无线网络接入控制策略研究

发布时间：2018-04-04 05:23

本文选题：机械类业务　切入点：随机接入策略　出处：《北京邮电大学》2014年博士论文

【摘要】：无线接入技术的不断更新,新的业务不断涌现,促使无线通信系统飞速发展。总体来说,其发展蕴含了单系统演进和多系统融合。单系统演进包含无线通信系统的平滑过渡或者革新,多系统融合则是网络异构化。无论是单系统演进还是多系统融合,无线资源管理都在其中扮演着重要角色,用户接入是无线通信系统中实现无线资源管理的基础,良好的随机接入控制策略则是实现用户成功接入的保障。本文从随机接入的角度出发,针对单体制系统,重点考虑OFDMA系统,针对多系统,重点考虑异构融合系统。首先简单分析介绍了这两种系统,并进一步研究了这两种系统下随机接入控制策略的研究现状,然后介绍了部分常用算法,最后分别针对OFDMA系统、异构系统以及Machine-to-Machine (M2M)业务提出了结合各自特征的随机接入控制策略,本文的主要内容及贡献包括：针对OFDMA系统,分别提出了一种基于伪贝叶斯算法的接入策略和退避策略,这两种策略充分考虑OFDMA系统的特点并基于TDD模式提出。接入策略包含自适应分配随机接入信道资源策略和自适应接入概率值策略两部分。自适应分配策略根据系统的负载状态动态配置随机接入信道的数目,自适应接入概率策略则基于自适应分配随机接入信道资源策略的结果来调整终端用户接入概率大小,从而提高终端用户接入成功率。退避策略同样是根据系统的负载状态,采用快速接入与动态退避相结合的机制,通过预估的终端用户数目与随机接入信道数目相比,从而选择出合适的退避策略以及相应的退避时间,有效的降低了终端用户的平均时延。针对M2M业务提出了一种异构环境下的分流接入控制策略。本策略突破单一网络的传输能力限制,结合M2M业务海量、群组化等业务特征,将M2M业务分为多个业务子流在不同网络中传输。在场景建模上,首先设置接入点,由接入点承担着M2M业务的汇聚以及接入多个网络的功能。同时引入经济学概念,考虑收益和成本两个方面,然后利用排队论模型进行分析,建立效用函数模型。并将此模型化为具有等式和不等式约束的非线性规划问题,采用拉格朗日常数法和KKT法进行问题求解,所得结果即为不同网络的分流结果,并证明了结果全局最优。该策略充分考虑了M2M业务特点,使得分配结果效用最大化。针对异构环境下多终端多网络场景,提出了一种基于负载平衡的随机接入策略,该策略由接入选择机制和接入控制机制两个方面构成,接入选择机制主要主要解决用户“如何选择网络”的问题,它将终端用户所关心的多个属性化为空间中的一个维度后,同时采用改进的AHP方法减少主观影响,得到目标向量和实际向量,通过向量间的几何关系得到用户的可选网络集。接入控制机制则解决“用户接入到哪个网络”的问题,网络侧根据终端用户由接入选择机制选出的可接入网络集后,基于负载平衡的原理,将终端用户分配到合适的网络中。该机制充分考虑了终端侧和网络侧的利益,即提高终端用户的QoS,又让网络在负载平衡的基础上尽可能让更多的用户接入。
[Abstract]:Wireless access technology is constantly updated, new business are constantly emerging, promote the rapid development of wireless communication system. In general, its development contains a single system and multi system integration evolution. Single system evolution of the wireless communication system includes a smooth transition or innovation, multi system integration is the network isomerization. Either single or multi system evolution system fusion, wireless resource management will play a key role in the implementation of user access, is the foundation of radio resource management in wireless communication system, a good random access control strategy is to achieve successful user access security.
This article from the random access point, for a single system, focus on the OFDMA system for multi system, focus on the integration of heterogeneous system. Firstly introduces the simple analysis of the two systems, and further study the research status of the random access control strategy of the two systems, and then introduces some commonly used algorithms, finally, respectively. According to the OFDMA system, heterogeneous system and Machine-to-Machine (M2M) business proposed random access control strategy according to their own characteristics, the main contents and contributions of this paper include:
According to the OFDMA system, propose an access strategy of the pseudo Bayesian algorithm and backoff strategy based on these two strategies fully consider the characteristics of the OFDMA system is proposed based on TDD model. The access strategy includes adaptive random access channel resource allocation strategy and adaptive access probability values of two part strategy. Adaptive allocation strategy according to the number of dynamic configuration random access channel load system, adaptive access probability strategy adaptive allocation random access channel resource strategy based on the results of the adjustment of the terminal user access probability, so as to improve the success rate of terminal user access. The same backoff strategy according to the load of the system, the mechanism of using fast access and dynamic backoff combination, the number of terminal the user estimates compared with the number of random access channel, so as to select the appropriate backoff strategy and the corresponding The backoff time effectively reduces the average delay of the end user.
For the M2M business a shunt access in heterogeneous environment. The proposed control strategy to limit the transmission capacity of the network's breakthrough single strategy, combined with the M2M business group of massive, business characteristics, M2M business is divided into a plurality of sub stream transmission business in different networks. In scene modeling, first set the access point by the access point undertakes M2M service convergence and access network function. While introducing the concept of economics, consider two aspects of income and cost, and then use the queuing theory model analysis, establishes the utility function model. And this model into a nonlinear programming problem with equality and inequality constraints, the Lagrange constant and KKT method is used to solve the problem, the result is the dividing results of different networks, and prove the global optimal results. This strategy takes full account of the characteristics of M2M services, the allocation results of utility maximization .
Aiming at the heterogeneous environment of multi terminal network scenarios, proposed a load balance based on random access strategy, the two aspects of the strategy by the access selection mechanism and access control mechanism, access selection mechanism mainly to solve the user how to choose network ", multiple attributes it will concern for end users a dimension in the space, at the same time using improved AHP method to reduce the subjective effect, get the target vector and the actual vector, get optional network users set through the geometric relationship between vectors. The access control mechanism is to solve the" user access to which network ", according to the user selected by the network side terminal access selection mechanism to access the network in principle, based on load balancing, the terminal user is assigned to the appropriate network. This mechanism takes full account of the terminal side and network side interests, namely The QoS of high end users allows the network to allow more users to access as much as possible on the basis of load balance.

【学位授予单位】：北京邮电大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN925.93

【参考文献】