1 <?xml version=
"1.0" encoding=
"utf-8"?>
2 <rss version='
2.0' xmlns:lj='http://www.livejournal.org/rss/lj/
1.0/'
>
4 <title>Petter Reinholdtsen - Entries tagged mesh network
</title>
5 <description>Entries tagged mesh network
</description>
6 <link>http://www.hungry.com/~pere/blog/
</link>
10 <title>Dokumentaren om Datalagringsdirektivet sendes endelig på NRK
</title>
11 <link>http://www.hungry.com/~pere/blog/Dokumentaren_om_Datalagringsdirektivet_sendes_endelig_p__NRK.html
</link>
12 <guid isPermaLink=
"true">http://www.hungry.com/~pere/blog/Dokumentaren_om_Datalagringsdirektivet_sendes_endelig_p__NRK.html
</guid>
13 <pubDate>Wed,
26 Mar
2014 09:
50:
00 +
0100</pubDate>
14 <description><p
><a href=
"http://www.nuug.no/
">Foreningen NUUG
</a
> melder i natt at
15 NRK nå har bestemt seg for
16 <a href=
"http://www.nuug.no/news/NRK_viser_filmen_om_Datalagringsdirektivet_f_rste_gang_2014_03_31.shtml
">når
17 den norske dokumentarfilmen om datalagringsdirektivet skal
18 sendes
</a
> (se
<a href=
"http://www.imdb.com/title/tt2832844/
">IMDB
</a
>
19 for detaljer om filmen) . Første visning blir på NRK2 mandag
20 2014-
03-
31 kl.
19:
50, og deretter visninger onsdag
2014-
04-
02
21 kl.
12:
30, fredag
2014-
04-
04 kl.
19:
40 og søndag
2014-
04-
06 kl.
15:
10.
22 Jeg har sett dokumentaren, og jeg anbefaler enhver å se den selv. Som
23 oppvarming mens vi venter anbefaler jeg Bjørn Stærks kronikk i
24 Aftenposten fra i går,
25 <a href=
"http://www.aftenposten.no/meninger/kronikker/Autoritar-gjokunge-
7514915.html
">Autoritær
26 gjøkunge
</a
>, der han gir en grei skisse av hvor ille det står til med
27 retten til privatliv og beskyttelsen av demokrati i Norge og resten
28 verden, og helt riktig slår fast at det er vi i databransjen som
29 sitter med nøkkelen til å gjøre noe med dette. Jeg har involvert meg
30 i prosjektene
<a href=
"http://www.dugnadsnett.no/
">dugnadsnett.no
</a
>
31 og
<a href=
"https://wiki.debian.org/FreedomBox
">FreedomBox
</a
> for å
32 forsøke å gjøre litt selv for å bedre situasjonen, men det er mye
33 hardt arbeid fra mange flere enn meg som gjenstår før vi kan sies å ha
34 gjenopprettet balansen.
</p
>
36 <p
>Jeg regner med at nettutgaven dukker opp på
37 <a href=
"http://tv.nrk.no/program/koid75005313/tema-dine-digitale-spor-datalagringsdirektivet
">NRKs
38 side om filmen om datalagringsdirektivet
</a
> om fem dager. Hold et
39 øye med siden, og tips venner og slekt om at de også bør se den.
</p
>
44 <title>Dugnadsnett for alle stiller på Oslo Maker Faire i januar
2014</title>
45 <link>http://www.hungry.com/~pere/blog/Dugnadsnett_for_alle_stiller_p__Oslo_Maker_Faire_i_januar_2014.html
</link>
46 <guid isPermaLink=
"true">http://www.hungry.com/~pere/blog/Dugnadsnett_for_alle_stiller_p__Oslo_Maker_Faire_i_januar_2014.html
</guid>
47 <pubDate>Tue,
10 Dec
2013 19:
20:
00 +
0100</pubDate>
48 <description><p
>Helga
18. og
19. januar
2014 arrangeres
49 <a href=
"http://makerfaireoslo.no/no/program/dugnadsnett
">Oslo Maker
50 Faire
</a
>, og
<a href=
"http://www.dugnadsnett.no/
">Dugnadsnett for
51 alle
</a
> har fått plass! Planen er å ha et bord med en plakat der vi
52 forteller om hva Dugnadsnett for alle er for noe, og et lite verksted
53 der vi hjelper folk som er interessert i å få opp sin egen mesh-node.
