Application de réseaux Mesh dans un milieu confiné
|Authors:||Moutairou, Manani M.|
|Advisor:||Delisle, Gilles Y.|
|Abstract:||This thesis focuses on an optimization problem of the deployment of Wireless Mesh Network in an underground mine environment. The deployment of the network’s devices required analytical and experimental studies of the mining area. The experimental study allowed us to better define the topology of the area, to model the received signal power with statistics and especially to study the radio coverage of the network’s elements according to their location in the mine. Moreover, multihop and multiradio wireless mesh networks’ behaviour in the area is also addressed. To elaborate, the problem of the limitation of the total available capacity is due to the number of relays (hops) the packets are subjected to in the network. In regards to the analytical study, three approaches of the network deployment in the underground area were proposed. These different topological approaches bring different results in the network depending on the performance, the robustness and the total reduction of expenses. The first topological approach consists in finding one or several optimal positions of the gateway which allows us to reduce the congestion of the WMN network in very dense traffic periods. In a confined area, the topology of the network is closely linked to the geometric shape of the area due to overused established links (based on the adopted routing protocol) that compromise the total performance of the network. The second topological approach manages the problem of position of a gateway in a local way by forming tree based independent clusters. It requires prior studies on the behaviour of the network in the deployed area. These studies will allow us to determine the size of the tree that is necessary to unfold the installation of the network by clustering the elements of the network in the area. The size of the sub-network tree will closely take into account the traffic density in the area. This second approach reduces, and even resolves the problem of congestion that can occur in the first approach. Each cluster is strictly sized in such a way that it allows the disposal of a minimum resource at the level of users in the network whatever their positions are. It also answers the question of flexibility looked after in the underground area that will allow us to unfold network only in the location where needed. This flexibility of the network allows mining industries to make huge savings while purchasing network equipments. The most significant concept introduced is the algorithm “Competitive Greedy (CG)” that allows us to manage this network approach. Competitive Greedy algorithm requires more work than existing ones, but it does offer the best solution regarding the total cost of the network (optimum resolutions which allow us to reduce network expenses). These performances will thoroughly be introduced and this algorithm will allow us to manage the size of the cluster, the numbers of relays that each element of the network can support, and finally the maximum number of hops (network depth) which separates the clusters’ element with its associated gateway. The third topological approach is a little bit more expensive compared to the first two approaches because of the number of gateways required to assure the robustness of the WMN. It resolves vulnerability problem of the network in the way gateways are deployed according to the first two approaches. In fact, this third approach is very important as it is possible that the network can expose different failure and possible accidents in the environment that can cause severe security problem to the users.|
|Document Type:||Thèse de doctorat|
|Open Access Date:||16 April 2018|
|Collection:||Thèses et mémoires|
All documents in CorpusUL are protected by Copyright Act of Canada.