An improved pilot pattern design-based channel estimation in wireless communication using distribution ant colony optimization

Abstract

Multiple-Input Multiple-Output (MIMO) communication System-Orthogonal Frequency Division Multiplexing (OFDM) approach can produce reliable transmission of high data rates over broadband wireless medium. This approach makes use of the available spectrum and proves to be a potential candidate for meeting the high demand for data rate. Therefore, channel estimation is a crucial component in MIMOOFDM systems. There exist several issues in channel estimation like Carrier Frequency Offset (CFO) estimation, power allocation and orthogonal pilot optimization. However, computing an optimal pilot pattern is the complex problem. It severely reduces channel estimation accuracy. To overcome this issue, the proposed work focuses on the best possible design for the pilot pattern to get reliable channel estimation. In this proposed work, two pilot pattern design approaches like Carrier Frequency Offset (CFO) estimation, power allocation and orthogonal pilot optimization are combined to provide orthogonal pilot patterns. Creating orthogonal pilot patterns based on Generalized Shift Invariance Property (GSIP) for a MIMOOFDM has been proposed using two pilot pattern design approaches based on Improved Pilot Design with GSIP (IPDGSIP) and Tradeoff Pilot Design with GSIP (TPDGSIP). In PDGSIP, pilot patterns with high mutual coherence are initially obtained, then updated with the best pilot patterns. Here the optimal pilot patterns are generated by using Distribution-based Ant Colony Optimization (DACO) algorithm. Once a pilot pattern is established, TPDGSIP immediately generates new pilot patterns based on GSIP to utilize the pilot distance of the obtained pilot pattern fully. Results from MATLAB simulation demonstrate that the suggested technique performs better in channel estimation in terms of Bit Error Rate (BER), Normalized Mean Square Error (NMSE) and Symbol Error Rate (SER). The proposed PDGSIP-DACO system reduces the NMSE (6%), BER (8%) and SER (20%) when compared to the PDGSIP method.