained by examining factors that affect CML performance and modifying CML for use in a noncoherent system. To improve bit error rate performance, a turbo code for channel coding was added to both systems. This addition required the development of reliability metrics for soft-information transfer between the space-time block code detector and the turbo code decoder. Significant coding gains exceeding 8dB at a bit error rate of are achieved for the turbo-coded system when compared to that of an uncoded system.
| Identifer | oai:union.ndltd.org:uvic.ca/oai:dspace.library.uvic.ca:1828/452 |
| Date | 10 April 2008 |
| Creators | Chow, William. |
| Contributors | Gulliver, T. Aaron |
| Source Sets | University of Victoria |
| Detected Language | English |
Page generated in 0.002 seconds