Multicarrier modulation technique, Orthogonal Frequency Division Multiplexing (OFDM), is a modulation method that can be used for the high speed data radio communications with multipath fading. In this modulation scheme transmission is carried out in parallel on the different frequencies, which can easily be implemented using the fast fourier transform (FFT) algorithm. However, OFDM technique is known to be very sensitive to the carrier frequency offset (CFO) between oscillators in the transmitter and receiver. The frequency offset causes the intercarriere interference (ICI) among subcarriers of OFDM system and severely degrades the bit error rate performance of OFDM receiver.
For the last few decades, the relay-assisted communication systems have been received much attention as a means to enhance coverage, reliability and throughput of the cellular network. Especially, two-way relay networks, where nodes exchange information via the help of intermediate relay nodes, have attracted recent attention. A number of different protocols has been proposed for the two-way relay networks [24] - [27], where a relay node helps two end nodes exchange information. These protocols aim to improve the performance of the traditional four-phase relaying in terms of throughput, achievable rates, and power efficiency. In perspective of CFO estimator, the two-way relay networks are the special case in which the number of users is two.
We consider a multiuser uplink orthogonal frequency-division multiple access (OFDMA) system. The conventional scheme [1] using the interleaved pilot structure has the ambiguity problem [3]. The proposed pilot structure can reduce the ambiguity problem caused by MUSIC based algorithm when a CFO of one use
r is close to those for other users. The proposed pilot structure allows the sufficient frequency separation between subcarriers allocated to any two users. Simulation results show that the proposed CFO estimator performs better than the conventional estimator and maintains its performance as the CFO range increases. Since the proposed algorithm is robust to the ambiguity problem.
In interleaved pilot structure, The conventional CFO estimator based on MUSIC algorithm uses basically the full search algorithm to find in CFO range, $ 0 < \varepsilon < 1$. The cost function of CFO estimator is used to find the M largest peaks in CFO range. Hence, as the number of users (M) increases as the each CFO range and minimum difference of adjacent user`s CFO decreases. When the number of users is too large MUSIC based algorithm would not be used due to its ambiguity in determining the CFOs. We present the novel estimation algorithm and interleaved pilot structure to overcome this problem. In case of small number of users, the performance of novel estimation algorithm is slightly lower than that of MUSIC based algorithm. However, in case of large number of users, the performance of novel estimation algorithm is significantly better than that of MUSIC based algorithm.
In conventional CFO estimation in two-way relay networks [28], [29], it is possible to estimate the CFO between two sources but is not able to estimate the CFO between source and relay. In AF mode, sources should know the CFO between two sources. However, in DF mode, the relay has to decode the received signal and send re-encoded data. Hence, sources should know the CFO between source and relay. We propose the CFO estimator and modified interleaved pilot structure to estimate the CFOs between sources and relay in the two-way relay OFDM systems. The permutation matrix makes the relay rearrange sequences of signal for estimating the CFOs between sources and relay. We present a simple CFO estimation method based on the companion matrix using the received signal from the interleaved pilot structure. The proposed estimator would be extended to uplink OFDMA systems.
다중 사용자 직교 분할 시스템에서 인터리브드 파일럿 구조를 기반으로 한 반송 주파수 추정 기법은 MUSIC으로 대표된다. 그 추정 기법은 근본적으로 모호성의 문제를 가지고 있다. 모호성 문제를 해결 하는 방법은 크게 두가지로 생각 할 수 있다. 추정 기법은 바꾸지 않고 인터리브드 파일럿 구조를 변경 하는 것이고 다른 하나는 추정 기법을 새로 제안하는 것이다. 반송 주파수 추정 기법의 목적 함수에서 인접한 두 반송 주파수가 가까울 때 모호성 문제를 가지기 때문에 인접한 두 반송 주파수의 차이를 벌리기 위한 개선된 인터리브드 파일럿 구조를 제안 하였다. 그러나 사용자가 많아 질 경우 추정 기법의 기본적인 한계가 존재 하기 때문에 새로운 파일럿 구조와 새로운 추정기법을 재안 하였다. 또한 양방항 중계기 시스템은 반송 주파수 추정 기법의 관점에서 보면 사용자가 두명인 특수한 경우라고 생각할 수 있다. 그러나 기본의 방식들은 단말과 중계기와의 반송 주파수를 추정 하는 것이 아니라 두 단말 사이의 반송 주파수 오차만 추정 하는 기법이 제안 되어 졌다. AF 방식에서는 문제가 되지 않지만 DF 방식에서는 단말과 중계기 사이의 반송 주파수 오차를 추정 해야 한다. 중계기에서 신호의 순서를 바꾸는 아이디어를 사용해서 단말과 중계기 사이의 반송 주파수 오차를 추정 하는 기법을 제시 하였다. 또한 companion matrix의 특성을 이용해서 간단한 polynomial estimator 를 제안 하였다. 마지막으로 양방향 중계기에서 제안하였던 polynomial estimator를 다중 사용자 환경으로 확장을 하여 일반화 하였다. 기본적인 파일럿 구조는 인터리브드 방식을 사용하였고, MUSIC 방식과 성능을 비교 하였다.
