Cold topics in networking
12-16
☆─────────────────────────────────────☆
freedmind (张开思想翅膀) 于 (Wed Dec 17 19:08:56 2008) 提到:
Jon Crowcroft
The Computer Laboratory, University of Cambridge
Cambridge, UK
jon.crowcroft@cl.cam.ac.uk
DHT and Structured P2P
I think we can safely say that we have all seen enough distributed hash
tables to last a lifetime of meals of pork luncheon meat and spam. But m
ore seriously, DHTs are now prime time, not only in use for super-peer o
rganisation in operational file-sharing systems, including several comme
r-
cial P2P-TV systems, but even for controlling really illegal things like
the Storm Botnet.
Internet Coordinate Systems
We know where the Internet is, and we don’t need a compass to find it.
More importantly, you cannot create an absolute euclidean coordinate sys
tem for the Internet since it isn’t embeddable except approximately in
nine dimensions. Thus systems like Vivaldi and Meridian are perfectly fi
ne for finding the best (lowest latency) server out of a list, but do no
t tell you where to aim that cruise missile if you want to take out a pa
rticular data center. Triangle Inequality Violations pervade the Interne
t and no amount of trouble tickets will remove them.
Faster packet classification
Too many people are working on making packets go through routers faster
- this is a marginal business for sure, since any router algorithm for h
eader processing that doesn’t get packets through at line rate is reall
y wasting the capacity (and power - see above) and therefore is going to
get rejected for business case reasons.
BGP
The Border Gateway Protocol should be shut down by the border police. It
was probably cold when it was created. We have several proposals for re
placements, and indeed an entire body of work telling us what is wrong.
We even have a meta-replacement. Now is the time to deploy.
DoS
Denial of service is undefined since IP provides only best effort. How c
an we deny that?
Spam
Spam is only something old fogeys get as they are the only people still
trying to use e-mail instead of some chat system or messaging on a socia
l network.
Overlays
Overlays were only invented as a disruptive idea in the “Looking over t
he Fence” report to make people think about doing things differently -
that didn’t mean one should do overlays (or DTN) seriously. No, really.
TINA
The Intelligent Network Architecture is coming back as we see people str
uggling to justify the GENI costs, and proposing massively complex in-ne
twork smarts. Perhaps one might list the Knowledge Plane here as one of
life’s less successful gedanken experiments.
TCP+AQM
Last time I looked there were over 5000 papers on TCP and AQM published.
This cannot be sensible. Many of the papers used a version of NS TCP co
de that was buggy, so it is probable that the authors should all go back
and check, and have an entirely new conference and journal on the topic
of “Retractions of Erroneous TCPAQM papers”.
Multicast
No-one uses it much so we should declare it dead. Watch this space, thou
gh, as I believe multicast is about to become hot again any decade now,
since several large IPTV deployments now use it, and people are starting
to compare them with P2P-TV (see above).
Newarch
New architectures will never get deployed (for sound business reasons),
so we should declare network architecture a cold topic.
Self similarity, long range dependence, large deviations
The papers on this topic are starting to look remarkably like papers a f
ew years ago on this topic, which means the temperature is falling fast,
from one of my heuristics. Indeed, self-similarity of papers (and the u
se of temperature and entropy as metrics) is a sure fire indication of f
rost.
MANET
Mobile Ad Hoc Networking is a topic with a great future behind it. Now w
e have Mesh, and Metro WiFi, and 3G cellular in most of the world, why w
ould anyone deploy a network that doesn’t work very well? Actually, thi
s reminds me - there may be another heuristic - when papers say how much
better they are than previous work that had achieved 45% connectivity,
when they achieve 53% connectivity, and fail to look at other systems th
at achieve 99.999% connectivity, you know you are approaching zero Kelvi
n in this area. I might
want to include handovers here too.
Self Organising Mobile Wireless Sensor Networks
Most sensors are not mobile. Most sensors sit on a device
and can be configured. Ask an oil man if they would use a biologically i
nspired algorithm to detect they are about to have a well go haywire.
