Wireless Throughput Calculations and Limitations

https://documentation.meraki.com/MR/WiFi_Basics_and_Best_Practices/Wireless_Throughput_Calculations_and_Limitations


This article provides details on MR Access Points' advertised maximum data transfer rates and how they are calculated, including: 

  • Limitations and factors affecting throughput
  • Maximizing throughput
  • Testing Client-to-Access Point throughput
  • Cisco Meraki Access Point Throughput Rates

Limitations and Factors Affecting Throughput

First and foremost, throughput to the Internet is capped by the ISP and devices upstream of the APs. Also, 802.11 is a shared medium and is limited by other devices connected to the wireless. Therefore throughput should always be considered aggregate throughput. Interference (radio, physical, electrical) and the distance from client device to the Access Point are two major factors that have a negative impact on observed maximum throughput. Physical obstacles, other wireless networks and even everyday household devices like computers, microwaves, and televisions increase interference significantly, especially on the 2.4GHz band.

 

The half-duplex nature of wireless combined with other overhead also means that the actual aggregate throughput is typically 50 percent or less of the data rate. It is theoretically possible for 802.11n-capable wireless clients to achieve speeds as high as 100Mbps or more depending on the MIMO capabilities of the AP and the wireless client. However, wireless clients operating at 802.11b/g/a can cause 802.11n users to slow to less than 54Mbps because the radio must adjust to the lowest common denominator. 

 

The nature of WiFi technology makes throughput hard to predict. Therefore, network administrators should maintain reasonable expectations for connection speeds and keep the above factors in mind.

Maximizing Throughput

Optimal throughput rates and wireless performance in general can be improved by mounting APs correctly (high up on a wall or on the ceiling). Additionally, placing APs away from kitchens and other areas with high interference will also result in better performance. In a mesh environment, throughput rates will be significantly better with fewer hops to the gateway.

 

Meraki recommends that the end user is located no more than 3 hops away from the gateway. Each hop will reduce the bandwidth by 50%. For example, a 6 Mbps connection to a gateway will reduce to 3 Mbps at the second hop and 1.5 Mbps at the third hop.

Testing Client-to-AP Throughput

When testing the throughput of a Cisco Meraki MR Access Point, it is important to remember that any advertised value is the theoretical total maximum data transfer rate (transmit and receive) for the AP’s radio(s). Device-to-AP speeds should always be tested using a tool like the Access Point to Client Speed Test available on the Local Status page, my.meraki.com, under Client survey tools (and not a tool like speedtest.net). The Access Point to Client Speed Test tool is especially useful when distinguishing between bandwidth constraints on the Meraki network (ex. repeater AP to gateway) and bandwidth constraints of the Internet connection (gateway to Internet).

Cisco Meraki Access Point Throughput Rate Calculations

Please note that the following throughput rates are theoretical for the whole access point, any individual client will be subject to many additional environmental factors that can affect their throughput.

Indoor APs

MR12: Theoretical maximum data transfer rate of 300Mbps 

One 2.4GHz radio

Multiple input, multiple output with two spatial streams

150Mbps x 1 radio x 2 streams = 300Mbps

See Datasheet here

 

MR16/18: Theoretical maximum data transfer rate of 600Mbps

One 2.4GHz radio

One 5GHz radio

Multiple input, multiple output with two spatial streams

150Mbps x 2 radios x 2 streams = 600Mbps

See Datasheet here

 

MR24/26: Theoretical maximum data transfer rate of 900Mbps

One 2.4GHz radio

One 5GHz radio

Multiple input, multiple output with three spatial streams

150Mbps x 2 radios x 3 streams = 900Mbps

See Datasheet here

 

MR32: Theoretical maximum data transfer rate of 1170Mbps

One 2.4GHz radio

One 5GHz radio

Multiple input, multiple output with two spatial streams on 80MHz-wide channels

292.5Mbps x 2 radios x 2 streams = 1.2Gbps

See Datasheet here

 

MR34: Theoretical maximum data transfer rate of 1755 Mbps

One 2.4GHz radio

One 5GHz radio

Multiple input, multiple output with three spatial streams on 80MHz-wide channels

292.5Mbps x 2 radios x 3 streams = 1.75Gbps

See Datasheet here

 

Outdoor APs

MR62: Theoretical maximum data transfer rate of 300Mbps

One 2.4GHz radio

Multiple input, multiple output with two spatial streams

150Mbps x 1 radio x 2 streams = 300Mbps

See Datasheet here

 

