Differences between revisions 14 and 28 (spanning 14 versions)

Agenda

The agenda will be set after the workshop submission deadline, 1st of May. We will try to keep a maximum of 4 hours of talks per day, including breaks.

Proposed events

Talks

Time

Title

Abstract

Type

Speaker

Slides & Documentation

||One hour||Troubleshooting MTU problems on IPv4 and IPv6 networks|| Ninux Roma is a Native IPv6 network with legacy support to IPv4.
Usually we expect 1500 bytes of MTU and a boring NOC life. Unfortunately life is different. All kind of tunneling (VLAN, GRE, TINC, MPLS) are deployed on the Internet, eating bytes out of the MTU. Some mesh routing protocols use tunneling techniques themself. IPv4 packet fragmentation introduces bad performances but a working IPv4 end to end connectivity even if some link have a smaller MTU than needed. In the IPv6 Internet a MTU problem can blackhole your traffic. In this talk we will see some well known tools (ping6, radvd), and we will understand how to use them to check the end to end MTU.||Talk||Saverio Proto||Projector required||

Dynamic Frequency Selection in 5 GHz mesh networks

The 5 GHz band is an interesting alternative to the crowded 2.4 GHz ISM band to set up mesh networks. However, DFS (Dynamic Frequency Selection) support is required in many countries to avoid disturbing primary users (e.g. military or weather radar systems), and DFS support was not available in open source drivers until last year.

In this talk I'll present what DFS is and how it can be handled, summarise the current status in open source drivers and userspace tools, and discuss limitations and requirements to get it running in mesh networks.

Length: 60 minutes (30-40 minutes talk time and 10-15 minutes for questions)
Requirements: Projector.
Speaker: Simon Wunderlich

Hunting a digital fox at 5GHz

5GHz networks are generally deployed to cover larger distances and thus cover a wide area. Even a single or a few “evil” Wi-Fi transmitters may significantly degrade the performance of other devices and hinder a portion of the free spectrum useless. The talk is based on real events and will teach you the basic approach to analysing the spectrum at a given location and choosing a frequency, pinpointing the source of a wireless node performing a De-Auth attach on your node or generally spamming the spectrum, likely as a result of an infection with malicious code.

Length: 45min with 30min for the talk and 15min for discussion/demonstration.
Speaker: Musti

WiBed, a testbed platform for WiFi experiments

Wibed is a platform for facilitating the quick and cost-efficient acquisition, deployment, and manage-ment of testbeds based on commodity IEEE802.11 routers and enabling experimentation with wireless technology including the modification of low-level system components such as physical and link layer mechanisms, and network and transport layer protocols.

Format: lightening talk? Workhop? I don't know... Maybe a lightening talk is too short... but a complete talk is too big.
Length: Around 30 minutes.
Requirements: projector and some beer.
Proposed by Pau (p4u)

Network Coding for Mesh Networks

The talk is a follow up on Martin Hundeboll's talk at BattleMeshv6 about the potential of network coding in meshed networks. Based on some examples the importance of protocol design for network coding is highlightes. Furthermore the basics of network coding are repeated in a nutshell to give the audience the basic understanding.

Proposed by Frank Fitzek
Preferred day for the presentation: Friday afternoon

Nodeshot / CitySDK

A talk about the new version of nodeshot and the work that has been done on the CitySDK project regarding common APIs, smart participation / active citizenship.

Proposed by Frederico (ninux.org)

Netengine

Netengine is a new project started at the end of 2013. It is an abstraction layer to extract information from network devices. This talk will cover the basic concepts and explain the work being done on it plus the work the GSOC student will have accomplished before the event.

Proposed by Frederico (ninux.org)

Are Wireless Community Networks really decentralized networks?

The topology of three networks (Ninux, FF-Graz, and FF-Wien) have been analysed to try to understand:

Are they really decentralized topologies?
Are they robust to failures/attacks/interception?
For those that use OLSR, why don't use MPRs?

leonardo_wireless_decentralized.pdf Here is the original paper on this, limited to Ninux, presented it at CNBUB last year and at the Ninux-day in Rome. There is now much more data and some deeper analysis.

Proposed by Leonardo Maccari (researcher at the university of Trento, part of the Ninux network).
Preferred dates: Leonardo will be there from Wed to Sun, so thursday-friday-saturday are best.
Length: one hour.
Requirements: A projector.

Meetings

DIY ISP Meeting

Meeting of diy-isp from all over the world. We would like to continue our discussion that started at the 30C3 and FOSDEM and see how we can grow as community and share our knowledge about all the topics involved in changing the internet landscape.

Diyisp.org Project wiki, Pad with notes from the FOSDEM meeting, Pad with notes from the 30c3 meeting
Length: one to two hours
Requirements: a place for 15+ people and a projector would be nice
Proposed by: Philipp Borgers

BATMAN developers community meeting

Batman-adv users (wireless communities, commercial projects or other) may present their way of using batman-adv in real world setups to the present developers and batman users. The goal is to foster exchange about where and how batman-adv is used today, what are the challenges / difficulties while doing so and what can be done to further improve batman-adv.

Batman-adv Project wiki
Length: one to two hours
Requirements: a place for 15+ people and a whiteboard would be nice
Proposed by: Mark Lindner

CommunityCoin, a cryptocurrency for community networks

Based on the BitCoin block chain, the community coin (CCN) is a crypto-currency made for network Communities. The coin reward is not based on the block finding but in the contribution and participation of the community individuals.

There is nothing to present because there is nothing yet implemented. However it might be an interesting topic to discuss about to see what the community network members around the world think.

Format: panel discussion
Requirements: Projector. Definitely I will need some beer here
Proposed by Pau

Workshops

Wireless optical networking with KORUZA

After a brief introduction to wireless optical / free-space optical networking history and an overview of the technology you will be prompted to discuss ideas and play around with a low-cost 3D printable wireless optical system KORUZA (www.koruza.net) and get tutored into establishing 100m long 1Gbps links, so one the design files are released open-source and parts/units made available, you will be able to start deploying such links as well.

Length: 60min
Speaker: Musti

Amos sector antenna

Franz Streibl has proposed to give a talk / workshop about the new Amos sector antenna open-hardware design. He will offer a workshop of approximately one hour length where people can build this antenna from prefabricated elements for the 2.4 GHz WiFi band, if there is sufficient interest. We might break this up into a separate talk and workshop.

Paper about the antenna: piamos_72_140323m.pdf
Speaker: Franz Streibl

Social events

Thursday morning: Activity outside of the sublab and without laptops! More details to follow.

more to be announced/proposed.

Agenda in progress (not final)

Monday 12th May 2014