Analysis And Optimization Of Cooperative Networks Under Constraint Backhaul

Interference and fading are two key limitations of the coverage and reliablitity, especially for the users near the cell edge where the qulity of service (QoS) deterriat-tes sharply due to the strong interference from the nearby cells. To address this issue, joint multicell processing (MCP) or cooperation of multi-point (CoMP) is adopted as a superior approach to eliminate inter-cell interference. By the cooperation of the Base Stations (BSs), joint decoding in the uplink and the joint procoding in the downlink are employed to overcome the interference effect. Teoretically, the interference can be eliminated entirely when the BSs share all the data and perfect channel state informa-tion (CSI) among the cooperative BSs. However, in the practical deloyment of multicell processing, the backhaul via which the BSs exchange information with each other are constrained by the finite-capacity or delay owning to the imperfect backhaul links.Thus, the BSs will utilize the compress-and-forward scheme to exchange information with each other.For the capacity-constraint backhaul, the problems about raising the utilization of the finite capacity of backhaul links are studied, such as the multicell cooperative decoding scheme from the user terminal, the selection of quantization schemes from the Base stations and the performance analyses of joint detection in the central processor, the main results are as follows:1) The designs of the multicell cooperative decoding scheme are studied under the constraint of the finite backhaul capacity. Under the two-users and two-BSs scenario and considering that the BSs have the ability of decoding,the users’ messages are di-vided to independent decoding messages in the BSs and joint decoding messages in the CP. The transimit signals are generated by superposed coding scheme in the user ter-minal. The central precosser solve the opitimized problem of maximizing the sum rate of the user rates in the system based on the known channel state information, then the optimized power allocation is feedback to the users. The research results shows that the proposed cooperative decoding scheme can effectively improve the utilization of the backhaul capacity. Further, the BS decoding messages are divided to private message and common message based on the same message and the different message between the users respectively. The ability of the BSs decoding can be improved further owing to introduce the private message and the common message, thus the backhaul capacity can be utilized more effectively. In addition, the conventional method of obtaining the achievable rate region is simple exhaustive search, while the new method is proposed to obtain the boundary of the achievable rate region, which is our main concern. Moreover, the complexity of our proposed method is lower than that of exhaustive search.2) Owing to the finite backhaul capacity, the BSs need compress the signals firstly then forward to the central processor for joint decoding. For the quantization scheme in the BSs, we study the relationship between the quantization noise power and the transmit power of users, and two cases are considered:users have the SINR requirement and users have no SINR requirement. When the user have no SINR requirement, joint optimization of the users’transmit power and the quantization noise power to maxmize the weighted sum rate of the users’rate. And a algorithm is proposed to solve the non-convex opitmized problem and obtain the global optimal solution. While the users have no SINR requirement, there exsits an extra constrant coupling the transmit power and the quantization noise power. By analyzing the relationship, the upper bound and lower bound of the quantization noise power are derived to satisfy the users’SINR requirement.The analyses and results give some guidence when deploying the multicell processing technology in practice.3) The performance of joint detection in the multicell cooperative system is ana-lyzed under the constraint of backhual capacity. The performance is analyzed for the scenario of finite number of cooperative BSs and the scenario of infinite number of co-operative BSs. The detection matrix of joint detection is dervied under the constraint backhual links. Although the performance comparisons of linear detections are the same as the conventional system, the user rate can not increase with the increasing of the trans-mit power due to the constraint of finite backhual capacity. The rate will saturate rapidly at the higher SNR region. Furthermore, the deployment of so large-scale antenna arrays at one BS may be not feasible in practice, these motivate us to analyze the performance of the large-scale cooperative system with linear detectors under constraint backhaul. The performance of large-scale cooperative cells to resist the small-scale fading for dif-ferent linear detection schemes are analyzed.Different from the cooperation of finite BSs, the salutary effect of large-scale cooperation is that the effect of uncorrelated re-ceiver noise is eliminated completely and the fast fading will converge to a fixed value in distribution. Moreover, the analyzed results are very closed to the results by numeri-cal simulation. Thus, the analyzed results can be used as reference for furture research, which reduce the complexity of numerical simulitions.