MR66: Theoretical maximum data transfer rate of 600Mbps

One 2.4GHz radio

One 5GHz radio

Multiple input, multiple output with two spatial streams

150Mbps x 2 radios x 2 streams = 600Mbps

See Datasheet here

 

MR72: Theoretical maximum data transfer rate of 1170Mbps

One 2.4GHz radio

One 5GHz radio

Multiple input, multiple output with two spatial streams on 80MHz-wide channels

292.5Mbps x 2 radios x 2 streams = 1.2Gbps

See Datasheet here


  • 1
    点赞
  • 0
    收藏
    觉得还不错? 一键收藏
  • 0
    评论
### 支持向量机非线性回归通用MATLAB程序解析 #### 一、概述 本文将详细介绍一个基于MATLAB的支持向量机(SVM)非线性回归的通用程序。该程序采用支持向量机方法来实现数据的非线性回归,并通过不同的核函数设置来适应不同类型的数据分布。此外,该程序还提供了数据预处理的方法,使得用户能够更加方便地应用此程序解决实际问题。 #### 二、核心功能与原理 ##### 1. 支持向量机(SVM) 支持向量机是一种监督学习模型,主要用于分类和回归分析。对于非线性回归任务,SVM通过引入核技巧(kernel trick)将原始低维空间中的非线性问题转换为高维空间中的线性问题,从而实现有效的非线性建模。 ##### 2. 核函数 核函数的选择直接影响到模型的性能。本程序内置了三种常用的核函数: - **线性核函数**:`K(x, y) = x'y` - **多项式核函数**:`K(x, y) = (x'y + 1)^d` - **径向基函数(RBF)**:`K(x, y) = exp(-γ|x - y|^2)` 其中RBF核函数被广泛应用于非线性问题中,因为它可以处理非常复杂的非线性关系。本程序默认使用的是RBF核函数,参数`D`用于控制高斯核函数的宽度。 ##### 3. 数据预处理 虽然程序本身没有直接涉及数据预处理的过程,但在实际应用中,对数据进行适当的预处理是非常重要的。常见的预处理步骤包括归一化、缺失值处理等。 ##### 4. 模型参数 - **Epsilon**: ε-insensitive loss function的ε值,控制回归带宽。 - **C**: 松弛变量的惩罚系数,控制模型复杂度与过拟合的风险之间的平衡。 #### 三、程序实现细节 ##### 1. 函数输入与输出 - **输入**: - `X`: 输入特征矩阵,维度为(n, l),其中n是特征数量,l是样本数量。 - `Y`: 目标值向量,长度为l。 - `Epsilon`: 回归带宽。 - `C`: 松弛变量的惩罚系数。 - `D`: RBF核函数的参数。 - **输出**: - `Alpha1`: 正的拉格朗日乘子向量。 - `Alpha2`: 负的拉格朗日乘子向量。 - `Alpha`: 拉格朗日乘子向量。 - `Flag`: 标记向量,表示每个样本的类型。 - `B`: 偏置项。 ##### 2. 核心代码解析 程序首先计算所有样本间的核矩阵`K`,然后构建二次规划问题并求解得到拉格朗日乘子向量。根据拉格朗日乘子的值确定支持向量,并计算偏置项`B`。 - **核矩阵计算**:采用RBF核函数,通过`exp(-(sum((xi-xj).^2)/D))`计算任意两个样本之间的相似度。 - **二次规划**:构建目标函数和约束条件,使用`quadprog`函数求解最小化问题。 - **支持向量识别**:根据拉格朗日乘子的大小判断每个样本是否为支持向量,并据此计算偏置项`B`。 #### 四、程序扩展与优化 - **多核函数支持**:可以通过增加更多的核函数选项,提高程序的灵活性。 - **自动调参**:实现参数自动选择的功能,例如通过交叉验证选择最优的`Epsilon`和`C`值。 - **并行计算**:利用MATLAB的并行计算工具箱加速计算过程,特别是当样本量很大时。 #### 五、应用场景 该程序适用于需要进行非线性回归预测的场景,如经济预测、天气预报等领域。通过调整核函数和参数,可以有效应对各种类型的非线性问题。 ### 总结 本程序提供了一个支持向量机非线性回归的完整实现框架,通过灵活的核函数设置和参数调整,能够有效地处理非线性问题。对于需要进行回归预测的应用场景,这是一个非常实用且强大的工具。
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值