[Battlemesh] Battlemesh Digest, Vol 64, Issue 20

[Uros Platise]

>> Hi there,
>>
>> I am working on a new distributed Medium Access Layer - Protocol with
>> aim to overcome today's scalability problems of centralized topologies
>> and synchronous concepts as seen in internet and wifis.
>>
>> My vision and goal is to create a global scale mesh network, entirely
>> based on distributed approach, and to stimulate evolution of locally
>> running distributed services, which network is to be open and free.
> My first two questions would be "what hardware do you have in mind"
> and "are we talking about a wired or wireless MAC layer" ?
>
> Writing a good MAC layer for wireless medium that can deal with hidden
> stations is hard... getting hardware for it might be even harder.
>
>> If you are interested I can give a brief in few minutes in one of the
>> Lightening talks in Maribor, what's up in this project.
>>
>> I am also eager to see your testbench results in Maribor comparing
>> various routing layers, how far they can scale.
> I am not sure if I understand what you are saying.
>
> MAC layer is for getting access to the media. Routing layer is for
> forwarding IP packets.
> MAC layer has to deal with local density as a scalability issue,
> Routing layer with the total size of the network.
>
> How do you want to compare them?
>
> Henning
>
>

Yes it is about the MAC protocol which can deal with the hidden node
stuff, mobility, bandwidth allocation, multi-peer connections to support
advanced routing protocols, at the same time while working with either
low or high performance devices. The key thing is decentralization in
all aspects.

At the moment in our research phase we have 800/900 and 2.4 GHz custom
hw up to 0.5 Mbps, but not limited to, as design is to cover IoT
devices, which require very low bandwidth, and high-performance bulk
transfers, between notebooks, mobiles, etc. The design is also not
limited to wireless as some of the considerations are in progress to
deploy it on power line communications.

As our primary goal is to deliver the new MAC and then porting it to
support the hardware, on the other side we need to merge it with, say,
the best performance routing protocol, which is the topic of the
battlemesh, that's how I see the synergy. That's why I would like to see
which routing algorithms perform best, and can cover border cases.

Regards,
Uros