1. Outline of paper
(1) Introduction
(2) Theoretical analysis
(3) Empirical results
(4) Conclusions
2. Developing method for each paragraph
（1） the predominant development method used in this paragraph is example
Many results have been published on the use of antenna diversity for temporal fade reduction assuming diverse channel models. The Ricean/Rayleigh case is treated for example in [1]–[4].
A treatment of the outage probability for combiners in Weibull fading channels was recently presented in [5], and the analysis of an EGC in correlated Nakagami-m channels can be found in [6].
（2） the predominant development method used in this paragraph is induction
Using empirical data, we conclude that space diversity effectively reduces temporal fluctuation
even for correlated branch signals.
（3） the predominant development method used in this paragraph is definition
We consider the often used model of exponentially correlated branch processes [4], [7] and characterize fi (t) as:
where {ni (t)}i=1....M is a sequence of independent zero-mean Gaussian random processes of variance σ2 and α is the temporal correlation coefficient between neighboring branches.
（4） the predominant development method used in this paragraph is example
For instance, a K-factor greater than 17dB was measured at 90% of locations for a 1 antenna
system, as shown in Fig. 1b. At the same time, this K-factor results in temporal fluctuations not greater then 4.5dB 99% of the time as shown in Fig. 1a.
For example, for UB=10dB Ki max must exceed 21dB for uncorrelated branches and 25dB when correlation is 0.8. Measured pairs of UB and Ki max for each location are also plotted on this chart for array sizes as indicated.
（5） the predominant development method used in this paragraph is deduction
We observed that for typical urban fixed wireless scenarios, the use of a 4 branch EGC combiner significantly reduces the likelihood of high temporal fluctuation when compared to a single antenna system placed at random in the same environment. Results of extensive empirical work have shown for example, that the K-factor that is exceeded at 90% of locations increases by about 7dB through the use of a 4 branch combiner with uncorrelated branches. Correspondingly, the temporal fluctuations go down from 4.5dB to 2dB for 99% of time. On the other hand, even for significant temporal correlation between branches, at least one antenna of the array usually will provide a signal stable enough to result in this correlation being of little consequence.
3. Writing style
(1) The use of many synonymous words avoiding repetition.
(2) The avoidance of contractions
(3) The use of tentative forms of expression.
(4) The use of precise words
(5) First and second person pronouns are instead of passive voice.
4. Important sentential type
(1) For mobile links the signal envelopes are usually described by Ricean or Rayleigh random processes which become independent for antenna separations in the order of one wavelength or more
(2) Temporal fades due to moving scatterers will exist for any location but will generally be much weaker than spatial variations (branch unbalance), even for non-line-of-sight conditions.
(3) This contrasts with the typical approach for mobile channels, where changes in position can
generate much deeper temporal fades and where performance is quantified through reduction in outage probability rather than by channel stability.
(4) The antenna output statistics were found to be very well described by Ricean distributions with K-factors large enough for the Gaussian approximation to result in negligible errors.
(5) On the other hand, even for significant temporal correlation between branches, at least one antenna of the array usually will provide a signal stable enough to result in this correlation being of little consequence.
(6) This, combined with the coherencetime [8], provides information useful to characterize the requirements of signal amplitude tracking.
5. Important words
Temporal correlation 时域相关性
temporal fading 时域衰减
fixed wireless links 固定无线链路
equal gain combiner 等增益合并器
spatial diversity 空域分集
diverse channel models 多信道模型
outage probability 损耗概率
temporal fluctuations 时域波动
Gaussian random processes 高斯随机过程
6．Idiomatic expressions
in the order of（固定搭配）
assumption is well justified（动词与副词）
at the same position （介词与名词）
In contrast （固定搭配）
This corresponds to （动词与介词）
empirical data （形容词与名词）
but will generally be much weaker（动词与副词）
even for non-line-of-sight conditions（介词与名词）
with relatively high K-factors （副词与形容词）
are normally small （动词与副词）
will typically either have or not have（动词与副词）
7. Articles and Tenses
Articles:
The Ricean/Rayleigh case
the analysis of an EGC
a single spatial sample
an M-branch
the typical approach
Since a combiner
Tense:
many results have been published
a FWL will typically either have or not have
we conclude that
Each branch can thus be modeled as
were found to be
8. Reference method
The Ricean/Rayleigh case is treated for example in [1]–[4].
A treatment of the outage probability for combiners in Weibull fading channels was recently presented in [5], and the analysis of an EGC in correlated Nakagami-m channels can be found in [6].
Thus, the often made assumption of i.i.d. envelopes is well justified in mobile radio [1], [3]
even for non-line-of-sight conditions. These temporal fluctuations are typically Ricean with relatively high K-factors [8]–[10].
From (5) we can find the statistics of temporal fluctuation for the combiner output.
Using (5) we can define the regions of acceptable branch K-factors
and that the assumption σ2 = σ2 is well justified, as reported in [10].
9.Gain
From the learning of this paper , I have known some problems should be paid attention to and this will be beneficial to the composition of my graduation thesis .To sum up ,there are some points：Firstly, we should use noun phrases to replace subordinate clauses and make greater conciseness through the avoidance of repetition. Secondly, first and second person pronouns should be instead of passive voice，in order to present the article’s objectivity .Thirdly, more tentative forms of expression should be used rather than affirmative. Finally, I realize some intuitive figures making us better understanding and let the paper more interesting and vivid. I will improve my skill of English academic writing through reading large numbers of thesis and exercising how to use words, phrase and sentences.
论文：
IEEE COMMUNICATIONS LETTERS, VOL. 11, NO. 2, FEBRUARY 2007 129
Temporal Fade Reduction for Fixed Wireless Diversity Receivers with
Unbalanced and Correlated Branches
Reinaldo A. Valenzuela, Fellow, IEEE, Luciano Ahumada, Student Member, IEEE,
Rodolfo Feick, Senior Member, IEEE, and Sven Gysling
Abstract— Temporal correlation between diversity branches
degrades the fade mitigation expected in a multi-antenna diver- sity receiver. We study the case when the receive power is unequal across the diversity array. We also include the effect of correlation among the received signals. The results are of particular interest when the diversity receiver is stationary because the antennas will be unbalanced since they experience a single instantiation of the spatial fade statistics. We present a simple analytical model applicable to fixed wireless links (FWLs). Experimental results illustrate the improvement in temporal link stability resulting from the use of an equal gain combiner (EGC).
Index Terms— Temporal correlation, temporal fading, fixed wireless links, equal gain combiner, spatial diversity.