In this dissertation, we investigate the Markov chain analytical model, which is applied in three focuses of current interest: the distortion of video quality, the performance of directional MAC (D-MAC) protocols, and the performances of IEEE 802.11 wireless networks in the presence of various greedy behavior strategies.
First, to seek an innovative computable formula to evaluate objective video quality, we discover formulas to estimate the video quality distortion due to packet loss in IEEE 802.11 wireless networks. We also derive the expressions to estimate the video quality where some IEEE 802.11 wireless nodes manipulate their backoff schemes. Based on the formulas, numerical results of video quality are provided for different wireless network scenarios. An optimization of setting parameters of IEEE 802.11 to minimize the distortion is also discussed. As a conclusion, the formulas provide us a transition from setting parameters of IEEE 802.11 wireless networks to the objective video quality of wireless networks and allow us to compute the objective video quality if the parameters of IEEE802.11 wireless networks are given.
Second, we calculate the performance of a D-MAC protocol with a new Markov chain equivalently. As an example, our approach is applied to single-hop wireless networks where nodes are directional and distributed randomly. We obtain theoretical throughput and delays of each node in the D-MAC wireless networks. Numerical results show that the average throughput of wireless network, a 64-node network for example, will increase at a faster pace for the RTS/CTS access mechanism than the basic access mechanism while transmitting antenna elements increase from one to eight. Numerical results also show the performance of D-MAC wireless network depends on the parameters of D-MAC protocol and the number of antennae, as well as the topology.
Third, we provide a series of analytical formula to obtain the performance of the IEEE 802.11 wireless networks in the presence of various greedy behavior strategies. Unlike previous research papers, our improvement provides the performance of nodes where the nodes have different arrival rates under non-saturation conditions. It also provides analytical performances for multi-hop networks, networks with hidden nodes, and networks combining multiple greedy strategies.
Clearly each user may have different applications, and as a result users have different arrival rates. We improve the Markov chain analytical model so that the non-saturation Markov is able to compute performance of wireless networks whose stations have heterogeneous traffic in this dissertation. Our analysis with the Markov chain analytical model is validated in multiple scenarios with heterogeneous traffic, and is proved by simulations.