The IEEE 802.11 standard offers a cheap and promising solution for small scale wireless networks. Due to the self configuring nature, WLANs do not require large scale infrastructure deployment, and are scalable and easily maintainable which incited its popularity in both literature and industry. In real environment, these networks operate mostly under unsaturated condition. We investigate performance of such a network with m-retry limit BEB based DCF. We consider imperfect channel with provision for power capture. Our method employs a Markov model and represents the most common performance measures in terms of network parameters making the model and mathematical analysis useful in network design and planning. We also explore the effects of packet error, network size, initial contention window, and retry limit on overall performance of WLANs.
Vertical handoff plays an important role in guaranteeing users to be always connected in an overlapped multi-network environment. During the vertical handoff procedure, handoff decision is the most important step that affects the normal working of communication. An incorrect handoff decision or selection of a non-optimal network may result in undesirable effects such as higher costs, poor quality of service (QoS) experience, and even dropped communication. Among the existing vertical handoff decision algorithms, the Markov Decision Process (MDP) based algorithm by Stevens-Navarro et al. is promising due to its ability to achieve the optimal expected reward. However, the reward function used by this algorithm is flawed as it favors reducing expected number of vertical handoffs at the expense of diminished expected values of other QoS parameters. This paper presents an extended MDP based algorithm (EMDP) with novel reward function formulation. Analysis shows that EMDP outperforms the MDP based algorithm in terms of improved expected values of all QoS parameters considered while keeping the vertical handoff number reasonably low.