Mobility support in a tactical P2P publish/subscribe overlay

Subscriber mobility management for a P2P publish/subscribe overlay in mobile tactical networks

“…The prediction of the subscriber's next connected broker thus can be fairly accurate. Therefore, the work in [23] did not fully confirm the effectiveness of the proposed solution with the required scalability of practical MTNs. This key issue and several others are the focus of this paper, as indicated in Section 1.…”

“…There is a substantial difference in the handoff time of the proactive and reactive solutions. From the figure, the reactive approach's latency is by far the highest [23] for the two broker scenarios, the performance advantage of the proactive approach sustains, in some cases even increases, with more brokers in the network. Figure 7a-c depicts the performance comparison at different publishing rates, including metrics of message loss, duplication, and the overall throughput.…”

“…In [23], a proactive scheme was presented for the distinctive MTN environment. The key issue in the proactive based approach, i.e., the prediction of the next connected broker for a mobile subscriber was addressed with a concept of neighbor graph.…”

“…Such a scenario is limited compared to some bigger MTN environment (e.g., five to six broker nodes as discussed in later sections). Therefore, the work in [23] could not fully establish the scalability and applicability of the proposed proactive solution. We thus expand in this work a comprehensive evaluation of our proactive subscriber mobility management scheme, covering practical MTN scenarios involving multiple brokers and some worst case subscriber mobility models which are very challenging to the proactive scheme.…”

“…In [23], a novel P2P publish/subscribe mobility scheme based on a proactive approach was presented. However, in that work, the scenario considered in studying the core idea of the proactive algorithm involves only two brokers.…”