54 Jeg gleder meg til å se hvordan prosjektet blir mottatt der.
</p
>
56 <p
>Målet med dugnadsnett for alle i Oslo er å få på plass et datanett
57 for kommunikasjon ved hjelp av radio-repeaterstasjoner (kalt
58 mesh-noder) som gjør at en kan direkte kommunisere med slekt, venner
59 og bekjente i Oslo via andre som deltar i dugnadsnettet, samt gjøre
60 det mulig komme ut på internett via dugnadsnettet. Første delmål er å
61 kunne sende SMS-meldinger vha. IP-telefoni løsningen
62 <a href=
"http://www.servalproject.org/
">Serval project
</a
> mellom
63 deltagerne i Dugnadsnett for alle i Oslo. Formålet er å ta tilbake
64 kontrollen over egen nett-infrastruktur og gjøre det dyrere å bedrive
65 massiv innsamling av informasjon om borgernes bruk av datanett.
</p
>
67 <p
>Høres dette interessant ut? Bli med på prosjektet, fortell oss
68 hvor du kunne tenke deg å sette opp en radio-repeater (slik at folk i
69 nærheten kan finne hverandre ved hjelp av
70 <a href=
"http://flynor.net/mesh/mesh.php
">kartet over planlagte og
71 eksisterende radio-repeatere
</A
>), bli med på epostlisten
72 <a href=
"http://lists.nuug.no/mailman/listinfo/dugnadsnett
">dugnadsnett
73 (at) nuug.no
</a
> og stikk innom
74 <a href=
"irc://irc.freenode.net/#dugnadsnett.no
">IRC-kanalen
75 #dugnadsnett.no
</a
>. Så langt er det planlagt over
40
76 radio-repeatere, med VPN-forbindelser via Internet for å la de delene
77 av nettet som ikke når hverandre via radio kunne snakke med hverandre
83 <title>Dugnadsnett for alle, a wireless community network in Oslo, take shape
</title>
84 <link>http://www.hungry.com/~pere/blog/Dugnadsnett_for_alle__a_wireless_community_network_in_Oslo__take_shape.html
</link>
85 <guid isPermaLink=
"true">http://www.hungry.com/~pere/blog/Dugnadsnett_for_alle__a_wireless_community_network_in_Oslo__take_shape.html
</guid>
86 <pubDate>Sat,
30 Nov
2013 10:
10:
00 +
0100</pubDate>
87 <description><p
>If you want the ability to electronically communicate directly with
88 your neighbors and friends using a network controlled by your peers in
89 stead of centrally controlled by a few corporations, or would like to
90 experiment with interesting network technology, the
91 <a href=
"http://www.dugnadsnett.no/
">Dugnasnett for alle i Oslo
</a
>
92 might be project for you.
39 mesh nodes are currently being planned,
93 in the freshly started initiative from NUUG and Hackeriet to create a
94 wireless community network. The work is inspired by
95 <a href=
"http://freifunk.net/
">Freifunk
</a
>,
96 <a href=
"http://www.awmn.net/
">Athens Wireless Metropolitan
97 Network
</a
>,
<a href=
"http://en.wikipedia.org/wiki/Roofnet
">Roofnet
</a
>
98 and other successful mesh networks around the globe. Two days ago we
99 held a workshop to try to get people started on setting up their own
100 mesh node, and there we decided to create a new mailing list
101 <a href=
"http://lists.nuug.no/mailman/listinfo/dugnadsnett
">dugnadsnett
102 (at) nuug.no
</a
> and IRC channel
103 <a href=
"irc://irc.freenode.net/#dugnadsnett.no
">#dugnadsnett.no
</a
> to
104 coordinate the work. See also the NUUG blog post
105 <a href=
"http://www.nuug.no/news/E_postliste_og_IRC_kanal_for_Dugnadsnett_for_alle_i_Oslo.shtml
">announcing
106 the mailing list and IRC channel
</a
>.
</p
>
111 <title>Lets make a wireless community network in Oslo!
