Research On Multi-Rate MAC Protocol In Ad Hoc Networks

With the development of wireless communication technonogy, mobile Ad Hoc networks are becoming more and more attractive owing to the unique characteristics. Since the IEEE 802.11 standard supports multiple data rates at the physical layer, many rate adaptation mechanisms have been proposed to utilize this multi-rate capability. The performance of wireless network is considerably influenced by a phenomenon known as performance anomaly. In the recent years, many new approaches have been proposed to improve the performance of multi-rate wireless network by solving performance anomaly. However, when dealing with performance anomaly, traditional mechanisms have some limitations.In a multirate wireless network, if low-rate nodes initialize their CWmin only based on their own data rate, unnecessary backoff leads to performance degradation. Aming at this problem, a new solution is presented in this paper known as LevelCW (CW adaptation based on nodes'competitive level). The proposed mechanism allows competitive nodes in two hops share each other's data rate information with the help of ACK packet and select appropriate CWmin based on the ratio of their own data rate to their competitive nodes'data rate.In Ad Hoc networks, data rates of competitive nodes can't reflect their paths'rate information. In the proposed mechanism known as PathCW (CW adaptation based on path's competitive level), nodes acquire path's end-to-end data rate by following a cross-layer design approach. Competitive node calculates its path's competitive level, which represents the whole path's data transmission capability, and uses it as the criteria for allocating the channel. Nodes on the path with higher competitive level can have a more time share of channel.Some of the traditional multi-rate MAC and both of the proposed mechanisms are implemented on the network simulator NS-2 to investigate the performance. Simulaiton results show that compared with some traditional multi-rate MAC protocols, the proposed protocol can increase the aggregate throughput while decreasing the average end-to-end delay, because it can not only solve the performance anomaly in multi-hop networks but also improve the overall performance by avoiding unnecessary back-off.