1
00:00:23,100 --> 00:00:25,000
The 100 meter radio telescope 

2
00:00:25,100 --> 00:00:27,000
in Effelsberg, Germany.

3
00:00:27,100 --> 00:00:30,000
As one of the largest telescope dishes on Earth, 

4
00:00:30,100 --> 00:00:34,000
it is capable of detecting radio waves that have traveled

5
00:00:34,100 --> 00:00:40,000
billions of years, from the farthest quasars and galaxies.

6
00:00:45,100 --> 00:00:49,000
The larger the collecting area a radio telescope has, 

7
00:00:49,100 --> 00:00:52,000
the more sensitive it is to signals from the cosmos, 

8
00:00:52,100 --> 00:01:01,000
enabling observations of fainter sources.

9
00:01:01,100 --> 00:01:05,000
Greater telescope size also results in sharper images 

10
00:01:05,100 --> 00:01:08,000
with better resolution.

11
00:01:09,100 --> 00:01:11,000
"Obviously there's a limit to the 

12
00:01:11,100 --> 00:01:13,000
size of the telescope we can build. 

13
00:01:13,100 --> 00:01:15,000
Fortunately we have a trick for that.

14
00:01:15,100 --> 00:01:28,000
It's called interferometry."

15
00:01:28,100 --> 00:01:32,000
Using a technique called interferometry, 

16
00:01:32,100 --> 00:01:36,000
multiple telescopes simulate a single, larger telescope.

17
00:01:36,100 --> 00:01:42,000
They observe the same region of sky at the same time, gathering data.

18
00:01:45,100 --> 00:01:50,000
This data is sent to a super-computer called a correlator.

19
00:01:51,100 --> 00:01:55,000
The correlator synchronizes the data for every possible pair 

20
00:01:55,100 --> 00:01:59,000
of telescopes, and combines them all to create a single image.

21
00:02:01,100 --> 00:02:06,000
Interferometry can also be done on a scale of thousands of kilometers.

22
00:02:06,100 --> 00:02:12,000
This is called Very Long Baseline Interferometry, or VLBI.

23
00:02:14,100 --> 00:02:18,000
The European VLBI Network is a collaboration of 

24
00:02:18,100 --> 00:02:22,000
major radio astronomical institutes in Europe, 

25
00:02:22,100 --> 00:02:25,000
as well as Asia and Africa.

26
00:02:27,100 --> 00:02:31,000
With access to 27 telescopes in 13 countries, 

27
00:02:31,100 --> 00:02:36,000
the EVN is capable of simulating a telescope up to 10,000 km 

28
00:02:36,100 --> 00:02:40,000
in width - nearly as big as the face of the Earth.

29
00:02:42,100 --> 00:02:46,000
With so many stations at such great distances, 

30
00:02:46,100 --> 00:02:50,000
the EVN can produce images with better resolution 

31
00:02:50,100 --> 00:02:53,000
than the best optical telescopes.

32
00:03:09,100 --> 00:03:11,000
"VLBI gives us fantastic resolution. 

33
00:03:11,100 --> 00:03:16,000
In fact the resolution is so good that even in astronomical objects 

34
00:03:16,100 --> 00:03:19,000
we can see things move. That is something that is hardly possible 

35
00:03:19,100 --> 00:03:21,000
with any other technique."

36
00:03:21,100 --> 00:03:24,000
As with smaller interferometric arrays, 

37
00:03:24,100 --> 00:03:29,000
VLBI data is sent from the telescopes to a correlator.

38
00:03:33,100 --> 00:03:36,000
The EVN correlator is located here, 

39
00:03:36,100 --> 00:03:41,000
at the Joint Institute for VLBI in Europe, or JIVE, in the Netherlands.

40
00:03:51,100 --> 00:03:56,000
Specially built for this purpose, the correlator is continually 

41
00:03:56,100 --> 00:03:59,000
improved to ensure compatibility with upgrades made 

42
00:03:59,100 --> 00:04:02,000
at the various telescopes.

43
00:04:03,100 --> 00:04:06,000
Correlator upgrades also enable cutting edge capabilities 

44
00:04:06,100 --> 00:04:11,000
to support new astronomical questions, such as when and how 

45
00:04:11,100 --> 00:04:14,000
the first galaxies were formed.

46
00:04:15,100 --> 00:04:17,000
In traditional VLBI, 

47
00:04:17,100 --> 00:04:20,000
data is sent from the telescopes to the correlator 

48
00:04:20,100 --> 00:04:23,000
on magnetic tape or disks.

49
00:04:25,100 --> 00:04:29,000
Shipping disks here from around the world is costly.

50
00:04:29,100 --> 00:04:33,000
They can be damaged or lost in transit, or can be found to have 

51
00:04:33,100 --> 00:04:35,000
errors in the data.

52
00:04:35,100 --> 00:04:39,000
Moreover, there is a significant delay from the time of the 

53
00:04:39,100 --> 00:04:44,000
observation to the receipt of processed data by the astronomer.

54
00:04:44,100 --> 00:04:49,000
In addition, scheduling VLBI observations typically requires 

55
00:04:49,100 --> 00:04:53,000
significant advance coordination among the many institutes that 

56
00:04:53,100 --> 00:04:55,000
operate the telescopes.

