Article Preview
TopIntroduction
In a wireless channel fading, adds both constructively and destructively to the transmitted signal which distorts the received signal (Durai, 2012). To successfully recover the transmitted information the channel knowledge at the receiver is important. But, when the channel is highly time-varying and random, the assumption of perfect Channel State Information (CSI) becomes a bit complex. Adaptive arrays and multi-antenna techniques are emerging options for improving the wireless link and mitigating interference in a multiple access system. Also, diversity techniques combat fading by increasing the probability that the receiver Signal-to-Noise Ratio (SNR) remains above some arbitrary threshold relative to the average receiver SNR. This in turn improves the average Symbol Error Rate (SER) and Bit Error Rate (BER) relative to a system with same average SNR but no diversity. Moreover, increasing the transmit power of the signal in a wireless communication system leads to significant BER performance improvement. But that does not satisfy the demanding needs of low power consumption based systems and hence, increasing the transmit power of the system cannot be a good solution to such problems. Therefore, development of a more efficient technique is yet an open issue in the field of wireless communication. The most frequently used technique is the adaptive processing and use of multi-antenna systems. A few uses diversity and simple equalization techniques. When combined with Orthogonal Frequency Division Multiplexing (OFDM), these multi-antenna set-ups help to achieve high data rate transmission with significant communication quality.
So, here in this chapter a self-adaptive antenna system is designed which is tested for two common systems -OFDM based single input-single output (SISO) and multi input-multi output (MIMO) system with binary phase shift keying (BPSK) and a comparison between the two set-ups is provided. It is assumed that the receiver has no knowledge of the CSI. A threshold limit for the BER is set at the receiver and until that threshold is reached the receiver keeps on sending a feedback to the transmitter to keep increasing the effective aperture of the transmitting antenna. When in a situation the impact of fading is not very high the receiver even notifies the transmitter to decrease the effective aperture. And this is accomplished for every time instant. However, in a real time situation it is not possible to have a fixed BER threshold value at a particular instant of time due to the time-varying nature of the channel. As such, there exists some percentage of tolerance of the BER value and hence there exists a range for effective aperture values of the transmitting antenna corresponding to the range of BER values. Thus, this paper mainly focuses on a method to take priory decision to determine the best possible effective aperture values in a real time environment. For this an efficient classification or clustering algorithm is required. Different clustering algorithm can be used for clustering purpose but two very commonly used of the data clustering algorithms are Fuzzy C-Means (FCM) algorithm and Fuzzy K-Means algorithm. FCM algorithm is very popular among the two as it is easy to use, straight forward and very efficient. FCM allows one piece of data to belong to two or more clusters. This method was developed by Dunn in 1973 and improved by Bezdek in 1981 and it is frequently used in pattern recognition. In fuzzy clustering, one object can be clustered in more than one cluster according to the degree of membership function. Because of the ease of implementing FCM is used in this work to take the priori decision. The clustering method gives us two values of the effective aperture within which satisfactory performance of the system can be achieved. Finally, the radiation pattern of the transmitting antenna for these two limits and the effective aperture obtained before applying FCM are plotted for both SISO and MIMO set-ups. The directivity is also calculated and a comparative study between the two set-ups is provided. The rest of the chapter is organized into the following headings-Related work, background theory, the system model and the methodology, results & discussions, the future research direction and the conclusion.