</title>
112 <link>http://www.hungry.com/~pere/blog/Lets_make_a_wireless_community_network_in_Oslo_.html
</link>
113 <guid isPermaLink=
"true">http://www.hungry.com/~pere/blog/Lets_make_a_wireless_community_network_in_Oslo_.html
</guid>
114 <pubDate>Wed,
13 Nov
2013 21:
00:
00 +
0100</pubDate>
115 <description><p
>Today NUUG and Hackeriet announced
116 <a href=
"http://www.nuug.no/news/Bli_med___bygge_dugnadsnett_for_alle_i_Oslo.shtml
">our
117 plans to join forces and create a wireless community network in
118 Oslo
</a
>. The workshop to help people get started will take place
119 Thursday
2013-
11-
28, but we already are collecting the geolocation of
120 people joining forces to make this happen. We have
121 <a href=
"https://github.com/petterreinholdtsen/meshfx-node/blob/master/oslo-nodes.geojson
">9
122 locations plotted on the map
</a
>, but we will need more before we have
123 a connected mesh spread across Oslo. If this sound interesting to
124 you, please join us at the workshop. If you are too impatient to wait
125 15 days, please join us on the IRC channel
126 <a href=
"irc://irc.freenode.net/%
23nuug
">#nuug on irc.freenode.net
</a
>
127 right away. :)
</p
>
132 <title>Running TP-Link MR3040 as a batman-adv mesh node using openwrt
</title>
133 <link>http://www.hungry.com/~pere/blog/Running_TP_Link_MR3040_as_a_batman_adv_mesh_node_using_openwrt.html
</link>
134 <guid isPermaLink=
"true">http://www.hungry.com/~pere/blog/Running_TP_Link_MR3040_as_a_batman_adv_mesh_node_using_openwrt.html
</guid>
135 <pubDate>Sun,
10 Nov
2013 23:
00:
00 +
0100</pubDate>
136 <description><p
>Continuing my research into mesh networking, I was recommended to
137 use TP-Link
3040 and
3600 access points as mesh nodes, and the pair I
138 bought arrived on Friday. Here are my notes on how to set up the
139 MR3040 as a mesh node using
140 <a href=
"http://www.openwrt.org/
">OpenWrt
</a
>.
</p
>
142 <p
>I started by following the instructions on the OpenWRT wiki for
143 <a href=
"http://wiki.openwrt.org/toh/tp-link/tl-mr3040
">TL-MR3040
</a
>,
145 <a href=
"http://downloads.openwrt.org/snapshots/trunk/ar71xx/openwrt-ar71xx-generic-tl-mr3040-v2-squashfs-factory.bin
">the
146 recommended firmware image
</a
>
147 (openwrt-ar71xx-generic-tl-mr3040-v2-squashfs-factory.bin) and
148 uploaded it into the original web interface. The flashing went fine,
149 and the machine was available via telnet on the ethernet port. After
150 logging in and setting the root password, ssh was available and I
151 could start to set it up as a batman-adv mesh node.
</p
>
153 <p
>I started off by reading the instructions from
154 <a href=
"http://wirelessafrica.meraka.org.za/wiki/index.php?title=Antoine
's_Research
">Wireless
155 Africa
</a
>, which had quite a lot of useful information, but
156 eventually I followed the recipe from the Open Mesh wiki for
157 <a href=
"http://www.open-mesh.org/projects/batman-adv/wiki/Batman-adv-openwrt-config
">using
158 batman-adv on OpenWrt
</a
>. A small snag was the fact that the
159 <tt
>opkg install kmod-batman-adv
</tt
> command did not work as it
160 should. The batman-adv kernel module would fail to load because its
161 dependency crc16 was not already loaded. I
162 <a href=
"https://dev.openwrt.org/ticket/
14452">reported the bug
</a
> to
163 the openwrt project and hope it will be fixed soon. But the problem
164 only seem to affect initial testing of batman-adv, as configuration
165 seem to work when booting from scratch.
</p
>
167 <p
>The setup is done using files in /etc/config/. I did not bridge
168 the Ethernet and mesh interfaces this time, to be able to hook up the
169 box on my local network and log into it for configuration updates.
