Asma Ben Hassouna

A model for deploying an opportunistic MAC protocol in NS-2

Multiuser diversity refers to a kind of diversity present across different users in a fading environment. This diversity can be exploited by scheduling transmissions so that users transmit when their channel conditions are favorable. Exploiting multiuser diversity access, called also opportunistic access, conducts to a system efficiency that rises with the number of users. Different opportunistic IEEE 802.11 Medium Access Control (MAC) approaches have been proposed. The efficiency study of these approaches by NS-2 simulation is necessary for a performance comparison. The implementation of any opportunistic IEEE 802.11 MAC approach in NS-2 is based on using different cross-layer interactions between the physical layer and the link layer components which are the interface queue and the MAC sub layer. In this paper, we introduce the work that has been done to create a model for implementing these interactions. We also give the different functions we have implemented in NS-2 on the basis of this model. Any opportunistic approach can thus be implemented using these functions. Our contribution is providing an NS-2 extension that can be used to implement protocols exploiting multiuser diversity. We explain how to use our NS-2 extension by giving two examples showing the implementation of new opportunistic MAC protocols.

A low-complexity and high-throughput multi-user diversity based multicast scheme for MANETs

We propose a new multicast access scheme that is founded on a metric that allows high-throughput multicast in MANETs. The new access metric, called MOST (as Maximum Opportunistic Scheduling scheme Throughput), is computed for each multicast session and utilized to find the optimal number of transmissions and the best set of transmission data rates to use in order to achieve the highest multicast session throughputs. The new multi-user diversity (MUD) [1] based multicast scheme exploits the wireless broadcast advantage feature [2] and the available channel capacities to get the transmission scheme that offers the highest multicast session throughput without need to know the throughput values of all other possible transmission schemes. The multicast scheme is modeled using a transition tree that is the basis of the metric calculation. This tree is exploited to minimize the number of capacity combinations to inspect before getting the best one. Results prove that the proposed access scheme based on MOST metric achieves throughputs visibly higher than the other proposed schemes.

An opportunistic IEEE 802.11 medium access control protocol

In a wireless network, the signals transmitted between a source station and different receiver users most often have different channel fluctuation characteristics. This diversity that exists between users is named multiuser diversity (MUD) and can be exploited to improve the capacity of wireless networks. One way of exploiting MUD is by opportunistic scheduling of users, i.e. giving priority to users having the best channel conditions. To be able to take advantage of the MUD, a feedback protocol has to be deployed to notify the source station about the Carrier to Noise Ratio (CNR) or the channel gain of the mobile user with the finest channel conditions. In this paper, we are inspired from the splitting algorithm [1] to provide IEEE 802.11 with an opportunistic scheduling. We describe the new opportunistic MAC protocol and we give an overview of its implementation in NS-2. We show that the simulation of the splitting algorithm gives the same average feedback time as the analysis study. We compare the native IEEE 802.11 scheduling to the opportunistic one and prove that exploiting MUD increases the network capacity. We also show that this increase is higher with increasing the number of users. Finally, we studied the influence of the observed number of packets in the queue and we prove that maximizing this number does increase the network capacity and the medium access fairness.

Extreme throughput multicast in multi-user diversity wireless networks

While low transmission delays can meet the requirements of many applications, high throughput is the appeal of several others. Accordingly and taking into account the multi-user diversity (MUD) and the wireless broadcast advantage (WBA) features, we will address the problem of providing the maximum throughput while guaranteeing an acceptable transmission delay. We give an overview of the new rate-aware routing metric for multicast: EMTH. For each ongoing transmission, by doing rate adaptation and taking into consideration the link reliabilities, EMTH helps to choose the transmission rates that approximate the maximum multicast throughput expected. Results show that the proposed metric gives the best system performance and that it is possible to trade-off a very good throughput improvement for an insignificant delay increase.

Multi-rate interference-sensitive and conflict-aware multicast in wireless ad hoc networks

The broadcast nature of the wireless medium makes multicast communication subject to various challenges, especially the unreliability due to the interference [2] and the impact of the transmission data rate choice on the conflicts between communicating users. In fact, a fundamental trade-off exists between communication speed (transmission data rate) and communication range. Actually, the effect of interference is more important when the communication speed decreases, i.e. when communication range increases. In this paper, we propose a multiple rate multicast scheme that is applied to capture the effect of transmission conflicts on the wireless multicast throughput. This work exploits the diversity between users to provide an accurate and efficient method that enables each multicast transmitter, i.e. forwarder or sender, to select the data rates to use to serve its interested neighbours. The choice of the set of data rates, i.e. the choice of multi-rate multicast scheme, should be conflict sensitive in order to guarantee high multicast throughput in multi-rate multi-hop MANETs. We start by introducing two new concepts: The Transmission data Rate based Interference graph (TRIGraph) and Concurrent Multi-rate Multicast Transmitter set (CMMS). Then, we describe the use of these concepts to characterize the interference conflicts caused by multi-rate multicast transmissions. Unlike all the existing conflict graphs, TRIGraph and CMMS are not only used to model interference conflicts, but they are also used to choose the multicast data rates that reduce the effect of such network inconsistency on the system performance.

