home > Projects >
The goal of this project is to investigate fundamental properties and limitations of a large population of Terminodes, and to propose solutions to maximize the throughput (or, more generally, the quality of service) experienced by the users of a Terminode network. We model the network at an appropriate level of abstraction, which incorporates realistic constraints on the physical and MAC layer (interferences, power consumption, bit error rate) and on the network and application layers (circuit or packet switching, real-time or "elastic" traffic). The objectives of this individual project within the NCCR are both specific and general. The specific objective is to develop a theoretical basis for the various choices and options that are taken to design a Terminode network. The general objective is to develop methods and to obtain results that go far be-yond the Terminode network itself and leads to a knowledge base that will be a reference for many specific developments in this fast growing field, not only in mobile communication, but in stochastic geometry and self-organizing systems at large.
Connectivity under the Boolean model
Our current effort is devoted to the one of the most basic properties of a Terminode network: connectivity. Currently, most research results address the full connectivity of a network over a finite area, with a number of nodes tending to infinity. In this work instead, we let the area tend to infinity, but we keep the node density finite. We study the property of connectivity for both a purely ad-hoc network and a hybrid network, where fixed base stations can be reached in multiple hops. We assume here that power constraints are modeled by a maximal distance (the distance above which two nodes are not directly connected) : this enables use percolation theory for the Boolean model. We found that the introduction of a sparse network of base stations does significantly help in increasing the connectivity, but only when the node density is much larger in one dimension than in the other.
// More information here
Connectivity under the physical model
Up to now, we have used a very simple model, the Boolean model, where two nodes are connected to each other if and only if their distance is less than some value. In reality, interferences need to be taken into account. We consider therefore a model (the "physical model") where two nodes are connected to each other if and only if the signal to noise ratio at the receiver (noise is here the sum of a background noise, and of the interference contributions from neighboring nodes) is larger than some prescribed value. The problem is more complex, as now direct connections between two particular nodes depend on other nodes' positions, contrary to the Boolean model. Moreover, two parameters (coefficient of orthogonality of the code coping with interferences, emitting power) instead of one (power) are now in the picture. We are currently investigating the existence of a percolation phenomenon in this model, in particular the behavior of the percolation threshold with the parameter weighting the interferences contribution (the coefficient of orthogonality of the code) and the differences with the Boolean models.
// More information here
Self-organized Medium Access Control
The role of a Medium Access Control (MAC) protocol is to coordinate the access of the Terminodes to the wireless channel. In particular, a good MAC protocol should avoid collisions and schedule as many concurrent transmissions as possible in order to guarantee a high network throughput.
In multi-hop ad hoc network, the lack of a central authority renders the design of efficient MAC protocols complex and experimental studies on existing protocols often report poor performance. Our interest lies in the design of MAC protocols that can globally organize the transmissions in space using only the limited information available at each Terminode.
We identified two mechanisms: (i) the reaction-diffusion mechanism (more information here)
and (ii) the backoff mechanism (more information will be available shortly), that are able (under certain conditions) to create spatially efficient transmission schedules.
Vulnerability of ad hoc and sensor networks
Fault management (this term encompasses the detection and location
of the failure(s), as well as the protection mechanism to restore the traffic
around the failed facility) is essential in both the optical and wireless
networks, although the issues involved are quite different. This project is
out under grant DICS 1830 of the Hasler Foundation.
// More information will be available shortly.