Channel estimation using optimized pilot placement in OFDM system

Orthogonal Frequency Division Multiplexing has the advantage of transmitting large amount of data by splitting it into several parallel data streams that are transmitted simultaneously at different orthogonal frequencies. Thus, there is efficient use of the spectrum by allowing overlap. They are well adapted to frequency selective channels considering invariance within the time span of an OFDM symbol. In high mobility scenarios, when time-varying channel is considered, loss of orthogonality is encountered in sub-carriers. This condition provokes Inter Carrier Interference that degrades the system performance. There are several algorithms available to mitigate this effect. This work addresses channel estimation based on the positions of the pilots. Because pilots have strong impact on the mean square error of the channel estimates and influences the overall system performance. The problem of optimizing the pilot placement to reduce mean square error is NP hard and it is an integer combinatorial optimization problem. This work investigates the impact of the pilot arrangement on the channel estimate mean square error and provides an optimized pilot placement using hill climbing algorithm. The channel is estimated using least square and minimum mean square error criteria. Performance analysis of the two algorithms is performed for arbitrary pilot placement and optimized pilot placement. It is found that lower mean square error can be achieved using the proposed hill climbing method. Simulation results show that minimum mean square error have been found to have better estimation accuracy when compared to least square.