| Recently, network coding technology is an important breakthrough in the network and communication technologies. It innovates the traditional mode of data transmission and provides a new way to improve the transmission efficiency of wireless communication protocols. The unreliability of the wireless link and physical layer broadcast nature has laid a good foundation to exploit the performance advantages of network coding in a wireless network. However, the implementation of wireless network coding has to cope with many challenges such as limited bandwidth, the variance of topology, high packet loss ratio. Currently, the key research problem for network coding in wireless Ad Hoc networks is how to figure out a reasonable point to combine the advantages of wireless networks and network coding. Amongst all possible technical routines, the design of efficient MAC mechanism to support network coding in wireless Ad Hoc networks is one of the feasible solutions.At first, this thesis highlights DCF access mechanism and the backoff mechanism, both of which are utilized to enable wireless terminal access the shared channel. Markov Chain model is employed to characterize the IEEE802.11DCF protocol, and the system saturation throughput of X-shaped and random network is derived and analyzed. Numerical results are presented to unveil that, in the X-shaped network setup, the higher transmission data rate, the longer the length of the packet, or the smaller backoff window will give rise to a better saturation throughput performance. While in random network setup, the increased probability of collisions with the increase in the number of wireless nodes will lead to the degradation in the achieved system throughput. Moreover, it is shown that, if the number of wireless nodes in wireless network is small enough, the minimum backoff window parameter will have negligible influence on the achieved throughput performance. However, if the minimum backoff window is small, the probability of the same backoff time selected between different sites will be enlarged with the increase in the number of nodes, which implies the increased probability of collisions and the degradation in the system throughput peformance.Secondly, opportunistic network coding mechanism and its encoding/decoding algorithm are reviewed. The performance analysis model of MAC protocol with opportunistic network coding is addressed as well based on the Markov chain model. Then the coding opportunities at intermediate nodes are investigated, and the saturation throughput of the DCF protocol before and after the introduction of the opportunistic network coding are compared and analyzed. It is validated that the coding opportunities at intermediate node will be decreased when a larger threshold value is employed. Meanwhile, it is shown that a too small threshold value will also cause system perfonnance degradation. In order to determine the reasonable choice of the threshold value, this thesis introduces the node capacity improvement (NCI) parameter. And our analysis shows that the choice of the optimal threshold value has nothing to do with the node coding opportunities. Moreover, the transmit power of all the involved nodes should be carefully set since it will not affect the achieved system throughput as long as it is larger than a certain value. In addition, this dissertation analyzes the impact of the buffer size of all wireless nodes on the network coding opportunities and system throughput. The analysis results show that the greater the cache, the better the performance. However, when the buffer size is large enough, the average encoding number and the throughput remain substantially constant.Finally, simulation platforms over the OPNET are presented to enable the numerical simulations of MAC protocol with opportunities network coding. The paper introduces the function of each module at node layer and process layer, respectively. Then the protocol of IEEE802.11DCF and the opportunistic network coding-aware MAC protocol is realized base on this platform. Simulation results of the two protocols are analyzed for both X-shaped and random network in order to verify the numerical performance analysis results. And the simulation results are presented to validate that, DCF protocol with opportunistic network coding can be utilized to effectively improve network throughput, to reduce the number of transmissions of intermediate nodes and to reduce the transmission delay in wireless communication network. |
PDF Full Text Request