Search

Abstract: In mobile ad hoc networks (MANETs), the IEEE 802.11 DCF protocol is commonly used at the medium access control (MAC) layer to reduce collision and contention. However, mobile nodes, before send out data, need to wait for other in-process data transmission in the same range to be completed. In this paper, we address this performance degradation problem in the DCF mechanism and propose an overlapping and pipelining data transmission MAC protocol to enhance throughput. The protocol regulates the actions of the neighboring nodes of both the sender and the receiver and enables the utilization of the same channel simultaneously by these neighboring nodes. A multiple sub-channel technique is proposed to avoid the interference problem caused by the control and data signals of different communication mobile node pairs. With the regulated actions and the multiple channel technique, multiple pairs of mobile nodes are provided with the desired feature of utilizing the same data sub-channel simultaneously without interfering with one another. Although the overlapping data transmission technique can increase the number of simultaneous communication pairs, an efficient bandwidth allocation of the control and data sub-channels to maximize utilization is critically important. Therefore, we further propose a bandwidth allocation strategy in which the control signals and data signals are pipelined. Performance simulation results prove the efficiency of the proposed bandwidth allocation strategy with better channel utilization. In addition, the proposed MAC protocol with the bandwidth allocation strategy, combined with the overlapping and pipelining data sub-channel technique, outperforms the IEEE 802.11 DCF in throughput (kB/s) by about 43.76%. [Copyright &y& Elsevier]

In the design of fault-tolerant interconnection networks, disjoint paths and dynamic rerouting are often used to tolerate faults. Although providing two more disjoint paths between any source and destination pair could increase reliability, it might not help to decrease collision rate. Rather than using disjoint paths, the method of dynamic rerouting finds an alternative path to the destination when a packet encounters a faulty or busy element. To look after both sides of fault-tolerance and performance, the following items must be carefully concerned to design a dynamic rerouting network. The issues including fault-tolerant capability, low collision rate (throughput) and rerouting hops should be considered. In this paper, we propose a dynamic rerouting network, which can provide an alternative path when a packet meets a faulty or busy element. Besides, the network costs 0-rerouting hops to find the alternative path and performs better static characteristics than other related networks. From the simulation results, the simulation results show that our network has less collision rate and a better throughput than other fault-tolerant networks. [Copyright &y& Elsevier]