57
00:04:55,100 --> 00:04:59,000
This limits opportunity for unscheduled observations when 

58
00:04:59,100 --> 00:05:03,000
transient events are detected, such as supernovae and other 

59
00:05:03,100 --> 00:05:05,000
violent cosmic phenomena.

60
00:05:05,100 --> 00:05:08,000
Until now.

61
00:05:11,100 --> 00:05:17,000
In recent years, JIVE has made numerous updates to the EVN correlator.

62
00:05:17,100 --> 00:05:20,000
The most significant allow for data to be streamed over 

63
00:05:20,100 --> 00:05:24,000
fiber optic networks directly to the correlator.

64
00:05:24,100 --> 00:05:27,000
"This technique is called real-time,

65
00:05:27,100 --> 00:05:31,000
or electronic VLBI, or e-VLBI for short.

66
00:05:31,100 --> 00:05:35,000
And we use this technique to stream the data from the telescopes 

67
00:05:35,100 --> 00:05:40,000
directly to the correlator, where it's correlated in real-time."

68
00:05:40,100 --> 00:05:44,000
With such fast data delivery and processing, problems at the 

69
00:05:44,100 --> 00:05:48,000
telescope stations can be detected and corrected immediately, 

70
00:05:48,100 --> 00:05:51,000
saving valuable data.

71
00:05:51,100 --> 00:05:56,000
Astronomers receive their data in a matter of hours, rather than weeks.

72
00:05:56,100 --> 00:06:00,000
Furthermore, procedural changes make it possible to accommodate 

73
00:06:00,100 --> 00:06:05,000
observations on short notice, when transient activity is detected.

74
00:06:05,100 --> 00:06:07,000
"Some of the most interesting 

75
00:06:07,100 --> 00:06:09,000
astrophysical phenomena are 

76
00:06:09,100 --> 00:06:11,000
transient, and many of these are 

77
00:06:11,100 --> 00:06:13,000
located at great distances from us.

78
00:06:13,100 --> 00:06:17,000
To study emission from objects like collapsing stars or 

79
00:06:17,100 --> 00:06:21,000
accreting black holes, radio astronomers have to arrange 

80
00:06:21,100 --> 00:06:25,000
observations with the largest telescopes on a very short notice.

81
00:06:25,100 --> 00:06:29,000
Real-time e-VLBI at the EVN makes this much easier than 

82
00:06:29,100 --> 00:06:33,000
before, allowing us to observe many more transient sources.

83
00:06:33,100 --> 00:06:37,000
In the past years, we have been able to investigate a number 

84
00:06:37,100 --> 00:06:41,000
of supernovae, which are exploding stars, microquasars that have 

85
00:06:41,100 --> 00:06:45,000
highly collimated, relativistic outflows, or even a star that 

86
00:06:45,100 --> 00:06:48,000
has been ripped apart by a supermassive black hole."

87
00:06:48,100 --> 00:06:52,000
These upgrades were made possible through the EXPReS project.

88
00:06:52,100 --> 00:06:56,000
Coordinated by JIVE and funded by the European Union, 

89
00:06:56,100 --> 00:07:00,000
EXPReS upgraded the correlator for real-time processing.

90
00:07:00,100 --> 00:07:04,000
It also completed much of the "last mile" networking 

91
00:07:04,100 --> 00:07:06,000
to connect the telescopes.

92
00:07:06,100 --> 00:07:10,000
"Data from telescopes all over the world is streaming into a 

93
00:07:10,100 --> 00:07:15,000
hardware supercomputer which is at the core of our VLBI operations. 

94
00:07:15,100 --> 00:07:19,000
And what you see on the screen here is the total data rate 

95
00:07:19,100 --> 00:07:21,000
that has been coming into..."(fades)

96
00:07:21,100 --> 00:07:26,000
Now, e-VLBI is a regular service offered by the EVN. 

97
00:07:26,100 --> 00:07:31,000
e-VLBI observations have already led to scientific discoveries 

98
00:07:31,100 --> 00:07:35,000
that have improved our under-standing of how galaxies form 

99
00:07:35,100 --> 00:07:37,000
and evolve.

100
00:07:41,100 --> 00:07:45,000
Improvements in the past few years have made e-VLBI

101
00:07:45,100 --> 00:07:48,000
as good as traditional VLBI

102
00:07:48,100 --> 00:07:51,000
in terms of bandwidth and image resolution.

103
00:07:51,100 --> 00:07:56,000
Today JIVE is working to make it even better.

104
00:07:57,100 --> 00:08:01,000
Major efforts at JIVE are focused on a new correlator platform,

105
00:08:01,100 --> 00:08:06,000
required for increasing data rates planned by the EVN. 

106
00:08:07,100 --> 00:08:12,000
JIVE and EVN engineers are using the latest electronics to deliver 

107
00:08:12,100 --> 00:08:17,000
the processing power for future radio astronomy observations.

108
00:08:25,100 --> 00:08:31,000
In 2010, JIVE started coordination of the NEXPReS project. 