Small World Networks
This is a scam perpetrated by physicists to get publications in Nature o
n the AS Topology of the Internet, the structure of the web and the natu
re of human society, as well as being an albatross around the neck of ec
ologists modelling birdflight, and bee swarms. We know there are really
only three physicists in the world, so it isn’t surprising that the wor
st case connected network has two of them at most six armslength apart,
unless they get around a table, or all get
asked to the same fondue party. It is diametrically opposite random way-
point and Poisson point processes in the unit plane with Voronoi diagram
s, as a source of endless cute visualisations telling us nothing terribl
y surprising, and only rarely even useful.
☆─────────────────────────────────────☆
xbull (xbull) 于 (Wed Dec 17 19:52:55 2008) 提到:
真不幸...
掉进cold里面了
不知道现在真正hot和方向的什么？特别是Internet领域，刚随意翻了翻看这位大牛的t
alks，FIND/GENI搞得火热的 network virtualization 也被他给否了。
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: Jon Crowcroft
: The Computer Laboratory, University of Cambridge
: Cambridge, UK
: ...................
☆─────────────────────────────────────☆
freedmind (张开思想翅膀) 于 (Wed Dec 17 20:46:17 2008) 提到:
全文
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: : : : ...................
☆─────────────────────────────────────☆
freedmind (张开思想翅膀) 于 (Wed Dec 17 20:47:04 2008) 提到:
唉~~~
这个超大牛，否了一批东西啊。
比如smthvv热推的P2P
【 在 xbull (xbull) 的大作中提到: 】
: 真不幸...
: 掉进cold里面了
: 不知道现在真正hot和方向的什么？特别是Internet领域，刚随意翻了翻看这位大牛的t
: ...................
☆─────────────────────────────────────☆
xbull (xbull) 于 (Wed Dec 17 20:48:53 2008) 提到:
zz
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: 唉~~~
: 这个超大牛，否了一批东西啊。
热推的P2P
: ...................
☆─────────────────────────────────────☆
freedmind (张开思想翅膀) 于 (Wed Dec 17 20:59:21 2008) 提到:
真不知道Internet如何变革？
但我认为纯Clean-Slate方法一定不能实现。
【 在 xbull (xbull) 的大作中提到: 】
: 真不幸...
: 掉进cold里面了
: 不知道现在真正hot和方向的什么？特别是Internet领域，刚随意翻了翻看这位大牛的t
: ...................
☆─────────────────────────────────────☆
smthvv (享受工作&认真生活) 于 (Wed Dec 17 22:40:14 2008) 提到:
我没有热推哦,搞清楚再说..虽然我很看好这个技术,有可能是有次给人推荐了吧,记不清了.
对我来说,主要是把手上的事情做好,对你贴出来的像上面一些东西实在是没兴趣,也许是身不由己吧.再说理想和水平也比较低,也总是顾不上有哪个大牛又冒出来说上一句啥.
要说计算机网络里,像active network是绝对的cold了,可是很多思想被引入到了新的一些网络形态中(比如WSN节点重编程等等),所以核心思想我想有些是不会cold的.
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: 唉~~~
: 这个超大牛，否了一批东西啊。
: 比如smthvv热推的P2P
: ...................
☆─────────────────────────────────────☆
dmesg2 (小葱拌豆腐) 于 (Wed Dec 17 23:12:59 2008) 提到:
嗯，好多CCR的文章读起来都很高屋建瓴
还有个10 papers系列来着？
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: Jon Crowcroft
: The Computer Laboratory, University of Cambridge
: Cambridge, UK
: ...................
☆─────────────────────────────────────☆
justicezyx (谈笑总关情) 于 (Wed Dec 17 23:24:05 2008) 提到:
p2p死掉是因为工业应用已经很成熟，学校能做的工作不多。
其他他提到的问题也都是这样，研究重点已经转移到工业界了。
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: 唉~~~
: 这个超大牛，否了一批东西啊。
: 比如smthvv热推的P2P
☆─────────────────────────────────────☆
xbull (xbull) 于 (Thu Dec 18 07:50:23 2008) 提到:
跟大牛们不一致的想法确实不怎么好...
再说，谁又不是在把手上的事情做好呢？ 人家帖子里提到某某，是因为觉得某某值得被提到；要是不愿意被扬名，可以站内bz把帖子给删掉就可以了；不感兴趣，不Re就能体现出来了，不是非得要叫出来。这样，也就不会影响版面的讨论气氛了
我觉得贴些大牛的言论和观点上来讨论挺好的。
【 在 smthvv (享受工作&认真生活) 的大作中提到: 】
: 我没有热推哦,搞清楚再说..虽然我很看好这个技术,有可能是有次给人推荐了吧,记不清了.
