Figure 2-2 Cell clustering:Depiction of a three-cell reuse pattern
The RF spectrum available for the geographic service area is assigned to each cluster, such that cells within a cluster do not share any channel . If M channels make up the entire spectrum available for the service area, and if the distribution of users is uniform over the service area, then each cell is assigned M/N channels. As the clusters are replicated over the service area, the reuse of channels leads to tiers of co-channel cells, and co-channel interference will result from the propagation of RF energy between co-channel base stations and mobile users. Co-channel interference in a cellular system occurs when, for example, a mobile simultaneously receives signals from the base station in its own cell, as well as from co-channel base stations in nearby cells from adjacent tiers. In this instance, one co-channel forward link (base station to mobile transmission) is the desired signal, and the other co-channel signals received by the mobile form the total co-channel interference at the receiver. The power level of the co-channel interference is closely related to the separation distances among co-channel cells. If we model the cells with a hexagonal shape, as in Figure 2-2, the minimum distance between the center of two co-channel cells, called the reuse distance DN, is
DN?3NR （2-1） where R is the maximum radius of the cell (the hexagon is inscribed within the radius).
5
Therefore, we can immediately see from Figure 2-2 that a small cluster size (small reuse distance DN), leads to high interference among co-channel cells.
The level of co-channel interference received within a given cell is also dependent on the number of active co-channel cells at any instant of time. As mentioned before, co-channel cells are grouped into tiers with respect to a particular cell of interest. The number of co-channel cells in a given tier depends on the tier order and the geometry adopted to represent the shape of a cell (e.g., the coverage area of an individual base station). For the classic hexagonal shape, the closest co-channel cells are located in the first tier and there are six co-channel cells. The second tier consists of 12 co-channel cells, the third, 18, and so on. The total co-channel interference is, therefore, the sum of the co-channel interference signals transmitted from all co-channel cells of all tiers. However, co-channel cells belonging to the first tier have a stronger influence on the total interference, since they are closer to the cell where the interference is measured.
Co-channel interference is recognized as one of the major factors that limits the capacity and link quality of a wireless communications system and plays an important role in the tradeoff between system capacity (large-scale system issue) and link quality (small-scale issue). For example, one approach for achieving high capacity (large number of users), without increasing the bandwidth of the RF spectrum allocated to the system, is to reduce the channel reuse distance by reducing the cluster size N of a cellular system . However, reduction in the cluster sizeincreases co-channel interference, which degrades the link quality.
The level of interference within a cellular system at any time is random and must be simulated by modeling both the RF propagation environment between cells and the position location of the mobile users. In addition, the traffic statistics of each user and the type of channel allocation scheme at the base stations determine the instantaneous interference level and the capacity of the system.
The effects of co-channel interference can be estimated by the signal-tointerference ratio (SIR) of the communication link, defined as the ratio of the power of the desired signal S, to the power of the total interference signal, I. Since both power levels S and I are random variables due to RF propagation effects, user mobility and traffic variation, the SIR
6
is also a random variable. Consequently, the severity of the effects of co-channel interference on system performance is frequently analyzed in terms of the system outage probability, defined in this particular case as the probability that SIR is below a given threshold SIR0. This is
Poutpage?Pr[SIR?SIR0]??SIR00pSIR（x)dx （2-2）
Where p SIR(x) is the probability density function (pdf) of the SIR. Note the distinction between the definition of a link outage probability, that classifies an outage based on a particular bit error rate (BER) or Eb/N0 threshold for acceptable voice performance, and the system outage probability that considers a particular SIR threshold for acceptable mobile performance of a typical user.
Analytical approaches for estimating the outage probability in a cellular system, as discussed in before, require tractable models for the RF propagation effects, user mobility, and traffic variation, in order to obtain an expression for . Unfortunately, it is very difficult to use analytical models for these effects, due to their complex relationship to the received signal level. Therefore, the estimation of the outage probability in a cellular system usually relies on simulation, which offers flexibility in the analysis. In this chapter, we present a simple example of a simulation of a cellular communication system, with the emphasis on the system aspects of the communication system, including multi-user performance, traffic engineering, and channel reuse. In order to conduct a system-level simulation, a number of aspects of the individual communication links must be considered. These include the channel model, the antenna radiation pattern, and the relationship between Eb/N0 (e.g., the SIR) and the acceptable performance.
