Research On User Paring Algorithm In LTE-A Uplink MU-MIMO System For Linear Receiver

With the emergence of new kinds of communication services, the requirements for information rate and system capacity of wireless communication system is getting higher and higher. Since frequency resource is quite limited, raising the efficiency of frequency spectrum is a research hotspot in wireless communication field.Multiple input multiple output technology can make full use of space resources without increasing wireless transmission power and channel bandwidth. It also can enhance the channel capacity and improve the utilization rate of channel effectively. MIMO technology is widely used in a variety of wireless communication system, including LTE-A. However, due to the limited size, cost and battery capacity of user equipments, almost all UE only equipped with a single antenna. Thus traditional MIMO technology is very difficult to realize in the uplink of LTE-A. Vitual MIMO technology is proposed to solve this problem. It can pair multiple indepedent user equippments which are equipped with a single antenna to compose Vitual MIMO which have the same effect as tranditional MIMO. User pairing algorithm which is the key techonolgy of virtual MIMO directly affects the performance of the system. This paper mainly researches the pairing algorithm under linear receiver.Virtual MIMO fundamental in LTE-A uplink is first described in detail in this paper and some key technical characteristics of virtual MIMO is also introduced. Then virtual MIMO system model is described and its development in LTE-A system is briefly introduced. Four basic user pairing algorithms are studied and analyzed. Under linear detection, MSE can be described explicitly. A threshold for the MSE can be set to guarantee system performance. So system can obtain maximum transmission rate by making the number of pairing users as large as possible under specific MSE threshold. Based on this idea, this paper develops MSE based user pairing algorithm and two suboptimal algorithms for linear receiver to obtain maximum transmission rate under MSE constraint. MSE based user pairing algorithm uses the mean square error after detection directly. The other two suboptimal algorithms use the minimum eigenvalue and its estimation separately to approximately the computing mean square error. The computation accuracy of these three algorithms degrades and their computation complexity decreases in turn. In the last part of this paper, simulation is given under both signal carrier and multiple carriers. The simulation results demonstrate that they achieve maximum system throughput with MSE and SER performance guarantees. The performance of these three algorithms decreases, the two suboptimal algorithms can tradeoff between computation complexity and system performance.