Adyton: A Network Simulator for Opportunistic Networks
Adyton is an event-driven network simulator, written in C++, for Opportunistic Networks (a.k.a. Delay-Tolerant Networks) that is capable of processing contact traces. The Adyton simulator supports a plethora of routing protocols and real-world contact traces, while also providing several congestion control mechanisms and buffer management policies.
COORD: A multi-copy routing protocol for Opportunistic Networks
The Coordinated Delegation Forwarding (COORD) allows each node to make replication decisions by taking into account not only its own perception of the highest utility in the network but also the highest utility as perceived by other nodes. In this way, nodes can coordinate their replication decisions in order to avoid unnecessary packet replications and thus achieve a more efficient replication process. COORD is generic in the sense that it can be implemented on top of any utility metric. Furthermore, COORD reduces the routing cost without sacrificing performance and that this is true for any utility metric.
RLDP: A RLNC-based broadcast protocol for MANETs
The Random Linear network coding over Dominant Pruning (RLDP) protocol combines random linear network coding (RLNC) with deterministic broadcasting. More specifically, the proposed algorithm implements CDS (Connected Dominating Set) based forwarding rules “on the generation level” in order to allow the flow of packet generations over the CDS. The rationale is that the CDS will provide a more systematic and topology-aware pruning of redundant transmissions without impairing the coding efficiency of RLNC. To distributively construct a CDS in order to broadcast messages RLDP uses the Partial Dominant Pruning (PDP) algorithm.
NOB-CR: A XOR-based broadcast protocol for MANETs
We introduce a new broadcast approach called Network COding Broadcast with Coding-based Redundancy (NOB-CR). Our approach incorporates new techniques in the coding operation that further enhance the broadcasting performance in terms of delivery delay, energy efficiency and network resources utilization.
CCAF: A congestion control algorithm for Opportunistic Networks
In collaboration with Dimitrios-Georgios Akestoridis.
The Congestion Control with Adjustable Fairness (CCAF) algorithm provides a tunable trade-off between efficiency and fairness. CCAF relies on the social preferences of the nodes for dynamically tuning the aforementioned trade-off. Furthermore, it incarnates the described functionality in a generic manner that can be incorporated to virtually any utility-based routing protocol for Opportunistic Networks.
CrossWorld: A new paradigm for producing large scale synthetic traces
In collaboration with Athanasios Koufoulis.
CrossWorld uses real human traces as building blocks and models the interaction of users from different blocks. In this way, we are able to model a higher degree of separation while at the same time preserve certain characteristics observed in real traces such as the distribution of inter-contact time, the distribution of contact duration and the clustering coefficient.