英语翻译B.Kreless Channel and Link ModelWe consider a discrete t

英语翻译
B.Kreless Channel and Link Model
We consider a discrete time channel with stationary,
ergodic,slowly time-varying gain and additive
white Gaussian noise (AWGN) n ,( k ) ,where the subscript
a refers to the ith link and k refers to the kth time instant.
In this paper,our analysis will be based on static channel
gains.This is justified by the assumption of very slow fading
where the channel coherence time (the time over which the
channel remains roughly constant) is long enough so that the
control system converges to steady-state within a coherence
time interval.We assume that the channel power gain g;(k)
is independent of the channel input and the transmission
power Pi does not change as the channel gain varies.
Different link layer design choices (coding,modulation,
etc.) lead to different performance for the data transfer.We
assume a simple class of communication link designs as
shown in Figure 3.The figure shows the wireless link from
a sensor to a controller.We assume the same link model
for the wireless links including the links from controllers
to the actuators with the input of Ud(k) and the output
of u ( k ) .At the transmitter,the data is first quantized and
converted into a binary bit stream via a uniform quantizer
[2][5].The bit stream then goes through the channel
coder that uses linear block codes for error correction
and error detection.We assume BPSK modulation at the
transmitter.At the receiver,we assume matched filtering
followed by a maximum likelihood detector.The probability
of successful transmission ps for each packet can be easily
calculated given the link design,wireless channel gain and
transmit power.We assume time is slotted and we allow
retransmissions if there are extra time slots.An optional
feedback channel from the receiver to the transmitter may
exist.With such a feedback channel,the receiver sends an
“ACK” to the transmitter upon a successful transmission.
If the transmitter receives the ACK,it clears its transmit
buffer and does not transmit until a new packet arrives.We
assume the transmit buffer only has a capacity of one data
packet.Thus a packet will he discardedZ if it has not been
successfully transmitted by the end of the sample period.
From the control system point of view,the relevant
parameters from the wireless links are data rate,time delay
and probability of packet loss.’ For this purpose,we can
simplify the link model as in Figure 4.This simplied
model is sufficient to calculate all the communication parameters
that may affect the control performance.The data
rate is implicit in the covariance of the quantization noise
vq,i.Both the time delay distribution and the probability
of packet loss are determined by the MAC protocols,total
number of time slots and probability of successful transmission
p,.

乙kreless通道和连接模型
我们认为,离散时间与频道文具,
遍历,时间慢慢地变增益和添加剂
高斯白噪声（ awgn ）氮,（ k ）款,凡标
a指该提取环节和K是指以第k时间瞬间.
在本文中,我们的分析将基于静态频道
收益.这是合理的假设很慢衰落
地方频道相干时间（时间超过其中
频道仍然大体不变） ,是足够长,使该
控制系统收敛到稳定状态在一个连贯性
时间间隔.我们假定了信道功率增益G ; （十一）
是独立的通道输入及传输
电力丕并不能改变作为通道增益各不相同.
不同的链路层设计选择（编码,调制,
等） ,导致不同的表现为数据传输.我们
假设一个简单的类通信链路的设计
如图3所示.这个数字表明,无线连接
一个传感器向控制器中.我们承担同样的连接模型
为无线链路,包括三通,从控制器
该驱动器与输入的丁（ k ）和输出
铀（ k ）款.在发射机,数据是第一次量化,并
转换成二进制码流通过一个统一的量化
[ 2 ] [ 5 ] .该位流,然后通过渠道
编码器采用线性分组码纠错
和错误检测.我们假定BPSK调制在
发射机.在接收端,我们假设匹配滤波
随后进行了最大似然检测器.概率
成功传输的PS为每包可以很容易
计算鉴于连接设计,无线频道增益和
传输功率.我们假定时间是开槽,我们允许
转播,如果有额外的时间槽.可选
信息反馈渠道,从接收到发射机可能
存在的.这样的一个反馈渠道,接收器发送一个
" ack " ,以变送器基于一个成功的传播.
如果发射机接收ack ,它擦亮了,其传输
缓冲区并没有传递到一个新的包抵达.我们
承担传输缓冲区,只有有能力的一个数据
包.因此,一包,将他discardedz如果它没有被
成功地传输到去年底,该样本时期.
从控制系统的角度来看,有关
参数,从无线链路数据速率,时间延迟
和概率丢包' ,为达到此目的,我们可
简化环节模式见图4 .这简化
模型是足够的计算所有的通讯参数
这可能会影响到控制性能.数据
利率是隐含在方差的量化噪声
矢量量化,一无论是时间延迟的分布及概率
对丢包是由MAC协议,总
若干时段和概率的成功传输
磷.