170 The following files were changed and look like this after modifying
173 <p
><tt
>/etc/config/network
</tt
></p
>
177 config interface
'loopback
'
178 option ifname
'lo
'
179 option proto
'static
'
180 option ipaddr
'127.0.0.1'
181 option netmask
'255.0.0.0'
183 config globals
'globals
'
184 option ula_prefix
'fdbf:
4c12:
3fed::/
48'
186 config interface
'lan
'
187 option ifname
'eth0
'
188 option type
'bridge
'
189 option proto
'dhcp
'
190 option ipaddr
'192.168.1.1'
191 option netmask
'255.255.255.0'
192 option hostname
'tl-mr3040
'
193 option ip6assign
'60'
195 config interface
'mesh
'
196 option ifname
'adhoc0
'
197 option mtu
'1528'
198 option proto
'batadv
'
199 option mesh
'bat0
'
202 <p
><tt
>/etc/config/wireless
</tt
></p
>
205 config wifi-device
'radio0
'
206 option type
'mac80211
'
207 option channel
'11'
208 option hwmode
'11ng
'
209 option path
'platform/ar933x_wmac
'
210 option htmode
'HT20
'
211 list ht_capab
'SHORT-GI-
20'
212 list ht_capab
'SHORT-GI-
40'
213 list ht_capab
'RX-STBC1
'
214 list ht_capab
'DSSS_CCK-
40'
215 option disabled
'0'
217 config wifi-iface
'wmesh
'
218 option device
'radio0
'
219 option ifname
'adhoc0
'
220 option network
'mesh
'
221 option encryption
'none
'
222 option mode
'adhoc
'
223 option bssid
'02:BA:
00:
00:
00:
01'
224 option ssid
'meshfx@hackeriet
'
226 <p
><tt
>/etc/config/batman-adv
</tt
></p
>
229 config
'mesh
' 'bat0
'
230 option interfaces
'adhoc0
'
231 option
'aggregated_ogms
'
232 option
'ap_isolation
'
233 option
'bonding
'
234 option
'fragmentation
'
235 option
'gw_bandwidth
'
236 option
'gw_mode
'
237 option
'gw_sel_class
'
238 option
'log_level
'
239 option
'orig_interval
'
240 option
'vis_mode
'
241 option
'bridge_loop_avoidance
'
242 option
'distributed_arp_table
'
243 option
'network_coding
'
244 option
'hop_penalty
'
246 # yet another batX instance
247 # config
'mesh
' 'bat5
'
248 # option
'interfaces
' 'second_mesh
'
251 <p
>The mesh node is now operational. I have yet to test its range,
252 but I hope it is good. I have not yet tested the TP-Link
3600 box
253 still wrapped up in plastic.
</p
>
258 <title>Teaching vmdebootstrap to create Raspberry Pi SD card images
</title>
259 <link>http://www.hungry.com/~pere/blog/Teaching_vmdebootstrap_to_create_Raspberry_Pi_SD_card_images.html
</link>
260 <guid isPermaLink=
"true">http://www.hungry.com/~pere/blog/Teaching_vmdebootstrap_to_create_Raspberry_Pi_SD_card_images.html
</guid>
261 <pubDate>Sun,
27 Oct
2013 17:
00:
