Large-scale network simulations are resource and time intensive tasks due to a number of factors i.e., setup configuration, computation time, hardware, and en ergy cost. These factors ultimately force network researchers to scale-down the scope of experiments, either in terms of Simulation Entities (SEs) involved or in abridging expected micro-level details. The Cloud technology facilitates re searchers to address mentioned factors by the provisioning of Cloud instances on shared infrastructure. In this thesis, an academic Cloud SIM-Cumulus targeting the research institutions is proposed. The thesis is divided into three parts, each part discussing the contributions achieved in thesis. The first part of this thesis discusses the design and implementation of SIM Cumulus academic Cloud framework for the provisioning of Network-Simulation as-a-Service (NSaaS). SIM-Cumulus provides the framework of Virtual Machine (VM) instances specifically configured for large-scale network simulations, with the aim of efficiency in terms of simulation execution time. The performance of SIM-Cumulus is evaluated using large-scale Wireless Network simulations that are executed sequentially as well as in parallel. Simulation results show that SIM-Cumulus is beneficial in two aspects i.e., (i) promotion of research within the domain of computer networks through configured Cloud instances of network simulators (ii) consumption of considerably fewer resources in terms of simula tion elapsed time and usage cost. The execution of simulation in parallel involves the partitioning of simulation model into several components and each compo nent is assigned to separate execution units (Logical Processes (LPs)). Each LP is comprised of a set of SEs that can interact with local as well as remote SEs. However, the remote communication among SEs and synchronization manage ment across LPs are the two main issues related to the parallel and distributed executions of large-scale simulations. A number of migration techniques are used to mitigate the problem of high remote communication and lead to a reduction in remote communication among SEs. However, most of the existing migration strategies result in higher number of migration which lead to higher computation overhead. The second part of the thesis contributes Migration-based Adaptive Heuristic Algorithm (known as MAHA). MAHA provides dynamic partitioning of the simulation model based on runtime dynamics of the wireless network simula tions. The proposed algorithm uses an intelligent heuristic for migration decision in order to reduce the number of migrations with an ultimate goal to achieve bet ter Local Communication Ratio (LCR). The proposed algorithm is better in terms of achieving improved LCR with reduced number of migrations as compared to the existing technique(s). The third contribution of this thesis is related to imple mentation of adaptive SIM-Cumulus (A-SIM-Cumulus) that integrates Advanced Runtime Infrastructure System (ARTIS) and Generic Adaptive Interaction Archi tecture (GAIA) with the SIM-Cumulus framework. To obtain an insight into the performance gain, the simulation has been performed multiple times with different configurations and execution environments. The obtained results assert that the proposed algorithm significantly reduces the number of migrations and achieves a good speedup in terms of execution time for parallel (i.e., both multi-core and distributed) simulations on the A-SIM-Cumulus Cloud.