Research On Precoding For The Next Generation Wireless Network

With the increasement of the wireless devices and the explosive growth of the wireless ser-vices, existing network system cannot meet the requirements of future services. The next generation wireless network will combine new technologies such as new transmitted tech-niques and new network collaboration techniques etc, and provide 1000 times capacity than existing network. The next generation wireless network will continue to evolve on the base of the existing network, it will compose super dense network by combining such as relay, cog-nitive network and multi antenna technologies, etc., to enhance the capacity of the system. In order to provide high spectral efficiency of the system, full duplex technique may also be used by the next generation wireless network. Super dense network and the same frequency multiplexing will introduce interference to the wireless network. As a result, the cancelation of the interference between different users and the coordinating transmission of different net-works will become important issues for the next generation wireless network. Supported by the State Key Program of National Natural Science of China (61231008), by National Basic Research Program of China (2009CB320404), by Program for Changjiang Scholars and In-novative Research Team in University(IRT0852), by National Natural Science Foundation of China (61301168), and the 111 Project (Grant No. B08038), in this paper, we put precoding as the research content, start from different network scenarios which are compositions of the fu-ture network, include one-way relay assisted cellular network, two-way relay assisted cellular network, interference cognitive radio and cooperative cognitive radio, design the correspond-ing precodings to increase the performance of the system, such as spectral efficiency (SE) and energy efficiency (EE) etc. The main contributions are listed as follows.1. The downlink transmission of one-way relay assisted cellular network is studied by con-sideration of a weak direct link. To satisfy multiple services of users, a transmission scheme that supports different data streams is designed, and based on the transmission scheme, the corresponding structured precoding algorithms are proposed. In order to make the base station (BS) transmit data conveniently, precodings of the BS are designed for direct link and the relay link, respectively. As the receive data of the direct link is weak, the BS precoding that for the direct link is designed to make the interference of the BS-destination link as low as possible. For the relay link, different precodings are designed for the different stages. In the first stage, in order to avoid the interference being amplified by the relay, the BS precoding that for the relay link is designed to cancel the inter-link interference. In the second stage, precoding of the relay station (RS) is designed to maximize the SE. To reflect the characters of the precodings apparently and clearly, precoding structures are derived for the BS and the relay by different methods. The precoding structures are designed first by combining the null space vectors of useful signal (Null-based) or by combining the orthogonal subspace vectors of the interference signal (Orth-based), then the parameters that introduced in the structures are optimized. The obtained precoding structures can be not only used for solving the SE optimization problem, but also used for solving the EE optimization problem. Simulation results show that the pro-posed precoding design algorithms can obtain higher SE or higher EE compared with existing methods.2. The precoding designs for the two-way relay assisted cellular network are studied. As there are several users in the network, precodings are designed to achieve EE balancing of the users. Due to non-convexity and NP hard property of the worst case EE problem, we use convex ap-proximation and convex relaxation to modify the problem for better mathematical tractability. The Dinkelbach-type method is adopted to solve the max-min fractional problem. Then Alter-nating optimization method is used to solve the inner subproblem as its non-convexity. After fixing the precoding of one station (BS or RS), the subproblem is still nonconvex. Then suc-cessive convex approximate technique (SCA) and semidefinite programming (SDP) methods are used to obtain the precoding matrices of BS and RS. The proposed algorithm, which can be extended to the two way relay system with direct link model or the one way relay system, has a wide range of application. Simulation results show that the proposed algorithm converges fast and achieves better EE balancing compared with existing method.3. For the multiple input multiple output (MIMO) interference cognitive radio network, a low-complexity interference alignment (IA) algorithm is proposed to cancel the interference between the primary user (PU) and each secondary users (SUs). The cores of the framework are the problem formulation of the IA transceiver design based on the IA zero-interference condition and the solution to the IA formulation function. The proposed algorithm not only effectively aligns the interference into the null space of the desired signal, but also maximizes the size of desired signal subspace. Simulation results show that the proposed algorithm can achieve higher sum rate than existing IA methods and can achieve the tradeoff between the interference and the sum rate of the system.4. The precodings for the spectrum-sharing-based cognitive radio are designed to maximize the weighted sum rate (WSR) of SUs. In order to effectively utilize the spectrum of wireless network, SUs are often authorized to share spectrum with PUs in cognitive networks. To guarantee effective transmission of PUs and SUs, the interference from the SUs to the PUs is limited and only the SUs that have the information rates than the corresponding thresholds are allowed to transmit their data. In order to select proper SUs to transmit and to maximize their achievable WSR under rate requirement and interference constraint, we propose a joint admission control and precoding design algorithm for partial channel state information (CSI) case. The WSR maximization problem is first constructed by using approximation and convex relaxation according to the statistical property of the CSI. Then a penalty factor is introduced to check the feasibility of such problem and instruct the admission of the SUs. Beyond the feasible SUs set, the WSR is maximized by updating the achievable rate of each user with rate profile technique and polyblock outer approximation method. Simulation results show that the proposed precoding can achieve higher WSR than the precoding of existing baseline method.