Cooperative Precoding And Interference Coordination In Cellular Systems

With the popularity of smart phones, tablet computers and other mobile devices, mobile data traffic grows rapidly. In order to reach peak rate up to1Gb/s for high mobility and100Mb/s for low mobility, carrier aggregation, downlink multiple-input multiple-output, heterogeneous networks and other techniques are proposed during the standardization process of3rd Generation Partnership Project Long Term Evolution (LTE)-Advanced. For an OFDM-based cellular system, the orthogonality of subcarriers makes intra-cell inter-ference neglectable. However, when the system has universal frequency reuse, the transmission rate achieved by cell edge users is impacted by inter-cell in-terference (ICI). In a typical macro cellular scenario, cell edge throughput is a small fraction of the peak rate. Inter-cell interference coordination is a tech-nique that avoids or mitigates ICI by coordinating frequency, time or power resources among different cells. It is easy to implement, flexible and effec-tive. Coordinated beamforming/precoding is a strategy that coordinates space resource among transmitters. Since the precoding matrices are designed semi-statically or dynamically, according to the scheduling and channel statement information (CSI), it is a more impressive way to improve cell edge through-put.This thesis first studies multi-cell coordinated precoding in homogenous networks. Specifically, the cases of multi-user multiple input single output (MISO) transmission and single-user multiple input multiple output (MIMO) transmission in each cell are investigated respectively. Then, interference coor-dination between closed Subscriber group (CSG) femto base station (BS) and the operator of macro BSs in heterogeneous networks is analyzed. The main contribution of this thesis is as follows. 1. For the coordinated precoding in multi-cell MISO system, leakage-controlled minimum mean square error (LC-MMSE) is proposed. This pre-coding scheme enables transmitter to serve users in its cell as well as mitigate interference to neighboring cells. The signal to interference plus noise ratio (SINR) of LC-MMSE scheme is analyzed and the optimal regularization fac-tor that maximizes SINR is derived. A joint frequency and space coordination scheme is proposed, where MMSE precoding scheme is employed to transmit data to serving users on primary frequency band, and LC-MMSE precoding scheme is imposed at the transmitter to reduce interference leakage to adjacent cells on non-primary resources. Compared to signal to leakage plus noise ra-tio (SLNR) precoding, the proposed spatial coordination scheme requires less amount of feedback and lower computational complexity, and thus is more easy to implement in real system. Simulation results demonstrate that the proposed scheme can significantly mitigate ICI, without loss of average cell throughput, even with imperfect channel statement information at the transmitter.2. For the coordinated precoding in multi-cell MIMO system, the sys-tem is modeled as multi-link MIMO Gaussian interference channel. Cross link interference is mitigated by imposing null shaping constraint on the de-sign of transmit covariance matrices. Given the cooperation relationship, the multi-cell cooperative transmission is formulated as strategic game. The Nash Equilibrium (NE) is derived, and the existence and uniqueness of the NE is analyzed. The generation of cooperation relationship is formulated as coali-tion formation game, coalition graph formation game and network formation game, and corresponding algorithms are proposed respectively. The proposed coalition algorithms are Nash-stable, and the proposed network formation al-gorithm converges to a Nash Equilibrium. Simulation results reveal that the proposed coalition algorithms improve the transmission rate apparently when the number of antennas at the transmitters is large, and the network formation algorithm has significant advantage even at low SNR region and/or with small number of transmit antennas. 3. For the interference coordination between CSG femto BS and the oper-ator of macro BSs, first, stochastic geometry is used to model the locations and number of macro/femto BSs and the coverage probability (non-outage proba-bility) of indoor macro users and femto users is derived. Under the constraint of minimal coverage probability of macro users, the optimal active time ratio and transmit power that maximize outage throughput of femto users are de-rived. Assume that it is not necessary for the femto BS to mitigate interference to nearby macro users altruistically, and a femto BS adjusts the interference leakage according to the reward from the operator of macro BSs. We formu-late the interference coordination as price games, and obtain the closed-form of Nash equilibria. Simulation results demonstrate that time mute scheme per-forms better than power control scheme in handling indoor macro user coverage problem.