: 对我来说,主要是把手上的事情做好,对你贴出来的像上面一些东西实在是没兴趣,也许是身不由己吧.再说理想和水平也比较低,也总是顾不上有哪个大牛又冒出来说上一句啥.
: 要说计算机网络里,像active network是绝对的cold了,可是很多思想被引入到了新的一些网络形态中(比如WSN节点重编程等等),所以核心思想我想有些是不会cold的.
: ...................
☆─────────────────────────────────────☆
smthvv (享受工作&认真生活) 于 (Thu Dec 18 09:32:08 2008) 提到:
【 在 xbull (xbull) 的大作中提到: 】
: 跟大牛们不一致的想法确实不怎么好...
是啊,愁呢,这可咋办呀...
: 再说，谁又不是在把手上的事情做好呢？ 人家帖子里提到某某，是因为觉得某某值得被提到；要是不愿意被扬名，可以站内bz把帖子给删掉就可以了；不感兴趣，不Re就能体现出来了，不是非得要叫出来。这样，也就不会影响版面的讨论气氛了
"叫"是啥意思呢? 我觉得被扬名了吗? 我也是在他提到我说过P2P之后,发贴发表我的看法而已,影响谁了吗? 我所说的只是说了我自己的情况和看法,也不存在对lz的诋毁,也没说不让谁说话来"影响版面的讨论气氛了".
说话要想着讲,你的话实在让人很反感,我很明确的告诉你.
: 我觉得贴些大牛的言论和观点上来讨论挺好的。
是啊,没错.
☆─────────────────────────────────────☆
asceticism (Networker) 于 (Thu Dec 18 12:49:16 2008) 提到:
Such a sarcastic way of looking at research can only come from those big sho
ts. We are not at the level, so don't be fooled.
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: Jon Crowcroft
: The Computer Laboratory, University of Cambridge
: Cambridge, UK
: ...................
☆─────────────────────────────────────☆
vivianite (排骨精，冤大头，小短腿，怕老婆，笨到家) 于 (Thu Dec 18 22:41:06 2008) 提到:
Hey, don't treat this as a guideline, that's just bits and pieces of ideas!
the internet has been a product of engineers who always tie themselves up with problems right ahead -- otherwise they don't care
if you solve a problem that is happening, good;
if you solve a problem that will happen in 6 months or 1 year (when your paper's gonna be published), great, fantastic, keep on doing that please;
if you solve a problem that will happen in 5 years, shit, how do you know what's there after 5 years?
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: Jon Crowcroft
: The Computer Laboratory, University of Cambridge
: Cambridge, UK
: ...................
☆─────────────────────────────────────☆
paolo81 (ooooo) 于 (Fri Dec 19 12:06:38 2008) 提到:
Gallager's LDPC is re-discovered almost 30 years later.
【 在 vivianite (排骨精，冤大头，小短腿，怕老婆，笨到家) 的大作中提到: 】
: Hey, don't treat this as a guideline, that's just bits and pieces of ideas!
: the internet has been a product of engineers who always tie themselves up with problems right ahead -- otherwise they don't care
: if you solve a problem that is happening, good;
: ...................
☆─────────────────────────────────────☆
vivianite (排骨精，冤大头，小短腿，怕老婆，笨到家) 于 (Sat Dec 20 01:20:09 2008) 提到:
to deal with new problems people frequently pick up old solutions some ten years ago from waste recyclers. some people like to be the real initator, but I'd rather be the one who pick it up -- different views. :)
【 在 paolo81 (ooooo) 的大作中提到: 】
: Gallager's LDPC is re-discovered almost 30 years later.
☆─────────────────────────────────────☆
Banyan (Banyan) 于 (Thu Jan 15 18:05:28 2009) 提到:
呵呵，有几个就是Prof. Crowcroft前几年搞的呀
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: Jon Crowcroft
: The Computer Laboratory, University of Cambridge
: Cambridge, UK
: ...................
freedmind (张开思想翅膀) 于 (Wed Dec 17 19:08:56 2008) 提到:
Jon Crowcroft
The Computer Laboratory, University of Cambridge
Cambridge, UK
jon.crowcroft@cl.cam.ac.uk
DHT and Structured P2P
I think we can safely say that we have all seen enough distributed hash
tables to last a lifetime of meals of pork luncheon meat and spam. But m
ore seriously, DHTs are now prime time, not only in use for super-peer o
rganisation in operational file-sharing systems, including several comme
r-
cial P2P-TV systems, but even for controlling really illegal things like
the Storm Botnet.