00 +
0100</pubDate>
262 <description><p
>The
263 <a href=
"http://packages.qa.debian.org/v/vmdebootstrap.html
">vmdebootstrap
</a
>
264 program is a a very nice system to create virtual machine images. It
265 create a image file, add a partition table, mount it and run
266 debootstrap in the mounted directory to create a Debian system on a
267 stick. Yesterday, I decided to try to teach it how to make images for
268 <a href=
"https://wiki.debian.org/RaspberryPi
">Raspberry Pi
</a
>, as part
269 of a plan to simplify the build system for
270 <a href=
"https://wiki.debian.org/FreedomBox
">the FreedomBox
271 project
</a
>. The FreedomBox project already uses vmdebootstrap for
272 the virtualbox images, but its current build system made multistrap
273 based system for Dreamplug images, and it is lacking support for
274 Raspberry Pi.
</p
>
276 <p
>Armed with the knowledge on how to build
"foreign
" (aka non-native
277 architecture) chroots for Raspberry Pi, I dived into the vmdebootstrap
278 code and adjusted it to be able to build armel images on my amd64
279 Debian laptop. I ended up giving vmdebootstrap five new options,
280 allowing me to replicate the image creation process I use to make
281 <a href=
"https://people.skolelinux.org/pere/blog/A_Raspberry_Pi_based_batman_adv_Mesh_network_node.html
">Debian
282 Jessie based mesh node images for the Raspberry Pi
</a
>. First, the
283 <tt
>--foreign /path/to/binfm_handler
</tt
> option tell vmdebootstrap to
284 call debootstrap with --foreign and to copy the handler into the
285 generated chroot before running the second stage. This allow
286 vmdebootstrap to create armel images on an amd64 host. Next I added
287 two new options
<tt
>--bootsize size
</tt
> and
<tt
>--boottype
288 fstype
</tt
> to teach it to create a separate /boot/ partition with the
289 given file system type, allowing me to create an image with a vfat
290 partition for the /boot/ stuff. I also added a
<tt
>--variant
291 variant
</tt
> option to allow me to create smaller images without the
292 Debian base system packages installed. Finally, I added an option
293 <tt
>--no-extlinux
</tt
> to tell vmdebootstrap to not install extlinux
294 as a boot loader. It is not needed on the Raspberry Pi and probably
295 most other non-x86 architectures. The changes were accepted by the
296 upstream author of vmdebootstrap yesterday and today, and is now
298 <a href=
"http://git.liw.fi/cgi-bin/cgit/cgit.cgi/vmdebootstrap/
">the
299 upstream project page
</a
>.
</p
>
301 <p
>To use it to build a Raspberry Pi image using Debian Jessie, first
302 create a small script (the customize script) to add the non-free
303 binary blob needed to boot the Raspberry Pi and the APT source
308 set -e # Exit on first error
309 rootdir=
"$
1"
310 cd
"$rootdir
"
311 cat
&lt;
&lt;EOF
> etc/apt/sources.list
312 deb http://http.debian.net/debian/ jessie main contrib non-free
314 # Install non-free binary blob needed to boot Raspberry Pi. This
315 # install a kernel somewhere too.
316 wget https://raw.github.com/Hexxeh/rpi-update/master/rpi-update \
317 -O $rootdir/usr/bin/rpi-update
318 chmod a+x $rootdir/usr/bin/rpi-update
319 mkdir -p $rootdir/lib/modules
320 touch $rootdir/boot/start.elf
321 chroot $rootdir rpi-update
322 </pre
></p
>
324 <p
>Next, fetch the latest vmdebootstrap script and call it like this
325 to build the image:
</p
>
328 sudo ./vmdebootstrap \
331 --distribution jessie \
332 --mirror http://http.debian.net/debian \
341 --root-password raspberry \
342 --hostname raspberrypi \
343 --foreign /usr/bin/qemu-arm-static \
344 --customize `pwd`/customize \
348 --package ca-certificates \
351 </pre
></p
>
353 <p
>The list of packages being installed are the ones needed by
354 rpi-update to make the image bootable on the Raspberry Pi, with the
355 exception of netbase, which is needed by debootstrap to find
356 /etc/hosts with the minbase variant. I really wish there was a way to
357 set up an Raspberry Pi using only packages in the Debian archive, but
358 that is not possible as far as I know, because it boots from the GPU
359 using a non-free binary blob.
</p
>
361 <p
>The build host need debootstrap, kpartx and qemu-user-static and
362 probably a few others installed. I have not checked the complete
363 build dependency list.
</p
>
365 <p
>The resulting image will not use the hardware floating point unit
366 on the Raspberry PI, because the armel architecture in Debian is not
367 optimized for that use. So the images created will be a bit slower
368 than
<a href=
"http://www.raspbian.org/
">Raspbian
</a
> based images.
</p
>
373 <title>A Raspberry Pi based batman-adv Mesh network node
</title>
374 <link>http://www.hungry.com/~pere/blog/A_Raspberry_Pi_based_batman_adv_Mesh_network_node.html
</link>
375 <guid isPermaLink=
"true">http://www.hungry.com/~pere/blog/A_Raspberry_Pi_based_batman_adv_Mesh_network_node.html
</guid>
376 <pubDate>Mon,
21 Oct
2013 11:
40:
00 +
0200</pubDate>
377 <description><p
>The last few days I have been experimenting with
378 <a href=
"http://www.open-mesh.org/projects/batman-adv/wiki
">the
379 batman-adv mesh technology
</a
>. I want to gain some experience to see
380 if it will fit
<a href=
"https://wiki.debian.org/FreedomBox
">the
381 Freedombox project
</a
>, and together with my neighbors try to build a
382 mesh network around the park where I live. Batman-adv is a layer
2
383 mesh system (
"ethernet
" in other words), where the mesh network appear
384 as if all the mesh clients are connected to the same switch.
