Thesis
On detection of OFDM signals for cognitive radio applications
- Creator
- Rights statement
- Awarding institution
- University of Strathclyde
- Date of award
- 2019
- Thesis identifier
- T15423
- Person Identifier (Local)
- 201351573
- Qualification Level
- Qualification Name
- Department, School or Faculty
- Abstract
- As the requirement for wireless telecommunications services continues to grow, it has become increasingly important to ensure that the Radio Frequency (RF) spectrum is managed efficiently. As a result of the current spectrum allocation policy, it has been found that portions of RF spectrum belonging to licensed users are often severely underutilised, at particular times and geographical locations. Awareness of this problem has led to the development of Dynamic Spectrum Access (DSA) and Cognitive Radio (CR) as possible solutions. In one variation of the shared-use model for DSA, it is proposed that the inefficient use of licensed spectrum could be overcome by enabling unlicensed users to opportunistically access the spectrum when the licensed user is not transmitting. In order for an unlicensed device to make decisions, it must be aware of its own RF environment and, therefore, it has been proposed that DSA could been abled using CR. One approach that has be identified to allow the CR to gain information about its operating environment is spectrum sensing. An interesting solution that has been identified for spectrum sensing is cyclostationary detection. This property refers to the inherent periodic nature of the second order statistics of many communications signals. One of the most common modulation formats in use today is Orthogonal Frequency Division Multiplexing (OFDM), which exhibits cyclostationarity due to the addition of a Cyclic Prefix (CP). This thesis examines several statistical tests for cyclostationarity in OFDM signals that may be used for spectrum sensing in DSA and CR. In particular, focus is placed on statistical tests that rely on estimation of the Cyclic Autocorrelation Function (CAF). Based on splitting the CAF into two complex component functions, several new statistical tests are introduced and are shown to lead to an improvement in detection performance when compared to the existing algorithms. The performance of each new algorithm is assessed in Additive White Gaussian Noise (AWGN), impulsive noise and when subjected to impairments such as multipath fading and Carrier Frequency Offset (CFO). Finally, each algorithm is targeted for Field Programmable Gate Array (FPGA) implementation using a Xilinx 7 series device. In order to keep resource costs to a minimum, it is suggested that the new algorithms are implemented on the FPGA using hardware sharing, and a simple mathematical re-arrangement of certain tests statistics is proposed to circumvent a costly division operation.
- Advisor / supervisor
- Crockett, Louise H., (Louise Helen)
- Stewart, Robert.
- Resource Type
- DOI
- Date Created
- 2019
- Former identifier
- 9912778493402996
Relations
Items
| Thumbnail | Title | Date Uploaded | Visibility | Actions |
|---|---|---|---|---|
|
|
PDF of thesis T15423 | 2021-07-02 | Public | Download |