Recent advances in wireless technologies have helped the proliferation of wireless devices, ranging from hand-applications without resolving to complex resource provisioning mechanisms. Such an approach, however, is costly and resource inefficient. A compromised approach is to find techniques for increasing the network capacity without substantially changing the wireless network infrastructure. One promising approach is the recent development of network coding (NC) paradigm which has been shown to improve performance and efficiency of wireless networks. Potential benefits of network coding range from bandwidth and power efficiency to robustness and network dynamics. However, our current understanding on the optimal integration of network coding in the existing network protocols is rather limited. Furthermore, many NC benefits are often theoretically derived or obtained via simulations in idealized settings. To that end, the main scope of this dissertation aims at an in-depth understanding of network coding, its potential benefits, and trade-offs in typical real-world scenarios. The dissertation contributions can be summarized into three thrusts. In the first thrust, we consider single-hop wireless networks such as Wi-Fi or WiMAX networks, where the access point (AP) or base station (BS) has the ability to intercept and mix packet belonging to different flows from the Internet to multiple wireless users. We investigate a hybrid network coding technique to be used at a BS or AP to increase the throughput efficiency of the networks. Traditionally, to provide reliability, lost packets from different flows (applications) are retransmitted separately, leading to inefficient use of wireless bandwidth. Using the proposed hybrid network coding approach, the BS encodes these lost packets, possibly from different flows together before broadcasting them to all wireless users. In this way, multiple wireless receivers can recover their lost packets simultaneously with a single transmission from the BS. Furthermore, simulations and theoretical analysis showed that when used in conjunction with an appropriate channel coding technique under typical channel conditions, this approach can increase the throughput efficiency up to 3.5 times over the Automatic Repeat reQuest (ARQ), and up to 1.5 times over the HARQ techniques. In the second thrust, we investigate the achievable throughput for scenarios involving prioritized transmissions. Prioritized transmissions are useful in many multimedia networking applications where the transmitted data have an inherent hierarchy such that a piece of data at one level is only useful if all the pieces of data at all the lower levels are present. We investigate the achievable throughput of prioritized transmissions from a source to multiple receivers via a shared and lossy channel. In particular, we assume that the source is an oracle such that it knows precisely whether a packet is lost or received at any receiver in any future time slot, thus it can schedule the packet transmissions in such a way to maximize the receiver throughputs. We show that using network coding technique, the achievable throughput region for the broadcast scenarios can be substantially enlarged. Furthermore, for some erasure patterns, the achievable throughput using network coding technique is optimal in the sense that no scheme can do better. In addition, a class of approximate algorithms based on the Markov Chain Mote Carlo (MCMC) method have been proposed for obtaining the maximum sum throughput. Theoretical analysis and simulation results have been provided to verify the correctness and convergence speed of the proposed algorithms. In the third thrust, we propose a framework for adaptively optimizing the quality of service of multiple data flows in wireless access networks via network coding. Specifically, we consider scenarios in which multiple flows originate from multiple sources in the Internet and terminate at multiple users in a wireless network. In the current infrastructure, the wireless base station is responsible for relaying the packets from the Internet to the wireless users without any modification to the packet content. On the other hand, in the proposed approach, the wireless base station is allowed to perform network coding by appropriate linear mixing and channel coding of packets from different incoming flows before broadcasting a single flow of mixed or coded packets to all wireless users. Each user then uses an appropriate decoding method to recover its own packets from the set of coded packets that it receives. Theoretically, we showed that for the given channel conditions and QoS requirements, appropriate mixing and channel coding of packets across different flows can lead to substantial quality improvement for both real-time and non-real time flows. On the other hand, blind mixing can be detrimental. We formulate the mixing problem as a combinatorial optimization problem, and propose a heuristic algorithm based on the simulated-annealing method to approximate the optimal solution. Simulation results verify the performance improvement resulting from the proposed approach over the non-network coding and the state-of-the-art network coding approaches.