</p
>
386 <p
>My hardware of choice was the Linksys WRT54GL routers I had lying
387 around, but I
've been unable to get them working with batman-adv. So
388 instead, I started playing with a
389 <a href=
"http://www.raspberrypi.org/
">Raspberry Pi
</a
>, and tried to
390 get it working as a mesh node. My idea is to use it to create a mesh
391 node which function as a switch port, where everything connected to
392 the Raspberry Pi ethernet plug is connected (bridged) to the mesh
393 network. This allow me to hook a wifi base station like the Linksys
394 WRT54GL to the mesh by plugging it into a Raspberry Pi, and allow
395 non-mesh clients to hook up to the mesh. This in turn is useful for
396 Android phones using
<a href=
"http://servalproject.org/
">the Serval
397 Project
</a
> voip client, allowing every one around the playground to
398 phone and message each other for free. The reason is that Android
399 phones do not see ad-hoc wifi networks (they are filtered away from
400 the GUI view), and can not join the mesh without being rooted. But if
401 they are connected using a normal wifi base station, they can talk to
402 every client on the local network.
</p
>
404 <p
>To get this working, I
've created a debian package
405 <a href=
"https://github.com/petterreinholdtsen/meshfx-node
">meshfx-node
</a
>
407 <a href=
"https://github.com/petterreinholdtsen/meshfx-node/blob/master/build-rpi-mesh-node
">build-rpi-mesh-node
</a
>
408 to create the Raspberry Pi boot image. I
'm using Debian Jessie (and
409 not Raspbian), to get more control over the packages available.
410 Unfortunately a huge binary blob need to be inserted into the boot
411 image to get it booting, but I
'll ignore that for now. Also, as
412 Debian lack support for the CPU features available in the Raspberry
413 Pi, the system do not use the hardware floating point unit. I hope
414 the routing performance isn
't affected by the lack of hardware FPU
417 <p
>To create an image, run the following with a sudo enabled user
418 after inserting the target SD card into the build machine:
</p
>
421 % wget -O build-rpi-mesh-node \
422 https://raw.github.com/petterreinholdtsen/meshfx-node/master/build-rpi-mesh-node
423 % sudo bash -x ./build-rpi-mesh-node
> build.log
2>&1
424 % dd if=/root/rpi/rpi_basic_jessie_$(date +%Y%m%d).img of=/dev/mmcblk0 bs=
1M
426 </pre
></p
>
428 <p
>Booting with the resulting SD card on a Raspberry PI with a USB
429 wifi card inserted should give you a mesh node. At least it does for
430 me with a the wifi card I am using. The default mesh settings are the
431 ones used by the Oslo mesh project at Hackeriet, as I mentioned in
432 <a href=
"https://people.skolelinux.org/pere/blog/Oslo_community_mesh_network___with_NUUG_and_Hackeriet_at_Hausmania.html
">an
433 earlier blog post about this mesh testing
</a
>.
