Online feedback-based rating systems are gaining popularity. Dealing with unfair ratings in such systems has been recognized as an important but difficult problem. This problem is challenging especially when the number of regular ratings is relatively small and the unfair ratings contribute to a significant portion of the overall ratings. In this paper, we propose a novel algorithm to detect the unfair ratings that cannot be effectively prevented by existing state-of-the-art techniques. Our algorithm is particularly effective to detect malicious raters that collaboratively manipulate ratings of one or several products. The main idea of our algorithm is to use an autoregressive signal modeling technique combined with trust-enhanced rating aggregation. We are able to detect and filter out unfair ratings very accurately. Extensive experiments through simulations and real-world data have been performed to validate the proposed algorithm. The experimental results show significant improvements on detecting collaborative unfair raters over existing techniques.