Abstract

Increasing demand for high data rate wireless communication motivates the wireless engineers to develop advanced technologies to address such needs. LTE and LTE-Advanced are examples of such wireless technologies, which support high data rate and a large number of users. However, higher data rate communication requires more frequency bandwidth.
Recent studies have shown that the inefficient utilization of frequency spectrum is one of the main reasons for the scarcity of frequency bandwidth. Cognitive Radio Network is introduced as a promising solution for this problem. It increases the utilization of bandwidth, by intelligently sensing the channel environment and dynamically providing access to the available resources (frequency bands) for a secondary user.
In this thesis, we developed an algorithm for dynamically detecting and anticipating the existence of underutilized resources in LTE system. The algorithm should be a real-time operation, i.e. the decision on availability of a detected resource should be made within a time much less than scheduling update period of LTE. This is the only way that rest of the unused resources becomes usable. For each specific channel assignment, the algorithm requires to start sensing as soon as possible.
Therefore, we develop the algorithm in three main steps. The first step is to blindly detect and identify the LTE-Downlink signal using cyclostationarity property of OFDM scheme. The second step is the acquisition of the LTE-Downlink sub-frame timing, which is basically performed by detecting the Primary Synchronization Signal.
The third step is to detect unused resources, for the duration of their transmission. This step is using a frequency domain energy detector. By performing the first and second steps, the sub-frame timing and scheduling update instances are known. So basically the algorithm does not require any previous knowledge of the LTE signal.
We evaluate the performance of the proposed algorithm with respect to the tolerable amount of interference at the primary user side. Using the proposed algorithm, in average up to 81 % of unused resources can be used by the secondary user.