Internet Coordinate Systems
We know where the Internet is, and we don’t need a compass to find it.
More importantly, you cannot create an absolute euclidean coordinate sys
tem for the Internet since it isn’t embeddable except approximately in
nine dimensions. Thus systems like Vivaldi and Meridian are perfectly fi
ne for finding the best (lowest latency) server out of a list, but do no
t tell you where to aim that cruise missile if you want to take out a pa
rticular data center. Triangle Inequality Violations pervade the Interne
t and no amount of trouble tickets will remove them.
Faster packet classification
Too many people are working on making packets go through routers faster
- this is a marginal business for sure, since any router algorithm for h
eader processing that doesn’t get packets through at line rate is reall
y wasting the capacity (and power - see above) and therefore is going to
get rejected for business case reasons.
BGP
The Border Gateway Protocol should be shut down by the border police. It
was probably cold when it was created. We have several proposals for re
placements, and indeed an entire body of work telling us what is wrong.
We even have a meta-replacement. Now is the time to deploy.
DoS
Denial of service is undefined since IP provides only best effort. How c
an we deny that?
Spam
Spam is only something old fogeys get as they are the only people still
trying to use e-mail instead of some chat system or messaging on a socia
l network.
Overlays
Overlays were only invented as a disruptive idea in the “Looking over t
he Fence” report to make people think about doing things differently -
that didn’t mean one should do overlays (or DTN) seriously. No, really.
TINA
The Intelligent Network Architecture is coming back as we see people str
uggling to justify the GENI costs, and proposing massively complex in-ne
twork smarts. Perhaps one might list the Knowledge Plane here as one of
life’s less successful gedanken experiments.
TCP+AQM
Last time I looked there were over 5000 papers on TCP and AQM published.
This cannot be sensible. Many of the papers used a version of NS TCP co
de that was buggy, so it is probable that the authors should all go back
and check, and have an entirely new conference and journal on the topic
of “Retractions of Erroneous TCPAQM papers”.
Multicast
No-one uses it much so we should declare it dead. Watch this space, thou
gh, as I believe multicast is about to become hot again any decade now,
since several large IPTV deployments now use it, and people are starting
to compare them with P2P-TV (see above).
Newarch
New architectures will never get deployed (for sound business reasons),
so we should declare network architecture a cold topic.
Self similarity, long range dependence, large deviations
The papers on this topic are starting to look remarkably like papers a f
ew years ago on this topic, which means the temperature is falling fast,
from one of my heuristics. Indeed, self-similarity of papers (and the u
se of temperature and entropy as metrics) is a sure fire indication of f
rost.
MANET
Mobile Ad Hoc Networking is a topic with a great future behind it. Now w
e have Mesh, and Metro WiFi, and 3G cellular in most of the world, why w
ould anyone deploy a network that doesn’t work very well? Actually, thi
s reminds me - there may be another heuristic - when papers say how much
better they are than previous work that had achieved 45% connectivity,
when they achieve 53% connectivity, and fail to look at other systems th
at achieve 99.999% connectivity, you know you are approaching zero Kelvi
n in this area. I might
want to include handovers here too.
Self Organising Mobile Wireless Sensor Networks
Most sensors are not mobile. Most sensors sit on a device
and can be configured. Ask an oil man if they would use a biologically i
nspired algorithm to detect they are about to have a well go haywire.
Small World Networks
This is a scam perpetrated by physicists to get publications in Nature o
n the AS Topology of the Internet, the structure of the web and the natu
re of human society, as well as being an albatross around the neck of ec
ologists modelling birdflight, and bee swarms. We know there are really
only three physicists in the world, so it isn’t surprising that the wor
st case connected network has two of them at most six armslength apart,
unless they get around a table, or all get
asked to the same fondue party. It is diametrically opposite random way-
point and Poisson point processes in the unit plane with Voronoi diagram
s, as a source of endless cute visualisations telling us nothing terribl
y surprising, and only rarely even useful.