pSIR(x) 7
英文翻译
蜂窝无线通信系统的研究
摘要
蜂窝通信系统允许大量移动用户无缝地、同时地利用有限的射频（radio frequency，RF）频谱与固定基站中的无线调制解调器通信。基站接收每一个移动台发送来的射频信号，并把他们转换到基带或者带宽微波链路，然后传送到移动交换中心（MSC），再由移动交换中心连入公用交换电话网（PSTN）。同样的，通信信号也可以从PSTN传送到基站，再从这里发送个移动台。蜂窝系统可以采用频分多址（FDMA）、时分多址（TDMA）、码分多址（CDMA）或者空分多址（SDMA）中的任何一种技术。
1 概述
人们开发出了许多无线通信系统，为不同的运行环境中的固定用户或移动用户提供了接入到通信基础设施的手段。当今大多数无线通信系统都是基于蜂窝无线电概念之上的。蜂窝通信系统允许大量移动用户无缝地、同时地利用有限的射频（radio frequency，RF）频谱与固定基站中的无线调制解调器通信。基站接收每一个移动台发送来的射频信号，并把他们转换到基带或者带宽微波链路，然后传送到移动交换中心（MSC），再由移动交换中心连入公用交换电话网（PSTN）。同样的，通信信号也可以从PSTN传送到基站，再从这里发送个移动台。蜂窝系统可以采用频分多址（FDMA）、时分多址（TDMA）、码分多址（CDMA）或者空分多址（SDMA）中的任何一种技术。
无线通信链路具有恶劣的物理信道特征，比如由于传播途径中有再大的障碍物，会产生时变多径和阴影。此外，无线蜂窝系统的性能还会受限于来自其他用户的干扰，因此，对干扰进行准确的建模就很重要。很难用简单的解析模型来描述复杂的信道条件，虽然有集中模型确实易于解析求解并与信道实测数据比较相符，不过，即使建立了完美的信道解析模型，再把差错控制编码、均衡器、分集及网络模型等因素都考虑再链路中之后，要得出链路性能的解析在绝大多数情况下任然是很困难的甚至是不可能的。因此，在分析蜂窝通信链路的性能时，常常需要进行仿真。
跟无线链路一样，对蜂窝无线系统的性能分析使用仿真建模时很有效的，这是由于在时间和空间上对大量的随机事件进行建模非常困难。这些随机事件包括用户的位置、系统中同时通信的用户个数、传播条件、每个用户的干扰和功率级的设置（power level setting）、每个用户的话务量需求等，这些因素共同作用，对系统中的一个典
8
型用户的总的性能产生影响。前面提到的变量仅仅是任一时刻决定系统中的某个用户瞬态性能的许多关键物理参数中的一小部分。蜂窝无线系统指的是，在地理上的服务区域内，移动用户和基站的全体，而不是将一个用户连接到一个基站的单个链路。为了设计特定大的系统级性能，比如某个用户在整个系统中得到满意服务的可能性，就得考虑在覆盖区域内同时使用系统的多个用户所带来的复杂性。因此，需要仿真来考虑多个用户对基站和移动台之间任何一条链路所产生的影响。
链路性能是一个小尺度现象，它处理的是小的局部区域内或者短的时间间隔内信道的顺时变化，这种情况下可假设平均接收功率不变。在设计差错控制码、均衡器和其他用来消除信道所产生的瞬时影响的部件时，这种假设时合理的。但是，在大量用户分布在一个广阔的地理范围内时，为了确定整个系统的性能，有必要引入大尺度效应进行分析，比如在大的距离范围内考虑单个用户受到的干扰和信号电平的统计行为时，忽略瞬时信道特征。我们可以将链路级仿真看作通信系统性能的微调，而将系统级仿真看作时整体质量水平粗略但很重要的近似，任何用户在任何时候都可预计达到这个水平。
通过让移动台在不同的服务区内共享或者复用通信信道，蜂窝系统能达到较高的容量（比如，为大量的用户服务）。信道复用会导致公用同一信道的用户之间产生同频干扰，这是影响蜂窝系统容量和性能的主要制约因素之一。因此，在设计一个蜂窝系统时，或者在分析和设计消除同频干扰负面影响的系统方法时，需要正确理解同屏干扰对容量和性能的影响。这些影响主要取决于通信系统的状况，如共享信道的用户数和他们的位置。其他与传播信道条件关系更密切的方面，如路径损耗、阴影衰落（或叫阴影）、天线辐射模式等对系统性能的影响也很重要，因为这些影响也岁特定用户的位置而改变。本章我们将讨论在同频干扰情况下，包括一个典型系统中的天线和传播的影响。尽管本章考虑的例子比较简单，但提出的分析方法可以容易地进行扩展，以包括蜂窝系统的其他特征。
2 蜂窝无线系统
系统级描述：
如图2-1所示，通过把地理区域分成一个个称为小区的部分，蜂窝系统可以在这个区域内提供无线覆盖。把可用的频谱也分成很多信道，每个小区分配一组信道，每个小区中的基站都配备了可以同移动用户进行通信的无线调制解调器。从基站到移动
9