109
00:08:31,100 --> 00:08:35,000
With 14 partner institutes, JIVE is combining the 

110
00:08:35,100 --> 00:08:40,000
best of both worlds: the speed and flexibility of e-VLBI, 

111
00:08:40,100 --> 00:08:46,000
with the robustness and reliability of disk-based, recorded VLBI.

112
00:08:47,100 --> 00:08:52,000
One main activity of the project addresses network disruptions 

113
00:08:52,100 --> 00:08:57,000
to make sure data transport progresses smoothly.

114
00:08:57,100 --> 00:09:00,000
Data integrity is protected by buffering the data 

115
00:09:00,100 --> 00:09:03,000
at the telescopes and the correlator.

116
00:09:03,100 --> 00:09:07,000
Related upgrades will improve on-the-fly monitoring and 

117
00:09:07,100 --> 00:09:12,000
modification. This will enable automated correlation.

118
00:09:13,100 --> 00:09:17,000
Another NEXPReS activity is to make more efficient use of the 

119
00:09:17,100 --> 00:09:21,000
networks with "bandwidth on demand".

120
00:09:21,100 --> 00:09:24,000
"Right now we have static lightpaths for e-VLBI, 

121
00:09:24,100 --> 00:09:26,000
which are point-to-point connections.

122
00:09:26,100 --> 00:09:28,000
They're there all the time, whether you use them or not.

123
00:09:28,100 --> 00:09:31,000
So this is not a very efficient use of resources.

124
00:09:31,100 --> 00:09:34,000
So we are developing bandwidth on demand, which makes it possible 

125
00:09:34,100 --> 00:09:38,000
to dynamically allocate bandwidth at the times that you need it, 

126
00:09:38,100 --> 00:09:41,000
and not at the times that you don't."

127
00:09:41,100 --> 00:09:46,000
Moreover, dedicated lightpaths are limited to 1 gigabit per second.

128
00:09:46,100 --> 00:09:51,000
The EVN is already planning 4 and 10 gigabit per second upgrades.

129
00:09:52,100 --> 00:09:57,000
A third aspect of the project addresses the limited speed 

130
00:09:57,100 --> 00:10:01,000
of the existing centralized, hardware-based correlator.

131
00:10:01,100 --> 00:10:05,000
In recent years, JIVE has developed a software correlator, 

132
00:10:05,100 --> 00:10:08,000
the basis for an automated, distributed correlator.

133
00:10:08,100 --> 00:10:11,000
By leveraging  network and computing resources 

134
00:10:11,100 --> 00:10:15,000
that already exist in the EVN, NEXPReS will enable additional 

135
00:10:15,100 --> 00:10:20,000
VLBI observations with minimal impact on scarce resources.

136
00:10:21,100 --> 00:10:26,000
The last major challenge to combining traditional and e-VLBI 

137
00:10:26,100 --> 00:10:29,000
is in storing and archiving the tremendous amount data 

138
00:10:29,100 --> 00:10:31,000
generated in an observation.

139
00:10:31,100 --> 00:10:38,000
A typical EVN experiment generates 12 gigabits of data per second.

140
00:10:38,100 --> 00:10:42,000
That's about the same as one high-quality, feature-length movie 

141
00:10:42,100 --> 00:10:45,000
every second for 12 hours.

142
00:10:47,100 --> 00:10:51,000
NEXPReS is developing high-bandwidth, high-capacity 

143
00:10:51,100 --> 00:10:55,000
networked storage on demand to accommodate this data.

144
00:10:56,100 --> 00:11:00,000
Radio astronomers around the world will benefit from these 

145
00:11:00,100 --> 00:11:04,000
developments in next-generation e-VLBI.

146
00:11:08,100 --> 00:11:12,000
Furthermore, new developments of the EVN technologies 

147
00:11:12,100 --> 00:11:16,000
advanced by JIVE are essential for new telescopes.

148
00:11:16,100 --> 00:11:20,000
The high level of automation and distributed architecture 

149
00:11:20,100 --> 00:11:24,000
are also feeding into design of the Square Kilometre Array.

150
00:11:24,100 --> 00:11:27,000
Additionally, these developments are crucial for various 

151
00:11:27,100 --> 00:11:31,000
planetary science missions such as ESA's Venus Express, 

152
00:11:31,100 --> 00:11:36,000
and space-based VLBI telescopes such as RadioAstron.

153
00:11:37,100 --> 00:11:41,000
Work by JIVE is pushing the envelope of Europe's ICT 

154
00:11:41,100 --> 00:11:43,000
infrastructure.

155
00:11:43,100 --> 00:11:46,000
"It has been demonstrated that fundamental science has a 

156
00:11:46,100 --> 00:11:49,000
healthy impact on the viability of our economy.

157
00:11:49,100 --> 00:11:52,000
And it's not hard to imagine how the techniques that we deploy 

158
00:11:52,100 --> 00:11:58,000
in e-VLBI will influence other sciences, education, communication 

159
00:11:58,100 --> 00:12:00,000
and even entertainment."

160
00:12:02,100 --> 00:12:07,000
JIVE, forging a path to the future of Radio Astronomy.