☆─────────────────────────────────────☆
xbull (xbull) 于 (Wed Dec 17 19:52:55 2008) 提到:
真不幸...
掉进cold里面了
不知道现在真正hot和方向的什么？特别是Internet领域，刚随意翻了翻看这位大牛的t
alks，FIND/GENI搞得火热的 network virtualization 也被他给否了。
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: Jon Crowcroft
: The Computer Laboratory, University of Cambridge
: Cambridge, UK
: ...................
☆─────────────────────────────────────☆
freedmind (张开思想翅膀) 于 (Wed Dec 17 20:46:17 2008) 提到:
全文
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: : : : ...................
☆─────────────────────────────────────☆
freedmind (张开思想翅膀) 于 (Wed Dec 17 20:47:04 2008) 提到:
唉~~~
这个超大牛，否了一批东西啊。
比如smthvv热推的P2P
【 在 xbull (xbull) 的大作中提到: 】
: 真不幸...
: 掉进cold里面了
: 不知道现在真正hot和方向的什么？特别是Internet领域，刚随意翻了翻看这位大牛的t
: ...................
☆─────────────────────────────────────☆
xbull (xbull) 于 (Wed Dec 17 20:48:53 2008) 提到:
zz
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: 唉~~~
: 这个超大牛，否了一批东西啊。
热推的P2P
: ...................
☆─────────────────────────────────────☆
freedmind (张开思想翅膀) 于 (Wed Dec 17 20:59:21 2008) 提到:
真不知道Internet如何变革？
但我认为纯Clean-Slate方法一定不能实现。
【 在 xbull (xbull) 的大作中提到: 】
: 真不幸...
: 掉进cold里面了
: 不知道现在真正hot和方向的什么？特别是Internet领域，刚随意翻了翻看这位大牛的t
: ...................
☆─────────────────────────────────────☆
smthvv (享受工作&认真生活) 于 (Wed Dec 17 22:40:14 2008) 提到:
我没有热推哦,搞清楚再说..虽然我很看好这个技术,有可能是有次给人推荐了吧,记不清了.
对我来说,主要是把手上的事情做好,对你贴出来的像上面一些东西实在是没兴趣,也许是身不由己吧.再说理想和水平也比较低,也总是顾不上有哪个大牛又冒出来说上一句啥.
要说计算机网络里,像active network是绝对的cold了,可是很多思想被引入到了新的一些网络形态中(比如WSN节点重编程等等),所以核心思想我想有些是不会cold的.
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: 唉~~~
: 这个超大牛，否了一批东西啊。
: 比如smthvv热推的P2P
: ...................
☆─────────────────────────────────────☆
dmesg2 (小葱拌豆腐) 于 (Wed Dec 17 23:12:59 2008) 提到:
嗯，好多CCR的文章读起来都很高屋建瓴
还有个10 papers系列来着？
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: Jon Crowcroft
: The Computer Laboratory, University of Cambridge
: Cambridge, UK
: ...................
☆─────────────────────────────────────☆
justicezyx (谈笑总关情) 于 (Wed Dec 17 23:24:05 2008) 提到:
p2p死掉是因为工业应用已经很成熟，学校能做的工作不多。
其他他提到的问题也都是这样，研究重点已经转移到工业界了。
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: 唉~~~
: 这个超大牛，否了一批东西啊。
: 比如smthvv热推的P2P
☆─────────────────────────────────────☆
xbull (xbull) 于 (Thu Dec 18 07:50:23 2008) 提到:
跟大牛们不一致的想法确实不怎么好...
再说，谁又不是在把手上的事情做好呢？ 人家帖子里提到某某，是因为觉得某某值得被提到；要是不愿意被扬名，可以站内bz把帖子给删掉就可以了；不感兴趣，不Re就能体现出来了，不是非得要叫出来。这样，也就不会影响版面的讨论气氛了
我觉得贴些大牛的言论和观点上来讨论挺好的。
【 在 smthvv (享受工作&认真生活) 的大作中提到: 】
: 我没有热推哦,搞清楚再说..虽然我很看好这个技术,有可能是有次给人推荐了吧,记不清了.
