الفهرس 
AbstractOnly 14 pages are availabe for public view
 |  | from | 185 |  |  |
| | | from | 185 | | |
Abstract
A mobile ad hoc network is commonly referred to as MANET. MANET is a wireless network without infrastructure consisting of mobile nodes such as laptop computers, and smart phones. MANET is used in various applications including sensor networks and rescue and guidance systems in disaster situations. The mobility of the nodes, limited capacity (bandwidth) of the wireless channel, and limited power sources together with wireless transmission effects such as attenuation, and interference are combined to create significant challenges for MANET routing protocols. The main purpose of routing is to bring packets efficiently to their destination. There are several algorithms that address routing problem. All the previous work considers that their algorithms is better than the others using simulation with different metrics. This means that there is no a concise reference that can be used to measure the exactness of the proposed algorithm. Also, there is no agreement of the quality of service metrics. Since MANETs are suffering from low power battery, as soon as a node consumes its battery power, that node becomes a dead node. This node is not able to transmit or receive data. When all the network nodes have consumed their energy, then whole network becomes down. To solve these previous problems, the thesis works in two phases; the first one is to face the general MANET routing problem by introducing a new exact reference model based on combinatorial theory to find the shortest path (optimum route) from point of view of the exactness of the solution. Then the researcher modified the most famous routing model which is the Ant Colony Optimization algorithm by adding a pre-processing phase (training phase for parameters selection). The two models are compared using three different criteria. These criteria are the execution time, energy consumption, and the total cost. A simulation experiment is performed and the results show that the modified Ant Colony Optimization algorithm is superior in execution time but consumed more energy than the reference combinatorial and its total cost is greater than or equal to the reference one. The second phase is to face a subset of the MANET routing problem which is energy aware routing algorithms by introducing a survey in the form of taxonomy of the energy aware routing algorithms in MANET. This taxonomy is categorizing these algorithms into three types: 1. Modification of basic routing protocols, 2. Use of evolutionary algorithms, and 3.Use of linear programming. Then the thesis uses linear programming technique to solve the problem of finding a route satisfying the main objective of minimum energy consumption and other QoS requirements such as minimum delay and maximum packet delivery ratio. Two cases are considered; case 1: The traffic amount of a given request is transmitted into single path, and case 2: The traffic amount of a request can be distributed into parallel paths. A preprocessing step is performed first for network topology design. This step leads to formulate the first case as integer linear programming problem and the second case as linear programming problem instead of mixed integer linear programming problem. The two formulated problems are solved using Mat-lab and the obtained solutions are evaluated. The results show that the solution of second case is much better than the first case in terms of energy consumption and execution time. Packet delivery ratio in the second case is 100% while in the first case is only 76%. This thesis also compares the technique of ACO and linear programming in a specific simulation experiment which shows that linear programming is much better than ACO in both energy saving and processing time.