第一年，我們將會把重心放在OFDM系統頻帶漏損的問題上，我們將不採用IEEE std. 802.11a上所提供簡單但頻譜包含能力差的raised cosine window函式，而是試著使用具有較佳的停止帶衰減的除窗式濾波，為了降低此方法的成本，我們將會探討接收端個別通道擉立運做的可行性。
Abstract: Recently there have been considerable interests in orthogonal frequency division multiplexing (OFDM) systems. Due to the increasing demands on the transmission of vast amount of multimedia signals over wireless channel, broadband wireless communication technology has become more and more important. Among different techniques for broadband wireless transmission, OFDM is the most promising technology and it has drawn a lot of attention. However the popular OFDM systems suffer from a number of drawbacks. The frequency separation ability of the transmitting DFT filters is rather poor and their stopband attenuation decays at a rate of 1/f. To avoid causing interference to other users, costly analog filter is needed at the OFDM transmitter. Moreover it is well know that when the transmission channel has some spectral nulls, the OFDM systems suffer from the problem of performance flooring. It is the purpose of this three-year project to attempt to solve these problems. Modification techniques without changing the main structure of OFDM systems will be proposed. At the end of each year, simulations will be carried out for performance comparison.
In the first year, we will address the issue of spectral leakage of OFDM systems. Instead of the simple raised cosine function in IEEE std. 802.11a which provides little spectral containment ability, we will consider windows that have a better stopband attenuation. To reduce the cost of removing ISI caused by windowing, we will explore the possibility of channel independent processing technique at the receiver.
In the second year, we will address the issue of performance flooring. We propose to use the precoding technique.Unlike previous work, we will study nonredundant precoding technique that is suitable for the conventional popular OFDM systems. To reduce the cost of the receiver, we will employ channel independent processing technique. Moreover to further reduce the cost of transmitter, low cost precoders will be applied. If possible, we will explore the possibility of designing channel independent precoders that minimize the bit error rate (BER) and/or mean square error.
In the third year project, we will apply the spread spectrum codes to OFDM systems. The attractive spread signature codes proposed by Wornell will be considered for this application. In addition, we will use the theory of multirate signal processing to construct new spread spectrum codes.