Research On Signal Processing In Uplink And Downlink Transmission In Ultra Dense Wireless Networks

With the continuous progress of wireless communication technology,each generation of wireless communication system is developing by leaps and bounds.In the 5G communication system,a series of technologies are proposed to improve wireless network capacity and communication service quality.In order to improve the frequency spectrum efficiency in per unit area,the next generation wireless communication system will form ultra-dense wireless network(UDN)by densely deploying access points(AP)in some hot spots.Through densely deploying APs,the characteristics of AP are user-centered and user-centered.On the one hand,the distance between the user and the base station is reduced to ensure the effective transmission of data between the user and the base station.On the other hand,densely placed APs provide a large number of network interfaces for transmitting low-power signals,so that ultra-dense wireless networks not only effectively improve system performance,but also significantly reduce system power consumption.As a large number of APs are densely deployed in an ultra-dense wireless network system,coverage overlap between adjacent aps may occur,which leads to resource allocation problems in the system.Only reasonable deployment and networking of ultra-dense wireless networks can avoid a large amount of resource waste.However,due to the change of ultra-dense wireless network system model,the traditional cellular network model with a cell as the unit of multiplexing scheme is no longer applicable,so a new resource multiplexing scheme needs to be proposed to allocate resources.At the same time,such overlapping coverage can also lead to interference between APs,which is a very important factor limiting the improvement of system performance.Therefore,effective interference management schemes are required to reduce the interference between adjacent APs.However,the interference management schemes used in traditional cellular networks may not be suitable for ultra-dense wireless networks,so a new interference management scheme needs to be proposed.Firstly,in view of the channel characteristics of ultra-dense wireless networks,this thesis studies the uplink transmission,and proposes an efficient and low-complexity multi-user detection algorithm based on interference elimination under local cooperation,which provides a practical method of uplink multi-user access and detection for the actual deployment of ultra-dense wireless networks.First for ultra-dense wireless network system aggregation characteristics of channel matrix,the use of compressed sensing technology for channel estimation,the channel estimation method can automatically according to the channel strength user pairing with AP,can achieve multi-user diversity gain,both at the same time also can let the AP based on channel strength,through local collaboration between the AP and the adaptive interference elimination.Compared with traditional minimum mean square error detection,this method avoids matrix inverse operation and reduces the complexity.At the same time,there is no need to share channel information in the system,and the central processing unit of the system only needs to do some basic scheduling,which reduces the requirement of information sharing.The final simulation results show that the proposed algorithm has strong applicability and can achieve a good balance between complexity and performance.Secondly,this thesis focuses on the downlink transmission of ultra-dense wireless networks.After analyzing the spectral efficiency,the power distribution scheme is studied.Based on the consideration of performance and complexity,a power allocation method for a single AP based on partial user forced zero and local cooperation on demand is proposed.First of all,it is assumed that each AP knows the location information of users,and each AP only serves the users around it.The zero-forcing criterion is used to eliminate the interference between the users served,instead of eliminating the interference between all users,that is,the partial user-forcing zero-forcing algorithm.After pre-coding,power is allocated according to the channel large-scale fading factor.However,due to the random distribution characteristics of APs and users in ultra-dense wireless network system,users may not be able to meet their minimum communication requirements when they fall in the region with relatively sparse APs.In view of these users,a local cooperative cluster is formed,and the power is redistributed by optimization method.The simulation results show that this algorithm can avoid the computational complexity caused by global optimization of power resources and ensure the system performance.