sven (dot) franck (at) nexedi (dot) com
We introduce in this presentation the progress of Nexedi as part of OSTV project sponsored by French Governement (GDN framework).
The presentation has 4 parts.
We remind the goals of the project.
We then provide a status report.
We demonstrate results through a video.
We make a list of missing steps to finalise.
Automate the whole process of deployment and management of a base station and backhaul so that anyyone can become a telco within a few minutes (and be rewarded).
Nexedi's role is automate all processes from system configuration to billing (or clearing) so that anyone can deploy a 4G (and 5G) network within a few minutes.
This is supported by two components: OSTV Cloud and OSTV Node.
SlapOS Cloud has dramatically progressed. It supports now about every feature needed to deploy a commercial telecom network. It was presented at MWC. We demonstrated a complete system based on open source components, include smartphone OS powered by /e/ foundation.
SlapOS Node also dramatically progressed. Everything is now automated: installation, deployment, configuration, monitoring, etc. of eNodeB, EPC, edge services, accounting, etc. Layer-2 eNodeB-to-eNodeB peering and Layer-3 Mobile IP are still not integrated though.
Let us see how easy it is to deploy a node. Basically, we add a node on SlapOS cloud and SlapOS cloud tells us what to do on about any GNU/Linux distribution.
In sequence, the video shows:
We reused an existing video footage for the monitor. This video footage needs to be replaced with Amarisoft stack for consistency.
This video is an evolution of tutorials that we written in 2018, which includes all instructions to setup the NMS Master. To install OSTV Cloud, the command would be:
wget http://deploy.erp5.net/slapos-master-standalone; bash slapos-master-standalone
The script we launched automates the installation of Ansible on any GNU/Linux distribution (Debian, Centos, Ubuntu, SuSE, etc.).
It then launches an Ansible profile which configures base system and in particular packages (re6st, babel, slapos-node) and network interfaces.
Then, it configures the backaul IPv6 using re6st and babel.
Then it registers OSTV Node to OSTV Cloud using SLAP protocol.
From that point, user can request installation of any service: eNodeB, EPC, SIM database, edge HTTP server, etc.
Each service can be instantiated as many times as needed. A single OSTV Node can thus run multiple instances of multiple SDR stacks at the same time.
Overall, what still needs to be completed is:
For Layer-2 Peering, we need to integrate results of Milestone 7 and tasks: SO-OSTV-30-10-10 and SO-OSTV-30-10-20.
For Layer-3 IP mobility, we need to integrate Paris 7 results which were just completed a few days ago, and a few days after MWC 2019.
For Frequency marketplace, which is actually a condition for legal deployment by verifying that a frequency can be licensed for a given area, we wil get the help from a company in Finland called Fairspectrum. Sweden and nordic countries have a nice way of managing frequencies. If a frequency is not used somewhere by a telco, as part of the auction attribution, citizens may use it to provide the service which the telco does not provide. It would be good to do the same in France. This would solve once and for all the problem of uncovered areas (France has less LTE coverage than many third-world countries).
We need an RRH with open source drivers, else it creates huge security or compatiblity issues with Linux kernel. An Ethernet RRH would be ideal.
We need also a field test which can use all technologies at once.
We believe that 6 months is not enough to achieve this so it would be better to extend the project. One year is probably reasonable.