تجزیه و تحلیل عملکرد از توپولوژی ناآگاه از طرح های TDMA MAC برای شبکه های ad hoc با کنترل توپولوژی

Traditional omni-directional antennas result in increased mutliuser interference and are known to limit the performance of Medium Access Control (MAC) protocols for ad hoc networks. Topology control is the capability of a node to control the set of neighbor nodes. In this paper, the impact of using smart antennas and/or power control for topology control is investigated and the performance of TDMA MAC schemes with common frame, for which the assignment of time slots to a node is not aware of the time slots assigned to the neighbor nodes (topology-unaware), is studied. A comparison based on analytical models reveals the advantages of topology control, as well as its dependence on the mobility of the nodes and its resolution. Simulation results support the claims and the expectations of the analysis and show how the performance under topology control can be increased and how mobility affects it.

Ad hoc networks require no infrastructure and nodes are free to enter, leave or move inside the network without prior configuration, thus making the design of an efficient Medium Access Control (MAC) a challenging problem. CSMA/CA-based MAC protocols have been proposed [1], [2], [3], [4] and [5], as well as TDMA-based MAC protocols have also been proposed for ad hoc networks [6], [7], [8], [9] and [10]. Topology-Unaware TDMA MAC schemes, under which the assignment of time slots to nodes does not consider the time slots assigned to the neighbor nodes (nodes that a direct transmission is possible), have also been proposed [11], [12] and [15]. In particular, Chlamtac and Farago proposed the Deterministic Policy [11], whereas the Probabilistic Policy has been proposed and analyzed in [13], [14] and [15]. This analysis has shown that the Probabilistic Policy outperforms the Deterministic Policy under certain conditions. The aforementioned analysis was based on traditional omni-directional antennas, where the transmitting node did not have any topology control capabilities. Topology control is a node's capability of controlling the set of neighbor nodes and it may be achieved by adjusting the angle of the transmission beam and/or the transmission power and thus, control the interference caused to neighbor nodes when transmitting. The use of directional antennas for topology control is not a new idea and has been proposed in the past [16]. Nowadays, more sophisticated smart antennas can be used to adjust the angle of the transmission beam and even be incorporated into portable devices. Several MAC protocols have been proposed for ad hoc networks that exploit the capabilities of smart antennas. The majority of them is based on random access schemes (e.g. ALOHA or CSMA/CA) and enhancements of the RTS/CTS mechanism [17], [18] and [19]. Power control may also be used for topology control. The transmission power can be adjusted according to the location of the receiver, reducing the interference caused to neighbor nodes by the transmitting node [20], [21], [22], [23] and [24]. Resolution is an important factor of topology control. The higher the resolution of the topology control, the narrower the transmission beam of the smart antennas and/or the smaller the transmission range corresponding to a particular transmission. In this work, both the Deterministic Policy and the Probabilistic Policy are considered when topology control is applied (use of smart antennas and/or power control) and their performance is compared against that induced when no topology control is present (use of traditional omni-directional antennas). This comparison is based on an analytical approach and is supported by simulation results. The nodes' mobility is also taken into account, since it is expected to impact the performance especially under topology control. The (mobility) conditions under which topology control (for a given resolution) improves performance are also established here. Topology control is presented in Section 2. In Section 3, an ad hoc network is described and some key definitions are introduced. The Deterministic Policy and the Probabilistic Policy are presented in Section 4. In Section 5, expressions for the system throughput under both policies are derived with and without topology control. The mobility aspects are considered in Section 6, where the conditions, under which topology control with a certain resolution is beneficial for the system performance, are also established. Simulation results for network topologies with different characteristics are presented in Section 7. These results support the claims and the expectations introduced by the analytical comparison and show that the system throughput achieved under the Probabilistic Policy and under topology control can be rather high. On the other hand, it is shown that mobility degrades the system throughput especially under topology control with high resolution and therefore, topology control may not be desirable under certain conditions. Finally, Section 8 presents the conclusions.

In this paper, the performance of topology-unaware TDMA MAC policies (the Deterministic Policy and the Probabilistic Policy) is studied under sophisticated antennas and power control (topology control) and compared to that under the traditional omni-directional antennas (no topology control). The mobility factor is also taken into consideration and it is shown that topology control (although more vulnerable to mobility than under traditional omni-directional antennas) still improves the system performance under certain mobility conditions. Simulation results validate the claims and expectations. Initially, topology control is defined and its effect on node transmissions is investigated. In the sequel, both topology-unaware policies are presented and it is shown that, under topology control, increased system throughput is achieved when nodes are static. A discussion regarding mobility has shown that link failures are more likely under topology control than under no topology control, for given mobility conditions, leading to a possibly smaller system throughput. Analytical expressions based on average values are obtained and the conditions are established determining the mobility conditions under which topology control is beneficial. It is also shown that there exists a trade-off between the resolution of the topology control, the mobility and the system throughput achieved. Preliminary simulation results for an example network support the aforementioned arguments. Simulation results are obtained for four network topologies corresponding to different values of the topology density View the MathML source(|S|¯/D). The results demonstrate that the Probabilistic Policy outperforms the Deterministic Policy, even for dense topologies. When smart antennas are used, higher maximum system throughput is achieved and it is observed that the Probabilistic Policy is not so greatly affected by the topology density as under no topology control. However, the simulation results demonstrate the fact that under certain mobility conditions (high mobility) topology control may lead to smaller system throughput, depending on the resolution of the topology control as it is also indicated by the earlier analysis.