4.3 Rate and Candidate Selection Schemes
How to select the transmission rates and forwarding strategy for each node efficiently so that the network capacity can be globally optimized is still an open research issue. We have shown the example in Figure 4.3 that nodes transmitting at a lower rate may lead to a higher end-to-end throughput than when nodes are transmitting at a higher rate. Then, what criteria should node a follow to select transmission rate, forwarding candidates and candidate priority to approach the capacity? It is nontrivial to answer this question. Towards the development of distributed and localized OR protocol that maximize the capacity, in this section, we propose two rate and candidate selection schemes, one is enlightened by least-cost opportunistic routing (LCOR) as proposed in Dubois-Ferriere et al. (2007), and the other is inspired by geographic opportunistic routing (GOR) (Fussler et al. 2003; Zeng et al. 2007a,b,c; Zorzi and Rao 2003).
4.3.1 Least Medium Time Opportunistic Routing
In traditional routing, the medium time metric (MTM) (Awerbuch et al. 2006) and expected transmission time (ETT) (Draves et al. 2004) have shown to be good metrics to achieve high throughput. For OR, we define the opportunistic ETT (OETT) as the expected transmission time to send a packet from ni to any node in its forwarding candidate set .
where Lpkt is the packet length, ...