CODYSUN - Collaborative and Dynamic Approaches to Increasing Spectrum Utilisation

The project proposes a generalized finite state machine (FSM) that executes on various system entities (earth stations in motion, gateways, satellites) for avoiding interference while guaranteeing the priority of the co-locating terrestrial network and the GSO segment. This FSM requires information in the form of spectrum monitoring, node position, antenna characteristics, etc, that operate within the same system (intra-system) as well as entities operating in other systems (inter-system). This information is shared through proposed intra-system or inter-system control channels depending on whether the entities communicate within or outside the system.

The important event for the steady state is interference detection and this interference can be detected by a variety of ways such as signal to noise plus interference ratio (SINR) degradation, spectrum monitoring, link statistics, etc. The interference detection module is monitoring all active links when the system is operating in the steady state and in case of detecting an interference event, the interference management techniques, local DSS enablers, spectrum decision blocks are executed whose solve job is to move the system back to the steady state. Three DSS techniques (i) DSS1 – collaboration protocol, (ii) DSS2 – decentralized sensing and (iii) DSS3 – collaboration protocol and decentralized sensing and are inspired by the dynamic spectrum sharing solutions that have been successfully validated by team SCATTER in the DARPA Spectrum Collaboration Challenge (SC2). [1]

[1] https://www.darpa.mil/program/spectrum-collaboration-challenge

Results

An extensive simulation study was done for a designed simulation scenario and a number of key performance indicators such as throughput, packet error rate, uptime, etc. amongst others, were calculated for the three DSS techniques. The defined KPIs were used to benchmark the proposed DSS techniques against the base line (fixed spectrum allocation, status quo) in order to reach a conclusion on whether DSS techniques can help to enhance spectrum usage efficiency while minimizing resulting interference.

The average system uptime achieved in each case is presented below, taking into account that the scenarios were all designed to trigger multiple interference events between the communication entities based on frequency usage, 3D space positioning and antenna radiation patterns employed from each system.

As it can be seen, from an achieved ~80% uptime in the baseline scenario, where the systems cannot use frequency agility to avoid interference due to dynamic mobility of the entities present in the scenario, we managed to achieve 95-99% when employing any of the DSS techniques. There are still possible enhancements that can be used to reach the optimum 100% uptime in any case, which will be part of our future work.

To conclude, our work proved that fixed spectrum allocation is no longer needed since there are techniques that can bestow the frequency selection process to the systems themselves, making it possible to avoid interference in most if not in all cases, without the hassle of bookkeeping of frequency allocation. The added value is that techniques as the proposed ones come also with the added advantages of extreme scalability and minimum economic overhead once established.