Please use this identifier to cite or link to this item:
https://scholarbank.nus.edu.sg/handle/10635/182789
Title: | A STUDY ON DIRECT-SEQUENCE CODE-DIVISION MULTIPLE-ACCESS MULTI-USER DETECTION | Authors: | SIM HAK KEONG | Issue Date: | 1997 | Citation: | SIM HAK KEONG (1997). A STUDY ON DIRECT-SEQUENCE CODE-DIVISION MULTIPLE-ACCESS MULTI-USER DETECTION. ScholarBank@NUS Repository. | Abstract: | Direct-sequence code-division multiple access (DS-CDMA) is a multiple access technology that is attracting much attention recently. It is based on a spread-spectrum technology that spread the spectrum of the signal such that it occupies a bandwidth in excess of the minimum necessary to transmit the data. In this scheme, each user is assigned distinct spreading code to spread the data for transmission. With these preassigned spreading codes, all the users can transmit at the same time and in the same frequency band. Therefore, all the user signals overlap in time and frequency in this system, and the use of conventional single-user detection strategy causes multiple access interference (MAI) among the users. Multi-user detection is a scheme that can counteract the MAI problem, but the optimum multi-user detector is too computational intensive to be implemented practically. Hence, there is a need for sub-optimum multi-user detector that has a performance close to optimum and a reasonable complexity to make it technically feasible. This dissertation describes a few state-of-art sub-optimum multi-user detectors that have been proposed, such as the decorrelating detector (DD), decorrelating decision-feedback detector (DDFD) and improved DDFD (IDDFD). The DD is a linear detector that has many desirable properties, but causes noise enhancement, and requires matrix inversion operation which is difficult to implement in real time, especially in asynchronous CDMA systems. The DDFD uses a whitening filter and a decision feedback scheme to improve performance to the single user bound. The DDFD solves a number of the DDFD's problems, but increases complexity further. Although these detectors have many desirable features, they still have a number of drawbacks. Hence, I proposed the orthogonal decision-feedback detector (ODFD) [18,19] and the quasi decorrelating detector (QDD) are proposed here to reduce complexity and/ or improve performance. These two new schemes are first developed for the synchronous DS-CDMA systems, and then they are generalized to the more realistic asynchronous OS-CDMA systems, namely the asynchronous-channel ODFD (AODFD) and the one-shot AODFD for the first scheme, and the asynchronous-channel QDD for the second scheme. The ODFD is proposed mainly to reduce the complexity of the DDFD and the IDDFD. By using orthogonal matched filters in the ODFD, the computational intensive inverse-matrix operation needed in the whitening filter of the DDFD is negated. Subsequently, the updating algorithm is simplified significantly and becomes truly linear to the number of users. Simulations have shown that this reduction in complexity does not degrade performance. On the other hand, the QDD is proposed to reduce computational complexity and improve performance. The QDD is a linear detector that approximates the inverse matrix in the DD with a finite number of cascading stages. By changing the number of stages, the QDD allows a very flexible means to control the noise enhancement and the near-far resistance in the DD. With a better tradeoff between noise enhancement and near-far resistance, the QDD is found to perform at least the same as the DD. The QDD has a lower computational complexity than that of the DD because the former does not require matrix inversion. | URI: | https://scholarbank.nus.edu.sg/handle/10635/182789 |
Appears in Collections: | Master's Theses (Restricted) |
Show full item record
Files in This Item:
File | Description | Size | Format | Access Settings | Version | |
---|---|---|---|---|---|---|
b20648637.pdf | 11.45 MB | Adobe PDF | RESTRICTED | None | Log In |
Google ScholarTM
Check
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.