How MIMO cross-layer design enables QoS while detecting non-cooperative nodes in wireless multi-hop networks

Abstract

In Journal of Network and Computer Applications (JNCA). DOI: 10.1016/j.jnca.2014.07.011International audienceWireless Multi-hop Networks (WMNs) are based on the cooperation between nodes. The non-cooperative (selfish) nodes can affect the quality of services (QoS) delivered by the network. The solutions proposed in literature are based on the monitoring mechanism to detect non-cooperative nodes. However, the monitoring mechanism has to tackle a significant false alarm rate. The origin of these issues is mainly related to the interferences and the costs of the monitoring mechanism. In WMNs based on Single-Input Single-Output (SISO) technology, the interferences at the monitor (detector) node can affect the assessment and the accuracy of the monitor node's observation. In this paper, we use Multi-Input and Multi-Output (MIMO) technology to tackle these drawbacks and to perform the monitoring mechanism without affecting the QoS. We propose a new MAC protocol based on the well-known SPACE-MAC protocol, named MIMODog. The collision at the monitor node can be avoided by tuning the antennas' weights. Therefore, the signal coming from other nodes than the monitored one can be nullified. Thus, this solution allows an important improvement of the accuracy of the monitor node's observation. Moreover, we propose a monitoring capacity analysis using graph theory particularly Conflict Graph (CG), and asymptotic study. We illustrate that the capacity consumed in the case of MIMODog is costly compared to SPACE-MAC, but the accuracy of the observation is better. We demonstrate that the number of monitor nodes is $\Theta(\frac{M}{\sqrt{n\ln n}})$ for a MIMO network with randomly located nodes n, each equipped with M antennas. Indeed, numerical results nlnn illustrate that by using MIMODog, the network can have a constant improvement M on an asymptotic number of monitor nodes compared to SISO 802.11 DCF MAC

Fully cooperative and distributed focal underdetermined system solver compressive sensing recovery algorithm for wireless sensor networks

“…18 Although several types of research on the CS and its applications have been investigated so far, the distributed sensing applications, including multiple signals, are less studied. Such conditions are more common in wireless sensor networks (WSNs), [19][20][21][22] where each node recovers the sparse signal using its measurements obtained using its own allocated measurement matrix.…”