Cooperative strategies for imperfect CSI scenarios based on distributed Alamouti