Statistically optimal routing scheme in multihop wireless ad-hoc networks

Abstract

In this paper, we propose a novel distributed routing scheme in multihop wireless adhoc networks. We assume that the nodes are distributed according to a Poisson-Point-Process (PPP), and employ stochastic-geometric tools to derive a novel routing metric. The proposed Statistically-Optimal (SO) scheme, works in a decentralized manner in which the next relay selection exploits knowledge on the geographical location of each possible relay and on the instantaneous channel gain (fading diversity). The SO scheme maximizes the expected routing performance metric given the channel gains to the potential relays and their locations. Simulations results demonstrate that the SO scheme outperforms previously published schemes, and achieve performance which is close to the upper bound on ALOHA performance. Furthermore, the implementation of the suggested SO scheme is very simple, and the metric evaluation requires only minimal computations.

Optimal and Suboptimal Routing Based on Partial CSI in Random Ad-Hoc Networks

“…We also present three sub-optimal, low-complexity routing schemes that can be evaluated in a closed form. Finally, we show that the routing scheme of [16] (termed here narrow-bound-optimal (NBO) routing) can be viewed as an approximation of the SO routing [18]. Thus, we can compare the performance of all the routing schemes; we show that the NBO routing scheme is very close to the optimal routing (given only local information at each transmitting node).…”

“…The available knowledge of the probe transmitting node is given by M 0 , and the next hop selection is given by f (M 0 ) ∈ N 0 . The Asymptotic-Densityof-Rate-Progress (ADORP) performance metric is given by [16]:D…”

“…Conveniently, the ergodic rate is also easier for analysis. The network performance measure associated with the ergodic rate is termed the Asymptotic-Density-of-Rate-Progress (ADORP) [16]. The ADORP measures the average density of the product of the rate and the distance of each transmission.…”

“…The ADORP measures the average density of the product of the rate and the distance of each transmission. This gives a good indication of the capability of the network to deliver messages from sources to destinations (see more details in [16]). The ADORP is equivalent to the transport-capacity [17] except for the use of the ergodic rate instead of the outage rate.…”

“…Although [16] aimed to derive a routing metric that optimizes the network throughput in single antenna WANETs, the actual derivation ignored some of the data that are known in advance at the transmitting node. In this paper, we present the routing metric that maximizes the WANET throughput.…”