With sparse recovery algorithms for pilot assisted channel

With the recent development of digital communication systems and the emergence of new communication technologies have made the spectral efficiency of high concern. Transmitting of pilot signals on the available spectrum helps the receivers to estimating the corresponding communication channels efficiently. Various pilot designs for equally-spaced pilot sequence was used to estimate the equivalent channel based on the least squares (LS) channel estimation. In recent years, compressed sensing has emerged a great role of reconstructing sparse signal from a set of samples in wireless sensor networks. This is based on the concept that, the sparsity of a signal can be destructed to reconstruct it from fewer samples required by the Shannon Nyquist sampling theorem through optimization 1. Sensing algorithms are implemented to satisfy the metric’s criteria like reliability, complexity and loss in system throughput 2, 3. Compressed sensing theory reduces the amount of pilot symbols with large number of transmitters in MIMO system 4.

Sparse channel estimation is efficient than conventional algorithms due to the sparse nature of multipath wireless channels. Compressed sampling matching pursuit, orthogonal matching pursuit, and basis pursuit are various sparse recovery algorithms for pilot assisted channel estimation 5, 6. Comb type and block type are the basic pilot patterns which can be used in channel estimation based on LS, MMSE and Linear MMSE techniques 7-9.Optimization problem is formulated to minimize a mean square error of upper bound in pilot design 10, 11.   It uses pilot symbol design algorithm and joint pilot placement to increase the accuracy in channel estimation 12-15.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

In minimum error rate and maximum sum rate method, different pilot value has been assigned for each transmitter 16-19. The performances of this estimation, recovery, and data detection and receiver blocks of OFDM relay networks were improved at the presence of CFO and noise 20, 21. BER of CS based channel estimation is lower in frequency selective fading channel 22, 23.

In traditional channel estimation methods, equidistant pilot placement is commonly used as optimal. It uses random pilot patterns which degrade the performance of channel estimation. In case of CS based channel estimation, the pilot placement is based on reducing the mutual coherence of measured matrix. The mutual coherence value obtained for the traditional pilot patterns are high and it increases the bit error rate of the received signal.  So, in the present work pilot pattern for OFDM system, feedback based Least Square channel estimator and  pilot pattern is selected with improved shuffled frog leaping algorithm (ISFLA) is proposed.  It improves the channel estimation performance effectively overcome the ill effects of mismatching of delay spread and the Doppler frequency shift making it a more robust against the variation of doppler shift and delay time specially at higher Eb/No and also the proposed pilot allocation for CS based method provides better performance than traditional channel estimation methods with respect to MSE and BER.