: 对我来说,主要是把手上的事情做好,对你贴出来的像上面一些东西实在是没兴趣,也许是身不由己吧.再说理想和水平也比较低,也总是顾不上有哪个大牛又冒出来说上一句啥.
: 要说计算机网络里,像active network是绝对的cold了,可是很多思想被引入到了新的一些网络形态中(比如WSN节点重编程等等),所以核心思想我想有些是不会cold的.
: ...................
☆─────────────────────────────────────☆
smthvv (享受工作&认真生活) 于 (Thu Dec 18 09:32:08 2008) 提到:
【 在 xbull (xbull) 的大作中提到: 】
: 跟大牛们不一致的想法确实不怎么好...
是啊,愁呢,这可咋办呀...
: 再说，谁又不是在把手上的事情做好呢？ 人家帖子里提到某某，是因为觉得某某值得被提到；要是不愿意被扬名，可以站内bz把帖子给删掉就可以了；不感兴趣，不Re就能体现出来了，不是非得要叫出来。这样，也就不会影响版面的讨论气氛了
"叫"是啥意思呢? 我觉得被扬名了吗? 我也是在他提到我说过P2P之后,发贴发表我的看法而已,影响谁了吗? 我所说的只是说了我自己的情况和看法,也不存在对lz的诋毁,也没说不让谁说话来"影响版面的讨论气氛了".
说话要想着讲,你的话实在让人很反感,我很明确的告诉你.
: 我觉得贴些大牛的言论和观点上来讨论挺好的。
是啊,没错.
☆─────────────────────────────────────☆
asceticism (Networker) 于 (Thu Dec 18 12:49:16 2008) 提到:
Such a sarcastic way of looking at research can only come from those big sho
ts. We are not at the level, so don't be fooled.
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: Jon Crowcroft
: The Computer Laboratory, University of Cambridge
: Cambridge, UK
: ...................
☆─────────────────────────────────────☆
vivianite (排骨精，冤大头，小短腿，怕老婆，笨到家) 于 (Thu Dec 18 22:41:06 2008) 提到:
Hey, don't treat this as a guideline, that's just bits and pieces of ideas!
the internet has been a product of engineers who always tie themselves up with problems right ahead -- otherwise they don't care
if you solve a problem that is happening, good;
if you solve a problem that will happen in 6 months or 1 year (when your paper's gonna be published), great, fantastic, keep on doing that please;
if you solve a problem that will happen in 5 years, shit, how do you know what's there after 5 years?
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: Jon Crowcroft
: The Computer Laboratory, University of Cambridge
: Cambridge, UK
: ...................
☆─────────────────────────────────────☆
paolo81 (ooooo) 于 (Fri Dec 19 12:06:38 2008) 提到:
Gallager's LDPC is re-discovered almost 30 years later.
【 在 vivianite (排骨精，冤大头，小短腿，怕老婆，笨到家) 的大作中提到: 】
: Hey, don't treat this as a guideline, that's just bits and pieces of ideas!
: the internet has been a product of engineers who always tie themselves up with problems right ahead -- otherwise they don't care
: if you solve a problem that is happening, good;
: ...................
☆─────────────────────────────────────☆
vivianite (排骨精，冤大头，小短腿，怕老婆，笨到家) 于 (Sat Dec 20 01:20:09 2008) 提到:
to deal with new problems people frequently pick up old solutions some ten years ago from waste recyclers. some people like to be the real initator, but I'd rather be the one who pick it up -- different views. :)
【 在 paolo81 (ooooo) 的大作中提到: 】
: Gallager's LDPC is re-discovered almost 30 years later.
☆─────────────────────────────────────☆
Banyan (Banyan) 于 (Thu Jan 15 18:05:28 2009) 提到:
呵呵，有几个就是Prof. Crowcroft前几年搞的呀
【 在 freedmind (张开思想翅膀) 的大作中提到: 】
: Jon Crowcroft
: The Computer Laboratory, University of Cambridge
: Cambridge, UK
: ...................
相关文章：
- 研究802.11high throughput MAC机制的同行近来讨论一下(05-08)
- 请教OFDM解调的问题(05-08)
- 关于802.11信道实际利用率的问题(05-08)
- 請問一下什么原因會造成16QAM星座圖的扭曲(05-08)
- 传感器网络 VS 网络测量(05-08)
- 预研和无线算法啥区别？(05-08)
射频专业培训教程推荐
