Interference Management And Limitation Of Delay In The Next Generation Of Communication Networks

With the development of IP(Internet Protocol) based next generation of wireless communication, wireless capacity in the physical layer, traffic scheduling and routing in the backbone networks face unprecedented challenges. Thanks to the close attention from academia and industry, large amount of researches and proposals have been proposed to improve the utilization ratio of physical resources in 4G network, e.g,. OFDMA(orthogonal frequency division multiple access), Co MP(coordinated multi-point) and CA(carrier aggregation). These technologies indeed enhance the performance of physical layer and coverage of cell by resources adjustment, however, challenges are simultaneously brought to the interference management, e.g., BS(base station) clustering and user scheduling in Co MP, power allocation in the OFDMA scenario. On the other hand, the rapid increase of amount of IP traffic, especially in the 5G network, mmwave, massive-MIMO(massive multiple-input multiple-output) and ultra-densification are employed to support roughly 1000 ×wireless capacity from 4G to 5G. From the viewpoint of utilization of resources and energy efficiency, the next generation of wireless networks should achieve tradeoff between large scale resources from wired backbone network and wireless physical layer, which calls for a new structure of traffic scheduling, routing and wireless transmitssion in the network with the limitation of delay. To achieve this goal, in this paper, we first study the interference management by Co MP in the physical layer and propose a feasible dynamic Co MP to improve the wireless capacity. Take the wireless capacity into account, we further propose a delay limited packet scheduling, routing and wireless transmission from the viewpoint of system energy efficieny.From the viewpoint of interference management, we design a distributed dynamic Co MP with strong environment-adaptation. The distributed manner is simple and scalable. With the result of transmission and scheduling in Co MP, we further executes WF(water-filling) algorithm to get the initial power allocation which is dealt as the strategy space of power in the power game. Thereafter, we further design a framework of Nash non-cooperative game to adjust the initial power allocation. We prove the existence and uniqueness of the NE(Nash equilibrium) and give a search algorithm to find the NE. By simulation, we verify that the performance of Co MP can indeed be improved by our algorithms. The distributed scalable design makes our algorithm more preferable in the engineering.Although the high value in engineering of the heuristic algorithms is alluring, more focus from the theorem is on the optimization model to achieve joint solution to transmission, scheduling and power allocation in Co MP. Most of the existing algorithms are heuristic philosophy due to both discrete and successive variables involved in the optimization model. In this paper, to solve the problem above, we first propose a novel PFM mechanism(power-fractionizing mechanis) to change the discrete variables into successive variables by some limitations to an SP(signomial programming) optimization model. Based on GP(geometric programming), we give the joint local optimization solution in Co MP, empirically, the local optimization solution is also the globle optimization solution. By simulation, we verify the effectiveness of our algorithm.Besides, as metioned before, how to achieve energy efficient tradeoff between costs of “large scale” resources as well as the tradeoff between limitation of delay in wired routing, and wireless transmission in the next generation of wireless networks, especially in the 5G network is another challenge. Since the large amount of IP packets in the next generation of wireless networks, we adopted WDM(wavelength division multiplexing) as the wired connection between BSs. By configurate multiple tunable lasers at each node in the networks, we can transmit multiple packets from one BS simultaneously.With a view to the characteristics of “large scale” and “multiple parameters”, we propose the objective of joint optimization of energy efficiency, where the energy efficiency from wired connection to wireless transmission is integrated, i.e., energy efficiency from the standpoint of “wireless-wired-wireless”. Define the maximal delay in system as limitation of delay, by allocation of limitation of delay between wired exchange and wireless transmission, as well as the tradeoff between resources from both wired and wireless parts, we propose a delay limited high energy efficient integrated algorithm regarding traffic scheduling, routing in WDM and wireless transmission. The proposed algorithm can improve energy efficiency from the viewpoint of “wireless-wired-wireless” according to the bound of delay. By simulation, we verify the effectiveness of our algorithms.With different objectives of optimization in the networks, we employ Nash non-cooperative game, geometric programming and convex optimization, e.t.c., to achieve improve the performance of the next generation of wireless networks. All the algorithms fit for the existing framework of network. Moreover, all the algorithms we proposed in this paper can also be adopted by other networks for achieving better interference management and delay limited energy efficieny from the standpoint of “wireless-wired-wireless”.