</p
>
435 <p
>The mesh node was not horribly expensive either. I bought
436 everything over the counter in shops nearby. If I had ordered online
437 from the lowest bidder, the price should be significantly lower:
</p
>
439 <p
><table
>
441 <tr
><th
>Supplier
</th
><th
>Model
</th
><th
>NOK
</th
></tr
>
442 <tr
><td
>Teknikkmagasinet
</td
><td
>Raspberry Pi model B
</td
><td
>349.90</td
></tr
>
443 <tr
><td
>Teknikkmagasinet
</td
><td
>Raspberry Pi type B case
</td
><td
>99.90</td
></tr
>
444 <tr
><td
>Lefdal
</td
><td
>Jensen Air:Link
25150</td
><td
>295.-
</td
></tr
>
445 <tr
><td
>Clas Ohlson
</td
><td
>Kingston
16 GB SD card
</td
><td
>199.-
</td
></tr
>
446 <tr
><td
>Total cost
</td
><td
></td
><td
>943.80</td
></tr
>
448 </table
></p
>
450 <p
>Now my mesh network at home consist of one laptop in the basement
451 connected to my production network, one Raspberry Pi node on the
1th
452 floor that can be seen by my neighbor across the park, and one
453 play-node I use to develop the image building script. And some times
454 I hook up my work horse laptop to the mesh to test it. I look forward
455 to figuring out what kind of latency the batman-adv setup will give,
456 and how much packet loss we will experience around the park. :)
</p
>
461 <title>Oslo community mesh network - with NUUG and Hackeriet at Hausmania
</title>
462 <link>http://www.hungry.com/~pere/blog/Oslo_community_mesh_network___with_NUUG_and_Hackeriet_at_Hausmania.html
</link>
463 <guid isPermaLink=
"true">http://www.hungry.com/~pere/blog/Oslo_community_mesh_network___with_NUUG_and_Hackeriet_at_Hausmania.html
</guid>
464 <pubDate>Fri,
11 Oct
2013 14:
10:
00 +
0200</pubDate>
465 <description><p
>Wireless mesh networks are self organising and self healing
466 networks that can be used to connect computers across small and large
467 areas, depending on the radio technology used. Normal wifi equipment
468 can be used to create home made radio networks, and there are several
469 successful examples like
470 <a href=
"http://www.freifunk.net/
">Freifunk
</a
> and
471 <a href=
"http://www.awmn.net/
">Athens Wireless Metropolitan Network
</a
>
473 <a href=
"http://en.wikipedia.org/wiki/List_of_wireless_community_networks_by_region#Greece
">wikipedia
474 for a large list
</a
>) around the globe. To give you an idea how it
475 work, check out the nice overview of the Kiel Freifunk community which
476 can be seen from their
477 <a href=
"http://freifunk.in-kiel.de/ffmap/nodes.html
">dynamically
478 updated node graph and map
</a
>, where one can see how the mesh nodes
479 automatically handle routing and recover from nodes disappearing.
480 There is also a small community mesh network group in Oslo, Norway,
481 and that is the main topic of this blog post.
</p
>
483 <p
>I
've wanted to check out mesh networks for a while now, and hoped
484 to do it as part of my involvement with the
<a
485 href=
"http://www.nuug.no/
">NUUG member organisation
</a
> community, and
486 my recent involvement in
487 <a href=
"https://wiki.debian.org/FreedomBox
">the Freedombox project
</a
>
488 finally lead me to give mesh networks some priority, as I suspect a
489 Freedombox should use mesh networks to connect neighbours and family
490 when possible, given that most communication between people are
491 between those nearby (as shown for example by research on Facebook
492 communication patterns). It also allow people to communicate without
493 any central hub to tap into for those that want to listen in on the
494 private communication of citizens, which have become more and more
495 important over the years.
</p
>
497 <p
>So far I have only been able to find one group of people in Oslo
498 working on community mesh networks, over at the hack space
499 <a href=
"http://hackeriet.no/
">Hackeriet
</a
> at Husmania. They seem to
500 have started with some Freifunk based effort using OLSR, called
501 <a href=
"http://oslo.freifunk.net/index.php?title=Main_Page
">the Oslo
502 Freifunk project
</a
>, but that effort is now dead and the people
503 behind it have moved on to a batman-adv based system called
504 <a href=
"http://meshfx.org/trac
">meshfx
</a
>. Unfortunately the wiki
505 site for the Oslo Freifunk project is no longer possible to update to
506 reflect this fact, so the old project page can
't be updated to point to
507 the new project. A while back, the people at Hackeriet invited people
508 from the Freifunk community to Oslo to talk about mesh networks. I
509 came across this video where Hans Jørgen Lysglimt interview the
510 speakers about this talk (from
511 <a href=
"https://www.youtube.com/watch?v=N2Kd7CLkhSY
">youtube
</a
>):
</p
>
513 <p
><iframe width=
"420" height=
"315" src=
"https://www.youtube.com/embed/N2Kd7CLkhSY
" frameborder=
"0" allowfullscreen
></iframe
></p
>
515 <p
>I mentioned OLSR and batman-adv, which are mesh routing protocols.
