Inter-Carrier Interference Mitigation by Means of Precoding Michal ˇ Simko 1 , Qi Wang 1 , Paulo S. R. Diniz 2 and Markus Rupp 1 1 Institute of Telecommunications, Vienna University of Technology, Vienna, Austria 2 Department of Electronics - School of Engineering, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil Email: msimko@nt.tuwien.ac.at Web: http://www.nt.tuwien.ac.at/ltesimulator Abstract—In this work, we discuss the possibility of using precoding as means to mitigate Inter Carrier Interference (ICI) caused by temporal channel variations in Orthogonal Frequency Division Multiplexing (OFDM) systems. Many different precod- ing schemes have been introduced in the past. However, the promised gains of these techniques are not achievable nor re- alistic. In this paper, we introduce a practical low-cost precoding technique to mitigate ICI caused by Doppler spread. The gain in terms of Signal to Interference and Noise Ratio (SINR) is 0.5 dB at a velocity of 500 km/h. Index Terms—LTE, ICI, OFDM, Precoding. I. I NTRODUCTION Current Orthogonal Frequency Division Multiplexing (OFDM) based systems for wireless communication are sen- sitive to channel variations [1]. The time variant distortion such as the Doppler spread destroys the designed orthogonality between subcarriers and causes Inter Carrier Interference (ICI). Consequently, the performance of such a system is limited. A. Related Work In previous work, one could recognize principally two different techniques of mitigating ICI caused by the Doppler spread. A first type of techniques is based on advanced equalization techniques. In the time-invariant case, it is suf- ficient to use single tap equalizers. In [2], a so-called - tap equalizer was introduced. This equalizer considers also channel knowledge of the neighboring subcarriers, which results in better performance. However, this solution has not only higher complexity, but also requires more information about the channel to be estimated. The second type of ICI mitigation techniques is based on precoding. The data vector is spread over different subcarriers which reduces the impact of ICI on the system performance as introduced in [3]. The same data symbols are transmitted on adjacent subcarriers, but rotated by 180 degrees, effectively reducing ICI at the cost of 50% bandwidth reduction. The authors of [4] introduced a precoding method, which does not cause bandwidth reduction. However, a Maximum Likelihood (ML) receiver is required, which significantly increases the complexity. In [5] so called frequency domain Partial Response Coding (PRC) was in- troduced. This technique can be viewed as special type of precoding, where only adjacent subcarriers are precoded. This technique attracted a lot of interest in the past few years, due to its high theoretical gain in terms of reduction of the ICI power. However, the performance metric utilized by the authors does not necessary reflect performance of a real system. As already argued in [6], PRC as it is introduced in [5], does not improve system’s performance, it rather limits it. In [7] a precoding technique is proposed to mitigate ICI utilizing eigendecom- position of the channel matrix. This method is however not of practical relevance due to the requirements on extremely low-delay feedback between receiver and transmitter and its high computational complexity. B. Contribution In this paper, we will show how to use precoding techniques in the frequency domain to mitigate ICI of an OFDM based system. The main contributions of the paper are: ∙ We introduce a precoding technique that effectively miti- gate ICI caused by channel variation and does not require an ML receiver. ∙ As cost function we choose the Signal to Interference and Noise Ratio (SINR) that directly reflects performance of a transmission system. ∙ All data, tools, as well as implementations needed to reproduce the results of this paper can be downloaded from our homepage [8]. The remainder of the paper is organized as follows. In Section II, we describe the mathematical system model. In Section III, we extend our system model for usage of pre- coding. Furthermore, we show how to design a precoder that maximizes the post-equalization SINR. Finally, we present simulation results in Section IV and conclude our paper in Section V. II. SYSTEM MODEL In this section, we introduce a mathematical model of the considered transmission system under time-variant channels.