Performance Analysis And Interference Management Techniques Investigation For Massive MIMO Wireless Network

For future wireless network,it is required to support super high data rate and reliable communication at anytime and anywhere.Also,it is needed to provide high bit-rate transmission for ultra-low spatial-temporal correlation scenario,i.e.,the user with high speed.To satisfy the above requirements,massive multiple-input multiple-output(MIMO)technology is introduced to enhance the spectral efficiency and energy efficiency.Meanwhile,the wireless network deployment tends to be denser and heterogeneous to further improve the area spectral efficiency.Therefore,it is important to carry out the performance evaluation and enhancement,i.e.,the interference management for cell-edge user of massive MIMO systems.Nevertheless,the irregularity and heterogeneity of the network deployment lead the interference management to be intractable.To cope with the above problem,this paper focuses on the performance analysis of massive MIMO systems,especially on the performance evaluation of cell-edge user and the corresponding interference mitigation.Furthermore,this paper studies the interference management techniques for the massive MIMO based multi-tier heterogeneous wireless network.Specifically,the contribution of this paper is summarized as follows.Firstly,the performance evaluation for cell-edge user of massive MIMO wireless network.Based on the central limit theorem and polynomial approximation,this paper gives out the approximate closed-form expressions of the cell average spectral efficiency(CASE)for different locations of single-cell distributed massive MIMO systems under three dimensional model.By the aid of the closed-form expressions,we obtain the closed-form tight upper bound for the CASE of cell-edge user.Also,we study the CASE of the cell-center user.Simulation results validate our analysis,and show that the CASE is robust to the changes of the user’s location except the cell-edge.The CASE of cell-edge user is only about 50% of that of cell-center user.We then extend the analysis of cell-edge user performance to multi-cell scenario.Utilizing tools from stochastic geometry,we derive the accurate closed-form expression for the coverage probability of cell-boundary user in irregular multi-user MIMO cellular network.This closedform expression indicates that the performance of cell-boundary user is only about of 30% of that of inner cell user.With the help of the above closed-form expression,a simple closed-form expression for the lower bound of the coverage probability in single user case is obtained.This simple closed-form expression reveals how the massive MIMO improves the performance of cell-boundary user.As for the multi-user scenario,numerical results indicate that the coverage probability of the cell-edge user is determined by the ratio of number of base station(BS)antennas to that of users,and the performance of cell-edge user increases rapidly with this ratio.This conclusion reveals the potential of multi-user massive MIMO to enhance performance of cell-edge user in irregular wireless network.Moreover,according to the deployment gain,the analytical result based on the stochastic geometry approaches the accurate one which is gotten from the actual urban network deployment.Secondly,the interference management for cell-edge user in single-tier wireless network with massive MIMO.As for the single user case,a flexible cell-edge definition based on the stochastic geometry is proposed.Also,this definition can improve the probability of the recognition of cell-edge user.Then the approximate distribution for the number of cell-edge users in the typical cell is derived.By using the extra degrees of freedom(DoF)of massive MIMO to carefully design the precoding at BS,the interference from the neighbouring BS is nulled out for the cell-edge user.Under the above proposed interference cancellation scheme,the upper bound in closed-form expression for the success probability of cell-edge user,and the lower bound in closed-form expression for the success probability of cell-center user are given out.Numerical results show the tightness of the above bounds.It is observed that the performance of the cell-edge user can be significantly enlarged with the interference cancellation strategy,but only a little reduction of the performance for the cell-center user is detected.As for the multi-user scenario,the above cell-edge definition for the single user case is not suitable here.Therefore,an elaborate cell-edge definition that is suitable for analysis of the fairness between the celledge user and cell-center user is given for this scenario.With the aid of stochastic geometry,the distribution for the number of the users located at the cell-edge is given out in closed-form expression.By using the extra DoF of massive MIMO,the precoding design aiming at nulling out the interference from the neighbouring BS for the cell-edge users and the corresponding cooperation strategy among the BSs are analyzed.Based on this interference management scheme,the closed-form expressions for the upper bound of coverage probability of the cell-edge user and the lower bound of coverage probability of the cell-center user are derived.Moreover,a simple closed-form expression for the coverage probability under typical massive MIMO regime is obtained according to the above analysis.This simple closed-form expression shows the potential of massive MIMO with interference cancellation on achieving the cell-centre-like experience.Numerical results show that the with the increase of users and the decrement of cell-edge,more performance improvement for the cell-edge user is discovered,and the analytical results tend to be more accurate.Thirdly,the interference management for massive MIMO enabled multi-tier heterogeneous wireless networks.The completely random point processes based on the stochastic geometry can not model the spatial correlations between BSs in real wireless network deployment.Therefore,the β-GPP(Ginibre point process)model is adopted here for investigating the interference management in multi-tier heterogeneous wireless networks.Thanks to the abundant DoF configured in massive MIMO,some excess DoF can be utilized to do the precoding design to mitigate the inter-tier interference.Based on this inter-tier interference mitigation scheme,the coverage probability for both macro-cell users and small-cell users is derived.Moreover,a simple closed-form expression for the coverage probability of macro-cell users under the typical massive MIMO regime is given out.This simple closed-form illustrates impact of massive MIMO on the user performance.Simulation results verify our analysis,and show that,β-GPP model is more accurate in performance evaluation than that with completely independent point processes model.The above precoding design can not null out all types of interferences in multitier heterogeneous wireless networks.Thus,this paper further proposes a hybrid interference management strategy that combines multi-domain interference cancellation schemes to further enhance the performance of massive MIMO heterogeneous networks.We formulate the determination of the interference management schemes as an optimization problem for maximizing the overall data rate.The suboptimal solution of this optimization problem is given out according to the D.C.program(difference of two convex functions).Numerical results show that the choice of optimal interference mitigation scheme depends on the network configurations,such as the deployments of BSs and the number of users.Also,it is detected that,compared with the fixed only one or two interference cancellation methods,significant enhancement of overall data rate can be obtained under the above hybrid interference management strategy.Fourthly,user mobility-aware interference management for massive MIMO wireless networks.Considering the channel aging,handoff,and etc.that are caused by the user mobility,this paper proposes a user association based interference management scheme aiming at maximizing the so called stepped data rate for two-tier heterogeneous wireless network with multi-user massive MIMO.The probability for different types of handoffs under open access and closed access is first derived.Then,with the assistance of complete Bell polynomials,the accurate closed-form expressions for the multi-user MIMO coverage probability considering the mobility of the user under the above two different kinds of access strategies are obtained.The impact of massive MIMO on the user performance in single user case is analyzed with these closedform expressions.Furthermore,for the multi-user scenario,the simple closed-form expressions of coverage probability under the typical massive MIMO regime for low-speed users are given out.These closed-form expressions reveal the mechanism of massive MIMO to improve the performance of users with low speed.As can be seen from the numerical results that the optimal association scheme depends on the user velocity,the density of small cell BSs and the corresponding bias factor,etc.Moreover,it is found that the proposed user association scheme can largely enhance the user performance compared with fixed open access strategy.Besides,the optimal access strategy can be well predicted by the above simple closed-form expressions even for the case that the user moves with medium velocity.To reduce the handoff probability,we further discuss the BSs cooperation based handoff prevention scheme for finite motions.The determination of cooperation area to prevent the handoff of the first transition is analyzed.Then the handoff probability for the second transition under this BS cooperation strategy is derived.Also,the impact of the correlation between these two transitions on the handoff probability is investigated.