A high-throughput routing for multiuser diversity multicast

In this paper, we introduce a new routing strategy that allows high-throughput and channel-aware multi-rate multicast in wireless networks. This new routing strategy, based on the so-called End-to-end Multicast Session Efficiency (EMSE), is founded on the multi-user diversity (MUD) and the wireless broadcast advantage (WBA) features. It aims at: i) joining users that provide the best-received multicast throughput, ii) guaranteeing the use of homogenous capacities, iii) considering the effect of the transmission/retransmission number needed to exploit the multi-rate diversity on the routing throughput and iv) constructing the multicast structure composed of paths that give the best end-to-end throughput. Compared to other routing strategies and multicast schemes, the new routing algorithms show considerable improvement in terms of the average throughput and an important reduction in the relays number needed to convey the multicast data.

A generic conflict-aware multicast routing overlay for wireless networks

Multicast routing is a fundamental issue in wireless networks and related routing strategies should select the lowest number of parallel conflicting transmissions. However, current works on this problem accounts only for conflicts occurring within the transmission ranges of the nodes. Particularly, no work is interested in studying conflicting transmissions while using multi-rate concurrent multicast transmissions. In this paper, we develop a conflict-aware routing model, which aims at increasing the probability of scheduling conflict-free transmissions and improving the throughput by extremely exploiting the network availabilities in terms of link capacities, using different rates. The suggested conflict-aware routing strategy is implemented by means of a set of greedy heuristic algorithms. This strategy, called the CAware routing strategy, is “generic” because it can be coupled with any throughput-based routing metric or any routing structure. The proposed algorithms are studied to justify their performance trends and to show that conflict-awareness gives considerable routing benefits.

Multicast throughput subject to delay in multi-rate wireless networks

Exploiting multi-user diversity (MUD) means to assign the resources to the user experiencing the best channel conditions. When multicast communication is considered, MUD means that a sender can serve the intended receiver users using one or many data transmission rates. In fact, MUD-based multicast schemes aiming to maximize the multicast throughput (e.g. DOMS and MOST) or to minimize the multicast delay (e.g. EMTT) are shown to significantly outperform other multicast schemes. In this paper, we suggest a distributed, stateless and multi-user diversity-based multicast scheme; the Multicast scheme for Opportunistic Throughput subject to Delay (MOTD), which is utilized to study the intrinsic trade-off between the multicast throughput and the multicast delay in multi-rate networks. Then, we study the performance of the previously mentioned multicast schemes and we compare them to MOTD. The evaluated features are the throughput and the transmission delay. Mainly, MOTD registers very good throughput values that are constrained by moderated transmission delays.

PEMSE: A high-throughput multicast routing protocol for multi-rate IEEE802.11

In the context of multi-rate communication, the throughput-based multicast routing that aims at maximizing the end-to-end throughput is very different from the commonly adopted delay-based multicast routing that aims at minimizing the end-to-end delay. Definitely, the delay-based approach is not suitable for applications that need to transfer large amounts of data (e.g. grid applications). In view of that, we introduce in this paper the first distributed routing strategy that allows high-throughput multicast in multi-rate IEEE802.11 networks. The new mesh-based routing, called the Protocol for End-to-end Multicast Session Efficiency (PEMSE), exploits the rate-diversity (MUD) and on the broadcast-nature of the wireless medium (WBA). It aims at: i) considering the effect of the number of transmissions/retransmissions, required to exploit the MUD, on the end-to-end multicast throughput and ii) establishing the routing structure that guarantees the best end-to-end multicast throughput. Simulations prove that PEMSE achieves considerable throughput improvement compared to the PUMA [11] routing protocol. They also show that PEMSE attains reasonable delays and delivery ratios.

A throughput-oriented reliable multicast in multi-rate wireless networks

As low transmission delays can meet the requirements of many applications, high-throughput is the request of various others. Therefore, with consideration of the multi-user diversity (MUD) and the wireless broadcast advantage (WBA) features, we deal with the problem of maximising the multicast throughput. Hence, we introduce two new rate-aware multicast metrics: EMTH and XMTH. EMTH helps to choose the transmission rates that estimate the best multicast throughput expected. XMTH can reach extreme throughput values if the high computational needs are satisfied. Both metrics carry out a rate adaptation process and take into consideration the link reliabilities. They can be used to choose multi-rate multicast schemes and to decide reliable multicast paths. Results show that the proposed metrics give the best system performance in terms of throughput. They confirm that, using EMTH and XMTH, it is possible to trade-off a very good throughput improvement for an insignificant delay increase.