3.1 EGOR Problem Formulation
3.1.1 Energy Consumption Model
Here we do not assume the promiscuous mode in which every node “overhears” the transmission within its range. Instead, being energy efficient, we assume nodes/sensors only listen to the transmissions intended for themselves. To achieve this, a second low-power radio (Vaidya and Miller 2005) can be used to wake up nodes that should participate in the EGOR or to inform the neighbors (including nodes giving negative advancement), which are not selected as forwarding candidates to shut down their data radios. Nodes can also only read the headers of packets for early rejection (Seada et al. 2004). For simplicity, we also ignore the energy consumption of the control packets,1 as control packets are usually much smaller than data packets. We only consider the energy consumption of packet transmission and reception.2 So the total energy consumption for one opportunistic forwarding attempt is:
where Etx and Erx are the packet transmission and reception energy consumption, respectively. Recall that r is the number of candidates in the forwarding candidate set .
3.1.2 Tradeoff Between EPA and Energy Consumption
As we proved in Lemma 2.1, the more nodes get involved in GOR, the larger the EPA can be. So the GOR that involves all ...