516 There are heaps of different protocols, and I am still struggling to
517 figure out which one would be
"best
" for some definitions of best, but
518 given that the community mesh group in Oslo is so small, I believe it
519 is best to hook up with the existing one instead of trying to create a
520 completely different setup, and thus I have decided to focus on
521 batman-adv for now. It sure help me to know that the very cool
522 <a href=
"http://www.servalproject.org/
">Serval project in Australia
</a
>
523 is using batman-adv as their meshing technology when it create a self
524 organizing and self healing telephony system for disaster areas and
525 less industrialized communities. Check out this cool video presenting
527 <a href=
"https://www.youtube.com/watch?v=
30qNfzJCQOA
">youtube
</a
>):
</p
>
529 <p
><iframe width=
"560" height=
"315" src=
"https://www.youtube.com/embed/
30qNfzJCQOA
" frameborder=
"0" allowfullscreen
></iframe
></p
>
531 <p
>According to the wikipedia page on
532 <a href=
"http://en.wikipedia.org/wiki/Wireless_mesh_network
">Wireless
533 mesh network
</a
> there are around
70 competing schemes for routing
534 packets across mesh networks, and OLSR, B.A.T.M.A.N. and
535 B.A.T.M.A.N. advanced are protocols used by several free software
536 based community mesh networks.
</p
>
538 <p
>The batman-adv protocol is a bit special, as it provide layer
2
539 (as in ethernet ) routing, allowing ipv4 and ipv6 to work on the same
540 network. One way to think about it is that it provide a mesh based
541 vlan you can bridge to or handle like any other vlan connected to your
542 computer. The required drivers are already in the Linux kernel at
543 least since Debian Wheezy, and it is fairly easy to set up. A
544 <a href=
"http://www.open-mesh.org/projects/batman-adv/wiki/Quick-start-guide
">good
545 introduction
</a
> is available from the Open Mesh project. These are
546 the key settings needed to join the Oslo meshfx network:
</p
>
548 <p
><table
>
549 <tr
><th
>Setting
</th
><th
>Value
</th
></tr
>
550 <tr
><td
>Protocol / kernel module
</td
><td
>batman-adv
</td
></tr
>
551 <tr
><td
>ESSID
</td
><td
>meshfx@hackeriet
</td
></tr
>
552 <td
>Channel / Frequency
</td
><td
>11 /
2462</td
></tr
>
553 <td
>Cell ID
</td
><td
>02:BA:
00:
00:
00:
01</td
>
554 </table
></p
>
556 <p
>The reason for setting ad-hoc wifi Cell ID is to work around bugs
557 in firmware used in wifi card and wifi drivers. (See a nice post from
559 "<a href=
"http://tiebing.blogspot.no/
2009/
12/ad-hoc-cell-splitting-re-post-original.html
">Information
560 about cell-id splitting, stuck beacons, and failed IBSS merges!
</a
>
561 for details.) When these settings are activated and you have some
562 other mesh node nearby, your computer will be connected to the mesh
563 network and can communicate with any mesh node that is connected to
564 any of the nodes in your network of nodes. :)
</p
>
566 <p
>My initial plan was to reuse my old Linksys WRT54GL as a mesh node,
567 but that seem to be very hard, as I have not been able to locate a
568 firmware supporting batman-adv. If anyone know how to use that old
569 wifi access point with batman-adv these days, please let me know.
</p
>
571 <p
>If you find this project interesting and want to join, please join
572 us on IRC, either channel
573 <a href=
"irc://irc.freenode.net/#oslohackerspace
">#oslohackerspace
</a
>
574 or
<a href=
"irc://irc.freenode.net/#nuug
">#nuug
</a
> on
575 irc.freenode.net.
</p
>
577 <p
>While investigating mesh networks in Oslo, I came across an old
578 research paper from the university of Stavanger and Telenor Research
579 and Innovation called
580 <a href=
"http://folk.uio.no/paalee/publications/netrel-egeland-iswcs-
2008.pdf
">The
581 reliability of wireless backhaul mesh networks
</a
> and elsewhere
582 learned that Telenor have been experimenting with mesh networks at
583 Grünerløkka in Oslo. So mesh networks are also interesting for
584 commercial companies, even though Telenor discovered that it was hard
585 to figure out a good business plan for mesh networking and as far as I
586 know have closed down the experiment. Perhaps Telenor or others would
587 be interested in a cooperation?
</p
>
589 <p
><strong
>Update
2013-
10-
12</strong
>: I was just
590 <a href=
"http://lists.alioth.debian.org/pipermail/freedombox-discuss/
2013-October/
005900.html
">told
591 by the Serval project developers
</a
> that they no longer use
592 batman-adv (but are compatible with it), but their own crypto based
593 mesh system.
</p
>