标题: 行动通讯网路之类神经乏晰无线资源配置机制
Radio Resource Allocation Schemes for Mobile Communication Networks Using Neural/Fuzzy Techniques
作者: 陈义升
Yih-Shen Chen
张仲儒
Dr. Chung-Ju Chang
电信工程研究所
关键字: 无线资源配置;行动通讯网路;类神经乏晰网路;radio resource allocation;mobile communication network;neural fuzzy technique
公开日期: 2003
摘要: 为了支援多媒体服务之丛集性传输和异质性服务品质之要求,精确设计无线资源配置机制是必备的,藉以有效提升系统资源使用效率。现有文献已指出,多媒体服务之特性为动态传输和多样化传输需求,因此有必要应用智慧型技术来解决传统无线资源配置问题。在本论文中,我们探讨行动通讯网路中,应用类神经乏晰技术之无线资源配置机制。

我们首先探讨分时多工行动通讯网路之无线资源配置机制。我们采用适应性网路式乏晰推论系统技术,提出一‘乏晰资源配置控制器’。该乏晰控制器为双层式架构,并挑选新连线需求之频宽需求、未来换手连线之频宽保留,以及空中介面效能为输入语意变数,所以,此控制器可最大化行动多媒体网路之统计多工特点所衍生之频宽增益。模拟结果显示,乏晰资源配置控制器可有效降低换手失败机率,但不会大幅增加塞机率。同时,与传统机制相比,此乏晰资源配置控制器可确保服务品质,并且提升系统效能。

接着我们探讨WCDMA系统之多速率传输控制机制。该多速率传输控制问题可模型化为一马可夫决策炼过程,而为了满足服务品质并提升资源使用率之目的,传输成本采用服务品质为参数。我们应用‘Q-learning’学习演算法来精确估算传输成本,因此设计了一‘Q-learning式多速率传输控制’机制。此外,我们也采用特征撷取法和RBFN类神经网路来解决Q函数近似问题,可以降低此机制之状态空间和记体体需求,并且增进Q-learning之收敛特性。模拟结果显示,Q-learning式多速率传输控制机制可提升WCDMA系统之系统容量、用户满意度,同时保证服务品质。

最后我们探讨多细胞WCDMA通讯系统之封包接取控制机制,并且提出一‘乏晰Q-learning式封包接取控制器’。该控制器包含一‘乏晰Q-learning式剩余资源预估器’,和一‘封包速率排程器’。该剩余资源预估器可依据负载状态,精确预估系统剩余资源;它选用自身细胞和相邻细胞之接收干扰量为输入语意变数,并采用感知式协调机制来将多细胞环境简化为单细胞环境。该排程器则采用改良式指数型排程原则,有效地为非及时性用户配置系统资源。模拟结果显示,在同质性和非同质性多细胞WCDMA环境中,该乏晰Q-learning式封包接取控制器可有效降低封包错误率,并提升非及时性讯务之传输速率。
To support bursty-transmission and heterogeneous quality of services (QoS) requirements for multimedia services, a well-designed sophisticated radio resource allocation scheme is required to effectively enhance the system utilization. Research has shown that the non-stationarity of work-loads, together with heterogeneous traffic characteristics and QoS constraints of multimedia services, constitute the necessity for applying intelligent techniques in future mobile multimedia networks. In this dissertation, the radio resource allocation schemes by using neural/fuzzy techniques for mobile communication networks are studied.

Firstly, the radio resource allocation scheme for TDMA-based mobile communication networks is investigated. The adaptive-network-based fuzzy inference system (ANFIS) is applied to propose a fuzzy resource allocation controller (FRAC). The FRAC is designed in a two-layer architecture and properly selects the capacity requirement of new call
request, the capacity reservation for future handoffs, and the air interface performance as input linguistic variables. Therefore, the statistical multiplexing gain of mobile multimedia networks can be maximized in FRAC.
Simulation results indicate that FRAC can keep the handoff call blocking rate low without jeopardizing the new call blocking rate. Also, compared to the conventional schemes,
FRAC can indeed guarantee QoS contracts and achieve higher system performance.

And then, the multi-rate transmission control scheme for WCDMA communication systems is studied. The multi-rate transmission control problem is modelled as a Markov decision process (MDP), where the transmission cost is defined in terms of the QoS parameters for enhancing spectrum utilization subject to QoS constraints. The Q-learning reinforcement algorithm is adopted to accurately estimate the transmission cost and propose a Q-learning-based multi-rate transmission control (Q-MRTC) scheme. In the meanwhile, the feature extraction method and RBFN network are successfully employed for the $Q$-function approximation. The state space and memory storage requirement are then reduced, and the convergence property of $Q$-learning algorithm is improved. Simulation results show that, for a multimedia WCDMA system, the Q-MRTC can achieve higher system throughput and better users' satisfaction while the QoS requirements are guaranteed.

Finally, the data access control scheme for multi-cell WCDMA systems is investigated. By using fuzzy Q-learning technique, a novel situation-aware data access manager (FQ-SDAM) is proposed. The FQ-SDAM contains a fuzzy
Q-learning-based residual capacity estimator (FQ-RCE) and a data rate scheduler (DRS). The FQ-RCE can accurately estimate the situation-dependent residual system capacity;
it appropriately chooses the received interferences from home-cell and adjacent-cell as input linguistic variables and simplifies the multi-cell environment into a single-cell one by applying a perceptual coordination mechanism. Also, the DRS can effectively allocate the resource for
non-real-time terminals by adopting a modified exponential rule which takes the interference influence on adjacent cells into consideration. Simulation results show that
the FQ-SDAM can effectively reduce the packet error probability and improve aggregate throughput of the non-real-time services in both the homogeneous and non-homogeneous multi-cell WCDMA environment.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT008813811
http://hdl.handle.net/11536/57890
显示于类别:Thesis


文件中的档案:

  1. 381101.pdf

If it is a zip file, please download the file and unzip it, then open index.html in a browser to view the full text content.