To Forward or not to Forward – that is the Question

Abstract

We introduced the use of two-hop forwarding to increase the throughput of an 802.11 network in our earlier work (Narayanan et al., Proceedings of IEEE WCNC’05, March 2005). Other researchers have also considered the benefits of forwarding in the 802.11 infrastructure mode to increase the total network throughput. But the high-data rate node that forwards data for other nodes will have to spend its energy transmitting this data. Previous work on forwarding implicitly assumed that in an enterprise network, the collective good is sufficient to justify this increased energy expense. However, it is important to address the advantages and the cost of participating in such schemes from the individual forwarding node’s perspective. Since a node cannot know whether there are other high-data rate nodes in the network capable and willing to forward data, it needs to assume that it is the only node with the capability to do so. In this paper, we focus our analysis on the cost benefit for such a forwarding node. We quantify the throughput improvement, medium access delay reduction and energy consumption for the forwarding node in a saturated network. The analysis and simulation results demonstrate that in a saturated network, participation in forwarding provides the high-data rate node with significant benefits in throughput and media-access-delay, while increasing the number of bits-per-joule even if it is the only node participating in data forwarding as suggested in this paper. The increase in the bits-per-joule is due to the reduction in the total amount of time needed by the high data rate node to transmit a given number of its own application bits. This results in savings in energy expenditure for the forwarding node. Based on these benefits, we conclude that it is unequivocally in the interest of a high data rate node to participate in two-hop forwarding schemes in 802.11 networks.