Comparison of Random, Temporal, Spatial, Geographic, and Trace-Based Mobility Models in Mobile Ad Hoc Networks

Every ad hoc network protocol focuses on a different set of measurements and traits. As a result, a protocol may function well in a setting with high mobility, few nodes, and low connectivity, but poorly in one with many nodes and low mobility. The majority of the currently utilized mobility models are employed to provide realistic movement patterns for MANET scenarios. Flexible testing with various topologies, mobility patterns, and physical and link-layer protocols is provided by the simulation. Since genuine systems are difficult to get, expensive, complex, or even dangerous (e.g., space simulations, aviation simulations), simulation offers a quick design option. Complex functions that cannot be evaluated with closed-form formulas or numerical methods are evaluated through simulation. The future topology of a network can be predicted, which enables Quality of Service (QoS) aware routing to choose a dependable link for data transmission between two points. The Random Waypoint (RWP) and Manhattan Grid can achieve more Packet Delivery Ratio (PDR) in Ad hoc On-Demand Vector (AODV) and Destination-Sequenced Distance Vector (DSDV) routing protocols. With the exception of Truncated Levy Walk (TLW), a delay for all mobility models is practically 0 ms for AODV and DSDV. Manhattan Grid outperforms competing mobility models in the case of throughput by 50% across all routing protocols. Due to its smooth intersection turns and halt durations, TLW has lower overhead than RWP, Gauss–Markov, and Manhattan Grid. Due to source routing feature, Dynamic Source Routing (DSR) with TLW offered a lower drop rate.