Research On Cross-layer Resource Allocation In Next Generation Mobile Communication System

In the next generation mobile communication system, time-varying character of radio links brings great challenges to broadband data services. With increasing demands of users and higher performance of user equipments, more and more video applications come to our daily life. Because of their inherent characteristics, video applications put forward strict requirements to network bandwidth and time delay. Wireless network broadband becomes a trend apparently, which makes diversification of mobile applications. However, different applications require different classes of QoS, which raises more requirements to the implementation of protocol layers. In order to respond to these challenges, cross-layer resource allocation has been employed to break the independence of different protocol layers, and restrain the negative impact brought by time-varying wireless channel. Cross-layer design can obtain multi-layer diversity gain, and improve the performance of resource allocation strategy. However, with great performance gain, cross-layer design, especially for video applications, brings enormous computation complexity to cross-layer optimizer, which asks for innovation of resource allocation strategies to achieve the SON design.In this paper, different layer parameters were adjusted dynamically to achieve the SON model, and efficient cross-layer resource allocation schemes were designed and implemented. A cross-layer scheduling scheme for video streaming based on fuzzy decision-making was proposed, and a QoS driven prediction function for video applications was modeled. The architecture and interface for inter-cell coordination were investigated, which were explored to restrain inter-cell interference in OFDM based networks. The main work of this paper includes four aspects.We proposed the theoretical model of cross-layer potential games. First of all, construction methods of potential functions for common cross-layer potential games were discussed. Then we analyzed the convergence of game iteration, as well as the existence and uniqueness of Nash Equilibrium. The finite improvement property of potential games was proved, as well as the stability of equilibrium. At last, the optimality of cross-layer potential games was analyzed and the indicator for optimality was presented. The theory of cross-layer potential games brought distributed framework for cross-layer design, and kept its low complexity with increasing network scale. On the other hand, with applicable convergence criterion cross-layer potential games could converge to the optimal solution as soon as possible, and cross-layer parameters could be allocated dynamically, which satisfies the requirements of SON for next generation mobile communication network.We modeled and implemented the simulation platform for cross-layer resource allocation. In terms of modular design, the platform was divided into four modules, such as application level simulation module, system level simulation module, link level simulation module and cross-layer optimizer. Different modules of platform worked in a asynchronous manner, so as to supply the real time character for video applications. As visualization interface was adopted in this platform, we could compare the original video with user received video in real time. Novel technologies and algorithms in research could be estimated quickly and intuitively, which curtailed the development cycle of new technologies, and cut down the development cost. Meanwhile, the platform helped researcheres to estimate different cross-layer algorithms in a visible way.We proposed novel SON based cross-layer resource allocation for video applications. Multi-objective fuzzy decision-making was adopted to model the cross-layer optimization problem for video applications, and interaction between different objectives was investigate sufficiently. Then based on analysis on characteristics of video content and video streaming, QoS driven prediction model was proposed, which predicted end-to-end video distortion exactly according to block error rate in transmission. At last, in order to reducing complexity of cross-layer optimization problem, cross-layer potential games framework was used to model the video scheduling problem, and distributed resource allocation was implemented. Gradient projection criterion was adopted as the convergence criterion of players in game iteration, in order to converge to optimal solution in a parallel way. SON based cross-layer resource allocation for video applications was a channel adaption method in nature. The concept of SON and cross-layer design were combined, cross-layer resource can be allocated adaptively with the variation of radio link. Cross-layer potential games framework was adopted to avoid the huge complexity caused by cross-layer design for video applications. Cross-layer resource allocation with low complexity and rapid convergence were implemented, while the convergence time would increase in linear with the quantity of game players. In addition, SON based cross-layer method allocated wireless resource according to users perceived video quality.We proposed positive cross-layer inter-cell interference coordination strategies. By means of coordination between neighbor cells, transmitted powers in subchannels were adjusted dynamically, and the wireless channel condition was improved, such that throughputs of the whole cell were increased and the power consumption of network was decreased, and the positive resource allocation scheme was proved to be effective. On this condition, positive cross-layer resource allocation scheme was proposed, transmitted power and scheduled user of subchannels with same frequency in neighbor cells were coordinated, and multi-layer diversity gain was obtained by wireless network, as well as the fairness between different users. Cross-layer potential games theory was used to model the cross-layer resource allocation problem, and PSO method was used to find the best respond strategies for players in each iteration of potential game, such that distributed problem solving could be achieved. Positive cross-layer inter-cell interference coordination scheme, by adjusting strategies in physical layer and MAC layer jointly, came up to restraining inter-cell interference in OFDM based network, reducing power consumption of network and increasing throughputs of mobile users. Meanwhile, with distributed architecture of algorithms, computational complexity brought by cross-layer design was impaired, and rapid convergence of interference coordination process was achieved.